USING SI UNITS IN AS TRONOMY

A multitude of measurement units exist within astronomy, some of which are

unique to the subject, causing discrepancies that are particularly apparent when
astronomers collaborate with other disciplines in science and engineering. The
International System of Units (SI) is based on a set of seven fundamental units
from which other units may be derived. However, many astronomers are reluctant
to drop their old and familiar systems. This handbook demonstrates the ease with
which transformations from old units to SI units may be made. Using worked exam-
ples, the author argues that astronomers would benefit greatly if the reporting of
astronomical research and the sharing of data were standardized to SI units. Each
chapter reviews a different SI base unit, clarifying the connection between these
units and those currently favoured by astronomers. This is an essential reference
for all researchers in astronomy and astrophysics, and will also appeal to advanced
students.

ri chard do d d has spent much of his astronomical career in New Zealand,

including serving as Director of Carter Observatory, Wellington, and as an Hon-
orary Lecturer in Physics at Victoria University of Wellington. Dr Dodd is Past
President of the Royal Astronomical Society of New Zealand.
U S I N G S I UNIT S
I N A S T RONOMY

RICHARD DODD
Victoria University of Wellington
 c a m b r i d g e u n ive r s i t y p r e s s
Cambridge, New York, Melbourne, Madrid, Cape Town,
Singapore, São Paulo, Delhi, Tokyo, Mexico City
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org
Information on this title: www.cambridge.org/9780521769174

© R. Dodd 2012

This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.

First published 2012

Printed in the United Kingdom at the University Press, Cambridge

A catalogue record for this publication is available from the British Library

Library of Congress Cataloguing in Publication data

Dodd, Richard.
Using SI units in astronomy / Richard Dodd.
p. cm.
Includes bibliographical references and index.
ISBN 978-0-521-76917-4 (hardback)
1. Communication in astronomy. 2. Metric system. I. Title.
QB14.2.D63 2011
522’.87–dc23 2011038728

ISBN 978-0-521-76917-4 Hardback

Cambridge University Press has no responsibility for the persistence or

accuracy of URLs for external or third-party internet websites referred to in
this publication, and does not guarantee that any content on such websites is,
or will remain, accurate or appropriate.
 Contents

Preface page ix
Acknowledgements xii
1 Introduction 1
1.1 Using SI units in astronomy 1
1.2 Layout and structure of the book 2
1.3 Definitions of terms (lexicological, mathematical
and statistical) 3
1.4 A brief history of the standardization of units in general 7
1.5 A brief history of the standardization of scientific units 8
1.6 The future of SI units 11
1.7 Summary and recommendations 11
2 An introduction to SI units 12
2.1 The set of SI base units 12
2.2 The set of SI derived units 12
2.3 Non-SI units currently accepted for use with SI units 13
2.4 Other non-SI units 14
2.5 Prefixes to SI units 14
2.6 IAU recommendations regarding SI units 20
2.7 Summary and recommendations 23
3 Dimensional analysis 24
3.1 Definition of dimensional analysis 24
3.2 Dimensional equations 25
3.3 Summary and recommendations 29
4 Unit of angular measure (radian) 30
4.1 SI definition of the radian 30
4.2 Commonly used non-SI units of angular measure 30

v
vi Contents

4.3 Spherical astronomy 36

4.4 Angular distances and diameters 46
4.5 Steradian 48
4.6 Summary and recommendations 53

5 Unit of time (second) 54

5.1 SI definition of the second 54
5.2 Definition of time 54
5.3 Systems of time or time scales 54
5.4 The hertz: unit of frequency 60
5.5 Angular motion 60
5.6 The determination of the ages of celestial bodies 66
5.7 Summary and recommendations 70

6 Unit of length (metre) 72

6.1 SI definition of the metre 72
6.2 Linear astronomical distances and diameters 72
6.3 Linear motion 83
6.4 Acceleration 88
6.5 Area 89
6.6 Volume 89
6.7 Summary and recommendations 91

7 Unit of mass (kilogram) 92

7.1 SI definition of the kilogram 92
7.2 The constant of gravitation 94
7.3 Masses of astronomical bodies 97
7.4 Density 106
7.5 Force 108
7.6 Moments of inertia and angular momentum 109
7.7 Summary and recommendations 111

8 Unit of luminous intensity (candela) 113

8.1 SI definition of the candela 113
8.2 Radiometry and photometry 113
8.3 Magnitudes 137
8.4 Summary and recommendations 142

9 Unit of thermodynamic temperature (kelvin) 146

9.1 SI definition of the kelvin 146
9.2 Temperature scales 147
9.3 Some examples of the temperatures of astronomical objects 149
 Contents vii

9.4 Blackbody radiation 151

9.5 Spectral classification as a temperature sequence 154
9.6 Model stellar atmospheres 165
9.7 Summary and recommendations 172
10 Unit of electric current (ampere) 174
10.1 SI definition of the ampere 174
10.2 SI and non-SI electrical and magnetic unit relationships 175
10.3 Magnetic fields in astronomy 183
10.4 Electric fields in astronomy 194
10.5 Summary and recommendations 195
11 Unit of amount of substance (mole) 197
11.1 SI definition of the mole 197
11.2 Avogadro’s constant and atomic masses 197
11.3 Astrochemistry and cosmochemistry 202
11.4 Summary and recommendations 204
12 Astronomical taxonomy 206
12.1 Definition of taxonomy 206
12.2 Classification in astronomy 206
12.3 Classification of stellar objects 207
12.4 Classification of Solar System objects 215
12.5 Astronomical databases and virtual observatories 216
12.6 Summary and recommendations 218
References 219
Index 226
 Preface

Other than derogatory comments made by colleagues in university physics

departments on the strange non-standard units that astronomers used, my first
unpleasant experience involved the Catalog of Infrared Observations published
by NASA (Gezari et al., 1993). In the introduction, a table is given of the 26 dif-
ferent flux units used in the original publications from which the catalogue was
compiled – no attempt was made to unify the flux measures. The difficulties of
many different ways of expressing absolute and apparent flux measures when try-
ing to combine observations made in different parts of the electromagnetic spectrum
became all too apparent to me when preparing a paper (Dodd, 2007) for a conference
on standardizing photometric, spectrophotometric and polarimetric observations.
This work involved plotting X-ray, ultraviolet, visible, infrared and radio frequency
measurements of selected bright stars in the open cluster IC2391 as spectra with
common abscissae and ordinates. Several participants at the conference asked if
I could prepare a ‘credit card’ sized data sheet containing the conversion expres-
sions I had derived. As is usually the case, I was otherwise engaged at the time in
comparing my newly derived coarse spectrophotometry with a set of model stellar
atmospheres, so the ‘credit card’ idea was not acted upon. However, the positive
response to my paper did make me realize that there was a need in the astronomical
community for a reference work which, at the least, converted all the common
astronomical measurements to a standard set. The answer to the question ‘Which
set?’ is fairly self evident since it was over 40 years ago that scientists agreed upon
a metric set of units (Le Système International d’Unités or SI units) based on three
basic quantities. For mass there is the kilogram, for length the metre and for time
the second. This primary group is augmented by the ampere for electric current, the
candela for luminous intensity, the kelvin for temperature and the mole for amount
of substance. From these seven it is relatively easy to construct appropriate physical
units for any occasion: e.g., the watt for power, the joule for energy or work, the
newton for force and the tesla for magnetic flux density.

ix
x Preface

The thirteenth-century Mappa Mundi illustration of the Tower of Babel. (© The

Dean and Chapter of Hereford Cathedral and the Hereford Mappa Mundi Trust.)

It is possible to express even the more unusual astronomical quantities in SI

units. The astronomical unit, the light year and the parsec are all multiples of the
metre – admittedly very large, and non-integral, multiples from ∼150 billion metres
for the astronomical unit to ∼31 quadrillion metres for the parsec. Similarly, one
solar mass is equivalent to ∼2 × 1030 kilograms and the Julian year (365.25 days)
to ∼31.6 million seconds. So in each of these cases we could use SI units, though
quite obviously many are unwieldy and a good scientific argument for using special
astronomical units may readily be made.
In many areas of astronomy, the combination of research workers trained initially
at different times, in different places and in different disciplines (physics, chem-
istry, electrical engineering, mathematics, astronomy etc.) has created a Babel1 -like
situation with multitudes of units being used to describe the same quantities to the
confusion of all.
Astronomers participate in one of the most exciting and dynamic sciences and
should make an effort to ensure the results of their researches are more readily
available to those interested who may be working not only in other branches of
astronomy but also in other fields of science. This can be done most readily by
using the internationally agreed sets of units.

1 The story of the Tower of Babel is set out in the Hebrew Bible in the book of Genesis, chapter 11, verses 1–9,
and relates to problems caused by a displeased God introducing the use of several different, rather than one
spoken language, to the confusion of over-ambitious mankind. A depiction of the Tower of Babel that appears
on the thirteenth-century Mappa Mundi in Hereford Cathedral Library is shown in the reproduction above.
 Preface xi

However, stern, but sensible, comments from the reviewers of the outline of the
proposed book, plus a great deal more reading of relevant astronomical texts on my
part, has led to a better understanding of why some astronomers would be reluctant
to move away from non-standard units. This applies particularly in the field of
celestial mechanics and stellar dynamics, where the International Astronomical
Union approved units include the astronomical unit and the solar mass. However,
this in itself should not act as a deterrent from adding SI-based units alongside the
special unit used, with suitable error estimates to illustrate why the special unit is
necessary.
In a recent book review in The Observatory, Trimble (2010) admitted to append-
ing an average of about two corrections and amplifications per page in not only the
book she had just reviewed but also in her own book on stellar interiors (Hansen
et al., 2004), and Menzel (1960) completed the preface to his comprehensive work
Fundamental Formulas of Physics by stating: ‘In a work of this magnitude, some
errors will have inevitably crept in.’
Whilst, naturally, I hope that this particular volume is flawless, I must confess
I consider that to be unlikely! The detection and reporting of mistakes would prove
of considerable value and, likewise, comments from readers and users of the book
on areas in which they believe it could be improved would be welcome. My own
experience using various well-known reference works and textbooks, to some of
which I had previously assigned an impossibility of error, was that they all contained
mistakes; some travelled uncorrected from one edition to the next and others in
which correct numerical values or terms in an algebraic equation in an earlier
edition were incorrectly transcribed to a later.
The most radical suggestions in this book are probably: a simple way of describ-
ing and dealing with very large and very small numbers; the use of a number pair of
radians rather than a combination of three time and three angular measures to locate
the position of an astronomical body; and the replacement of the current ordinal
relative-magnitude scheme for assigning the brightness of astronomical bodies by
a cardinal system based on SI units in which the brighter the object the larger the
magnitude.
Writing a book such as this takes time. Time during which new values of astro-
nomical and physical constants may become available. I have referenced the various
sources of constants published before the end of 2010 that were used in the prepara-
tion of tables and in the worked examples presented. Readers are invited to substitute
later values for the constants, as a valuable exercise, in the worked examples should
they so wish.
In conclusion, it is important to bear in mind that the primary purpose of this book
is to act as a guide to the use of SI units in astronomy and not as an astronomical
textbook.
 Acknowledgements

It is a pleasure to thank the following organizations for permission to reproduce

illustrations and text from their material.
The Bureau International des Poids et Mesures (BIPM) for permitting the use of
the English translations of the formal definitions for each of the SI units and some
of the tabular material contained in the 8th edition of the brochure The International
System of Units (SI) (BIPM, 2006).2
The Dean and Chapter of Hereford Cathedral for permission to use a print of
part of the Mappa Mundi that shows an imagined view of the Tower of Babel.
The Canon Chancellor of Salisbury Cathedral for permission to use their
translation from the Latin of clause 35 of the Magna Carta.
Writing a book such as this has benefitted considerably from the availability of
online data sources. Those which were regularly consulted included: the Astrophys-
ical Data Service of NASA; the United States Naval Observatory for astrometric
and photometric catalogues; the European Southern Observatory for the Digital
Sky Surveys (DSS) and the HIPPARCOS and TYCHO catalogues; SIMBAD for
individual stellar data; the Smithsonian Astrophysical Observatory for DS9 image
analysis software, and many of the other databases and virtual observatory sites
listed in Chapter 12.
At an individual level, the inspiration to start this work is due in part to: Mike
Bessel, Ralph Bohlin, Chris Sterken, Martin Cohen and other participants at the
Blankenberge conference on standardization who expressed an interest in the paper
I presented there. Denis Sullivan of the School of Chemical and Physical Sciences
of the Victoria University of Wellington provided the enthusiasm and logistical sup-
port to continue with this work, and with Mike Reid was responsible for improving
my limited skill with LATEX. Harvey McGillivray, formally of the Royal Observatory

2 Please note that theses extracts are reproduced with permission of the BIPM, which retains full internationally
protected copyright.

xii
 Acknowledgements xiii

Edinburgh, provided me with COSMOS measuring machine data of the double clus-
ter of galaxies A3266. The desk staff at Victoria University of Wellington library
and the librarian of the Martinborough public library were of great assistance in
sourcing various books and articles. The proofreading was bravely undertaken by
Anne and Eric Dodd. To all these people I express my thanks.
My aim was to write a book that would prove of use to the astronomical commu-
nity and persuade it to move towards adopting a single set of units for the benefit
of all. I hope it succeeds!
 1
Introduction

1.1 Using SI units in astronomy

The target audience for a book on using SI units in astronomy has to be astronomers
who teach and/or carry out astronomical research at universities and government
observatories (national or local) or privately run observatories. If this group would
willingly accept the advantages to be gained by all astronomers using the same
set of units and proceed to lead by example, then it should follow that the next
generation of astronomers would be taught using the one set of units. Since many
of the writers of popular articles in astronomy have received training in the science,
non-technical reviews might then also be written using the one set of units. Given
the commitment and competence of today’s amateur astronomers and the high-
quality astronomical equipment they often possess, it follows that they too would
want to use the one set of units when publishing the results of their research.
As to why one set of units should be used, a brief search through recent astronom-
ical literature provides an answer. Consider the many different ways the emergent
flux of electromagnetic radiation emitted by celestial bodies and reported in the
papers listed below and published since the year 2000, is given.

Józsa et al. (2009) derived a brightness temperature of 4 × 105 K for a faint

central compact source in the galaxy IC2497 observed at a radio frequency
of 1.65 GHz.
Bohlin & Gilliland (2004), using the Hubble Space Telescope to produce absolute
spectrophotometry of the star Vega from the far ultraviolet (170 nm) to the
infrared (1010 nm), plotted their results in erg . cm−2 . s−1 . Å−1 flux units.
Broadband BVRI photometric observations, listed as magnitudes, were made
by Hohle et al. (2009) at the University Observatory Jena of OB stars in two
nearby, young, open star clusters.
In the study of variable stars in the optical part of the spectrum it is quite common
to use differential magnitudes where the difference in output flux between

1
2 Introduction

the variable object of interest and a standard non-varying star is plotted against
time or phase (see, e.g., Yang, 2009).
An X-ray survey carried out by Albacete-Colombo et al. (2008) of low-mass
stars in the young star cluster Trumpler 16, using the Chandra satellite, gives
the median X-ray luminosity in units of erg . s−1 .
The integral γ -ray photon flux above 0.1 GeV from the pulsar J0205 + 6449 in
SNR 3C58, measured with the Fermi gamma-ray space telescope, is given in
units of photons . cm−2 . s−1 by Abdo et al. (2009).

These are just a few examples of the many different units used to
specify flux. Radio astronomers and infrared astronomers often use janskys
(10−26 W . m2 . Hz−1 ), whilst astronomers working in the ultraviolet part of
the electromagnetic spectrum have been known to use flux units such as
(10−9 erg . cm−2 . s−1 . Å−1 ) and (10−14 erg . s−1 . cm−2 . Å−1 ). So it would seem
not unreasonable to conclude that whilst astronomers may well be mindful of SI
units and the benefits of unit standardization they do not do much about it.
Among reasons cited in Cardarelli (2003) for using SI units are:

1. It is both metric (based on the metre) and decimal (base 10 numbering system).
2. Prefixes are used for sub-multiples and multiples of the units and fractions
eliminated, which simplifies calculations.
3. Each physical quantity has a unique unit.
4. Derived SI units, some of which have their own name, are defined by simple
expressions relating two or more base SI units.
5. The SI forms a coherent system by directly linking the mechanical, electrical,
nuclear, chemical, thermodynamic and optical units.

A cursory glance at the examples given above shows numerous routes to possi-
ble mistakes. Consider the different powers of ten used, especially by ultraviolet
astronomers. Some examples use wavelengths, some frequencies, and some ener-
gies to define passbands. One uses a form of temperature to record the flux detected.
In short, obfuscation on a grand scale, which surely was not in the minds of the
astronomers preparing the papers. For this book to prove successful it would need
to assist in a movement towards the routine use of SI units by a majority, or at the
very least a large minority, of astronomers.

1.2 Layout and structure of the book

The introductory chapter (1) contains the reasons for writing the book and the target
audience, definitions of commonly used terms, a brief history of the standardization
of scientific units of measurement and a short section on the future of SI units.
 1.3 Definitions of terms (lexicological, mathematical and statistical) 3

Descriptions of the base and common derived SI units, plus acceptable non-SI units
and IAU recommended units, are listed in Chapter 2 with Conférence Général des
Poids et Mesures (CGPM) approved prefixes and unofficial prefixes for SI units
with other possible alternatives.
Given the importance of the technique known as dimensional analysis to the
study of units, an entire chapter (3) is allocated to the method, including worked
examples. There are further examples throughout the book that illustrate the value
of dimensional analysis in checking for consistency when transforming from one
set of units to another.
Eight chapters (4–11) cover the seven SI base units plus the derived unit, the
radian. Each includes the formal English language definition published by the
Bureau International des Poids et Mesures (BIPM) and possible future changes
to that definition. Examples of the uses of the unit are given, including transforma-
tions from other systems of units to the SI form. Derived units, their definitions, uses
and transformations are also covered, with suitable astronomical worked examples
provided. Each chapter ends with a summary and a short set of recommendations
regarding the use of the SI unit or other International Astronomical Union (IAU)
approved astronomical units.
The book ends with a chapter (12) on astronomical taxonomy, outlining various
classification methods that are often of a qualitative rather than a quantitative nature
(e.g., galaxy morphological typing, visual spectral classification).
The subject matter of the book covers almost all aspects of astronomy but is not
intended as a textbook. Rather, it is a useful companion piece for an undergradu-
ate or postgraduate student or research worker in astronomy, whether amateur or
professional, and for the writers of popular astronomical articles who wish to link
everyday units of measurement with SI units.

1.3 Definitions of terms (lexicological, mathematical

and statistical)
The meaning of a word is, unfortunately, often a function of time and location and
is prone to misuse, rather as Humpty Dumpty said in Through the Looking Glass,
‘When I use a word, it means just what I chose it to mean – neither more nor less.’3
When discussing a subject such as the standardization of units, it is of paramount
importance to define the terms being used. Hence, words that appear regularly
throughout the book related to units and/or their standardization are listed in this
section with the formal definition, either in their entirety or in part, as given in

3 See Carroll L. (1965). Through the Looking Glass. In The Works of Lewis Carroll. London: Paul Hamlyn, p. 174.
4 Introduction

volumes I and II of Funk & Wagnalls New Standard Dictionary of the English
Language (1946).

1.3.1 Lexicological and mathematical

Unit
Any given quantity with which others of the same kind are compared for the pur-
poses of measurement and in terms of which their magnitude is stated; a quantity
whose measure is represented by the number 1; specifically in arithmetic, that num-
ber itself; unity. The numerical value of a concrete quantity is expressed by stating
how many units, or what part or parts of a unit, the quantity contains.

Standard
Any measure of extent, quantity, quality, or value established by general usage and
consent; a weight, vessel, instrument, or device sanctioned or used as a definite
unit, as a value, dimension, time, or quality, by reference to which other measuring
instruments may be constructed and tested or regulated.
The difference between a unit and a standard is that the former is fixed by
definition and is independent of physical conditions, whereas a standard, such as
the one-metre platinum–iridium rod held at the Bureau International des Poids et
Mesures (BIPM) in Sèvres, Paris, is a physical realization of a unit whose length
is dependent on physical conditions (e.g., temperature).

Quantity (Specific)
(1) Physics: A property, quality, cause, or result varying in degree and measurable
by comparison with a standard of the same kind called a unit, such as length,
volume, mass, force or work.
(2) Mathematics: One of a system or series of objects having only such relations,
as of number or extension as can be expressed by mathematical symbols; also,
the figure or other symbol standard for such an object. Mathematical quantities
in general may be real or imaginary, discrete or continuous.

Measurement
The act of measuring; mensuration; hence, computation; determination by judge-
ment or comparison. The ascertained result of measuring; the dimensions, size,
capacity, or amount, as determined by measuring.
The mathematical definition of a quantity Q, is the product of a unit U , and a
measurement m, i.e.,
Q = mU (1.1)
 1.3 Definitions of terms (lexicological, mathematical and statistical) 5

Q is independent of the unit used to express it. Units may be manipulated as

algebraic entities (see Chapter 3) and multiplied and divided.

Dimension
Any measurable extent or magnitude, as of a line, surface, or solid; especially one of
the three measurements (length, width and height) by means of which the contents
of a cubic body are determined; generally used in the plural. Any quantity, as length,
time, or mass, employed or regarded as a fundamental factor in determining the
units of other physical quantities (see Chapter 3); as, the dimensions of velocity are
length divided by time. The dimension of a physical quantity is the set containing
all the units which may be used to express it, e.g., the dimension of mass is the set
(kilogram, gram, pound, ton, stone, hundredweight, grain, solar mass . . .).

Accurate
Conforming exactly to truth or to a standard; characterized by exactness; free from
error or defect; precise; exact; correct.

Accuracy
The state or quality of being accurate; exactness; correctness.

Precise
Having no appreciable error; performing required operations with great exactness.

Precision
The quality or state of being precise; accuracy of limitation, definition, or
adjustment.

There is a tendency to use accuracy and precision as though they had the same
meaning, this is not so. Accuracy may be thought of as how close the average value
(see below) of the set of measurements is to what may be called the correct or actual
value, and precision is a measure (see standard deviation below) of the internal
consistency of the set of measurements. So if, for example, a measuring instrument
is incorrectly set up so that it introduces a systematic bias in its measurements, these
measures may well have a high internal consistency, and hence a high precision,
but a low accuracy due to the instrumental bias.

Error
The difference between the actual and the observed or calculated value of a quantity.
6 Introduction

Mistake
The act of taking something to be other than it is; an error in action, judgement or
perception; a wrong apprehension or opinion; an unintentional wrong act or step;
a blunder or fault; an inaccuracy; as a mistake in calculation.

1.3.2 Statistical
Statistics is an extensive branch of mathematics that is regularly used in astronomy
(see Wall & Jenkins, 2003). As a very basic introduction to simple statistics, defi-
nitions are given for the terms mean and standard deviation, which are commonly
used by astronomers and an illustration (Figure 1.1) of the Gaussian or normal
distribution curve showing how a set of random determinations of a measurement
are distributed about their mean value.

Mean
Consider a set of N independent measurements of the value of some parameter x
then the mean, or average, value, μ, is defined as:

1
N
μ= xi (1.2)
N
i=1

0.4

0.3

σ σ
f (x)

0.2

0.1

0.0
–3 –1 μ 1 3
x

Figure 1.1. A Gaussian distribution with μ = 0 and σ = 1 generated using equation

(1.4).
 1.4 A brief history of the standardization of units in general 7

Standard deviation
Given the same set of N measurements as above, the standard deviation, σ , is
defined as:
1 √
N
σ= (xi − μ)2 (1.3)
N
i=1

Gaussian distribution
For large values of N , the expression describing the Gaussian or Normal distribution
of randomly distributed values of x about their mean value μ is:
 
1 −1
(x−μ)2
f (x) = √ e 2σ 2 (1.4)
σ (2π)
Figure 1.1 shows the typically bell shaped Gaussian distribution with a mean value,
μ = 0, and standard deviation, σ = 1.
Occasionally published papers may be found that use expressions such as stan-
dard error or probable error. If definitions do not accompany such expressions then
they should be treated with caution, since different meanings may be attributed by
different authors.

1.4 A brief history of the standardization of units in general

The history of the development of measurement units is well covered in many
excellent books that range from those for children, such as Peter Patilla’s Measuring
Up Size (2000) and the lighthearted approach of Warwick Cairns in About the Size
of It (2007), to the scholarly and comprehensive Encyclopaedia of Scientific Units,
Weights and Measures by François Cardarelli (2003), and Ken Alder’s detailed
account of the original determination of the metre in the late eighteenth century,
The Measure of All Things (2004).
Everyday units in common use from earliest times included lengths based on
human anatomy, such as the length of a man’s foot, the width of a hand, the width
of a thumb, the length of a leg from the ground to the hip joint, and the full extent
of the outstretched arms. Greater distances could be estimated by, e.g., noting the
number of paces taken in walking from town A to town B. Crude standard weights
were provided by a grain of barley, a stone and a handful of fruit. Early measures
of dry and liquid capacity used natural objects as containers, such as gourds, large
bird eggs and sea shells. Given that many such units were either qualitative or
dependent on whose body was being used (e.g., King Henry I of England decreed
in 1120 that the yard should be the distance from the tip of his nose to the end of
8 Introduction

his outstretched arm), trading from one village to another could be fraught with
difficulties and even lead to violent altercations.
One of the earliest records of attempted standardization to assist in trade is set
out in the Hebrew Bible in Leviticus, 19, 35–36 (Moffatt, 1950): ‘You must never
act dishonestly, in court, or in commerce, as you use measures of length, weight, or
capacity; you must have accurate balances, accurate weights, and an honest measure
for bushels and gallons.’
Around 2000 years later, King John of England and his noblemen inserted a
clause in the Magna Carta (number 35 on the Salisbury Cathedral copy of the
document) that stated (in translation from the original abbreviated Latin text): ‘Let
there be throughout our kingdom a single measure for wine and for ale and for corn,
namely: the London quarter,4 and a single width of cloth (whether dyed, russet or
halberjet)5 namely two ells within the selvedges; and let it be the same with weights
as with measures.’
It would appear to be very difficult to introduce a new set of standard units
by legislation. Even the French, under Emperor Napoleon I, preferred a mainly
non-decimal system, which had more than 250 000 different weights and measures
with 800 different names, to the elegant simplicity of the decimal metric system.
This preference caused Napoleon to repeal the act governing the use of the metric
system (passed by the republican French National Assembly in 1795, instituting
the Système Métrique Décimal) and allowing the return to the ancien régime in
1812. The metric system finally won out in 1837 when use of the units was made
compulsory. One hundred and sixty years later it was the turn of the British to object
to the introduction of the metric system, despite such a change greatly simplifying
calculations using both distance and weight measurements.

1.5 A brief history of the standardization of scientific units

With the beginnings of modern scientific measurements in the seventeenth century,
the scientists of the time began to appreciate the value and need for a standardized
set of well-defined measurement units.
The first step towards a non-anthropocentric measurement system was proposed
by the Abbé Gabriel Mouton, who in 1670 put forward the idea of a unit of length
(which he named the milliare) equal to one thousandth of a minute of arc along
the North–South meridian line. Mouton may fairly be considered the originator of
the metric system, in that he also proposed three multiple and three submultiple
units based on the milliare but differing by factors of ten, which were named by

4 The London quarter was a measure that King Edward I of England decreed, in 1296, to be exactly eight striked
bushels, where ‘striked’ implied the measuring container was full to the brim.
5 ‘halberjet’ is an obsolete term for a type of cloth (Funk et al., 1946).
 1.5 A brief history of the standardization of scientific units 9

adding prefixes to milliare. The premature death of Mouton prevented him from
developing his work further.
The English architect and mathematician Sir Christopher Wren proposed in 1667
using the length of the seconds pendulum as a fixed standard, an idea that was sup-
ported by the French astronomer Abbé Jean Picard in 1671, the Dutch astronomer
Christiaan Huygens in 1673 and the French geodesist Charles Marie de la Con-
damine in 1746. Neither the milliare nor the length of the seconds pendulum was
chosen to be the standard of length however, with that honour going to a measure-
ment of length based on a particular fraction of the circumference of the Earth.
The metre, as the new unit of length was named, was originally defined as
one ten millionth part of the distance from the North Pole to the Equator along
a line that ran from Dunkirk through Paris to Barcelona. The survey of this line
was carried out under the direction of P. F. E. Méchain and J. B. J. Delambre. Both
were astronomers by profession, who took from 1792 to 1799 to complete the
task. A comparison with modern satellite measurements produces a difference of
0.02%, with the original determined metre being 0.2 mm too short (Alder, 2004). A
platinum rod was made equal in length to the metre determined from the survey and
deposited in the Archives de la République in Paris. It was accompanied by a one-
kilogram mass of platinum, as the standard unit of mass, in the first step towards
the establishment of the present set of SI units. Following the establishment of the
Conférence Général des Poids et Mesures (CGPM) in 1875, construction of new
platinum–iridium alloy standards for the metre and kilogram were begun.
The definition of the metre based on the 1889 international prototype was
replaced in 1960 by one based upon the wavelength of krypton 86 radiation, which
in turn was replaced in 1983 by the current definition based on the length of the path
travelled by light, in vacuo, during a time interval of 1/(299 792 458) of a second.
The unit of time, the second, initially defined as 1/(86 400) of a mean solar
day was refined in 1956 to be 1/(31 556 925.974 7) of the tropical year for 1900
January 0 at 12 h ET (Ephemeris Time). This astronomical definition was super-
seded by 1968, when the SI second was specified in terms of the duration of
9 192 631 770 periods of the radiation corresponding to the transition between two
hyperfine levels of the ground state of the caesium 133 atom at a thermodynamic
temperature of 0 K.
The unit of mass (kilogram) is the only SI unit still defined in terms of a manu-
factured article, in this case the international prototype of the kilogram which, with
the metre, were sanctioned by the first Conférence Général des Poids et Mesures
(CGPM) in 1889. These joined the astronomically determined second to form the
basis of the mks (metre–kilogram–second) system, which was similar to the cgs
(centimetre–gram–second) system proposed in 1874 by the British Association for
the Advancement of Science.
10 Introduction

A move to incorporate the measurement of other physical phenomena into the

metric system was begun by Gauss with absolute measurements of the Earth’s
magnetic field using the millimetre, gram and second. Later, collaborating with
Weber, Gauss extended these measures to include the study of electricity. Their
work was further extended in the 1860s by Maxwell and Thomson and others
working through the British Association for the Advancement of Science (BAAS).
Ideas that were incorporated at this time include the use of unit-name prefixes from
micro to mega to signify decimal submultiples or multiples.
In the fields of electricity and magnetism, the base units in the cgs system proved
too small and, in the 1880s, the BAAS and the International Electrotechnical Com-
mission produced a set of practical units, which include the ohm (resistance), the
ampere (electric current) and the volt (electromotive force). The cgs system for
electricity and magnetism eventually evolved into three subsystems: esu (electro-
static), emu (electromagnetic) and practical. This separation introduced unwanted
complications, with the need to convert from one subset of units to another.
This difficulty was overcome in 1901 when Giorgi combined the mechanical
units of the mks system with the practical electrical and magnetic units. Discus-
sions at the 6th CGPM in 1921 and the 7th CGPM in 1927 with other interested
international organizations led, in 1939, to the proposal of a four-unit system based
on the metre, kilogram, second and ampere, which was approved by the CIPM in
1946.
Eight years later at the 10th CGPM, the ampere (electric current), kelvin (ther-
modynamic temperature) and candela (luminous intensity) were introduced as base
units. The full set of six units was named the Système International d’Unités by
the 11th CGPM in 1960. The final base unit in the current set, the mole (amount of
substance), was added at the 14th CGPM in 1971.
The task of ensuring the worldwide unification of physical measurements was
given to the Bureau International des Poids et Mesures (BIPM) when it was estab-
lished by the 1875 meeting of the Convention du Mètre. Seventeen states signed
the original establishment document. The functions of the BIPM are as follows:

1. Establish fundamental standards and scales for the measurement of the principal
physical quantities and maintain the international prototypes.
2. Carry out comparisons of national and international prototypes.
3. Ensure the coordination of corresponding measuring techniques.
4. Carry out and coordinate measurements of the fundamental physical constants
relevant to these activities.

This brief enables the BIPM to make recommendations to the appropriate

committees concerning any revisions of unit definitions that may be necessary.
 1.7 Summary and recommendations 11

1.6 The future of SI units

It is evident from reading the above section on the development of units of measure-
ment that it is very much an ongoing project. Today’s definitions and measurements
are generally the best that are available now, but that is no guarantee that they will
still be so tomorrow.
The most recent candidates for change are the kilogram, the ampere, the kelvin,
and the mole. Under discussion are changing the definition of these units in the
following ways:
1. The kilogram is a unit of mass such that the Planck constant is exactly
6.626 069 3 × 10−34 J . s (joule seconds).
2. The ampere is a unit of electric current such that the elementary charge is exactly
1.602 176 53 × 10−19 coulombs (where 1 coulomb = 1 A . s (ampere second)).
3. The kelvin is a unit of thermodynamic temperature such that the Boltzmann
constant is exactly 1.380 650 5 × 10−23 J . K−1 (joules per kelvin).
4. The mole is an amount of substance such that the Avogadro constant is exactly
6.022 141 5 × 1023 mol−1 .
No changes are under immediate consideration for the standard units of time
(second), length (metre) or luminous intensity (candela).

1.7 Summary and recommendations

1.7.1 Summary
Astronomers tend to use a variety of units when describing the same quantity,
which has the potential to lead to confusion and mistakes. To a certain extent this
reflects the different training they may have received and the subsequent areas of
astronomy in which they have carried out research. The advantages of standard
systems in everyday use, as well as in astronomy, has unfortunately not meant that
such systems have been readily adopted by those they were designed to benefit.
The international system is currently the most widely accepted and used set of
physical units. It is still undergoing changes to the base unit definitions, mainly
to tie them to fundamental physical constants rather than Earth-based constants or
prototype model representations.

1.7.2 Recommendations
For astronomers who do not routinely use SI units, a simple approach to doing
so would be to convert whatever units are being used into their SI equivalents
and prepare research papers, presentations and lecture notes with the final results
given in both forms. This would lead to SI units becoming familiar, acceptable and,
hopefully, the universal system of choice.
 2
An introduction to SI units

The name Système International d’Unités (International System of Units), with the
abbreviation SI, was adopted by the 11th Conférence Générale des Poids et Mesures
(CGPM) in 1960.
This system includes two classes of units:
- base units
- derived units,
which together form the coherent system of SI units.

2.1 The set of SI base units

There are seven well defined base units in the SI. They are: the second, the metre,
the kilogram, the candela, the kelvin, the ampere and the mole, all selected by the
CGPM and regarded, by convention, to be dimensionally independent. Table 2.1
lists the base quantities and the names and symbols of the base units. The order of
the base units given in the table follows that of the chapters in this book.

2.2 The set of SI derived units

Derived SI units are those that may be expressed directly by multiplying or dividing
base units, e.g., density (kg . m−3 ) or acceleration (m . s−2 ) or electric charge (A . s).
Table 2.2 lists examples of SI derived units obtained from base units.
Special names have been assigned to selected derived units that are used to
prevent unwieldy combinations of base SI names occurring. Table 2.3 gives some
examples of such special names, with the derived unit expressed in terms of both
other SI units and of base SI units only. Note that the radian and steradian were
originally termed supplementary SI derived units.
Table 2.4 lists some examples of SI derived units whose names and symbols
include SI derived units with special names and symbols.

12
 2.3 Non-SI units currently accepted for use with SI units 13

Table 2.1. SI base units (BIPM, 2006)

Base quantity Name Symbol

time second s
length metre m
mass kilogram kg
luminous intensity candela cd
thermodynamic temperature kelvin K
electric current ampere A
amount of substance mole mol

Table 2.2. SI derived units (BIPM, 2006)

Derived quantity Name Symbol

area square metre m2

volume cubic metre m3
speed, velocity metre per second m . s−1
acceleration metre per second squared m . s−2
wavenumber reciprocal metre m−1
density, mass density kilogram per cubic metre kg . m−3
specific volume cubic metre per kilogram m3 . kg−1
current density ampere per square metre A . m−2
magnetic field strength ampere per metre A . m−1
concentration (of amount mole per cubic metre mol . m−3
of substance)
luminance candela per square metre cd . m−2

2.3 Non-SI units currently accepted for use with SI units

Table 2.5 lists examples of non-SI units that have been accepted for use with the
International System.
Table 2.6 gives three non-SI units that have been determined experimentally and
are also accepted for use with the International System. The first two values in
the table were obtained from Mohr et al. (2007) and that for the astronomical unit
from the USNO online ephemeris.6 Table 2.7 lists some other non-SI units that are
currently accepted for use with the International System.

6 See page K6, http://asa.usno.navy.mil/index.html

14 An introduction to SI units

Table 2.3. SI derived units with special names (BIPM, 2006)

Derived quantity Name Symbol Other SI Base SI

plane angle radian rad m . m−1

solid angle steradian sr m2 . m−2
frequency hertz Hz s−1
force newton N m . kg . s−2
pressure, stress pascal Pa N . m−2 m−1 . kg . s−2
energy, work, joule J N.m m2 . kg . s−2
quantity of heat
power, radiant flux watt W J.s m2 . kg . s−3
electric charge, coulomb C A. s
quantity of electricity
potential difference, volt V W . A−1 m2 . kg . s−3 . A−1
electromotive force
capacitance farad F C . V−1 m−2 . kg−1 . s4 . A2
electrical resistance ohm  V . A−1 m2 . kg . s−3 . A−2
magnetic flux weber Wb V.s m2 . kg . s−2 . A−1
magnetic flux density tesla T Wb . m−2 kg . s−2 . A−1
inductance henry H Wb . A−1 m2 . kg . s−2 . A−2
luminous flux lumen lm cd . sr cd
illuminance lux lx lm . m−2 cd . m−2
activity bequerel Bq s−1

2.4 Other non-SI units

In astronomy, many derived cgs units are in common use. The relationships between
these and their values in SI units are set out in Table 2.8.
The cgs system is based on the centimetre, the gram and the second, with other
units expressed in terms of these quantities. Unfortunately, electrical and magnetic
units can be expressed in three different ways leading to three different systems:
the cgs electrostatic system, the cgs electromagnetic system and the cgs Gaussian
system.
A final set of examples, given in Table 2.9, is of units that were more common
in older texts. If still used it is essential that the unit be redefined in SI terms.

2.5 Prefixes to SI units

In science in general and astronomy and astrophysics in particular, a huge range
of numbers is covered from, for example, the Planck time (∼5.4 × 10−44 s), the
time before which it is currently not possible to describe phenomena that might be
 2.5 Prefixes to SI units 15

Table 2.4. Some SI derived units whose names and symbols include SI derived
units with special names and symbols (BIPM, 2006)

Derived quantity Name Symbol Base SI

dynamic viscosity pascal second Pa . s m−1 . kg . s−1

moment of force newton metre N.m m2 . kg . s−2
surface tension newton per metre N . m−1 kg . s−2
angular velocity radian per second rad . s−1 s−1
angular acceleration radian per second squared rad . s−2 s−2
heat flux density, watt per square metre W . m−2 kg . s−3
irradiance
specific energy joule per kilogram J . kg−1 m2 . s−2
thermal conductivity watt per metre kelvin W . (m . K)−1 m . kg . s−3 . K−1
energy density joule per cubic metre J . m−3 m−1 . kg . s−2
electric field strength volt per metre V . m−1 m . kg . s−3 . A−1
electric charge coulomb per cubic metre C . m−3 m−3 . s . A
density
electric flux density coulomb per square metre C . m2 m−2 . s . A
permittivity farad per metre F . m−1 m−3 . kg−1 . s4 . A2
permeability henry per metre H . m−1 m . kg . s−2 . A−2
molar energy joule per mole J . mol−1 m2 . kg . s−2 . mol−1
radiant intensity watt per steradian W . sr−1 m2 . kg . s−3
radiance watt per square metre W . m−2 . sr−1 kg . s−3
steradian

Table 2.5. Non-SI units accepted for use with the

International System (BIPM, 2006)

Name Symbol Value in SI units

minute min 60 s
hour h 3600 s
day d 86 400 s
degree ◦ (π/180) rad
minute  (π/10 800) rad
second  (π/648 000) rad
litre l, L 10−3 m3
tonne t 103 kg
neper Np 1 Np = 1
bel B 1 B = (1/2) ln 10 (Np)
16 An introduction to SI units

Table 2.6. Non-SI units accepted for use with the International
System whose values in SI units are obtained experimentally
(BIPM, 2006)

Name Symbol Value in SI units

electronvolt eV 1.602 176 487 × 10−19 J

unified atomic mass unit u 1.660 538 782 × 10−27 kg
astronomical unit au 1.495 978 714 64 × 1011 m

Table 2.7. Some other non-SI units currently accepted

for use with the International System (BIPM, 2006)

Name Symbol Value in SI units

are a 102 m2
hectare ha 104 m2
bar bar 105 Pa
angstrom Å 10−10 m
barn b 10−28 m2

Table 2.8. Derived cgs units with special names

(BIPM, 2006). The mathematical symbol ∧ is used for
‘corresponds to’

Name Symbol Value in SI units

erg erg 10−7 J

dyne dyn 10−5 N
poise P 10−1 Pa . s
stokes St 10−4 m2 . s−1
gauss G ∧ 10−4 T
oersted Oe ∧ (1000/4π) A . m−1
maxwell Mx ∧ 10−8 Wb
stilb sb 10−4 cd . m−2
phot ph 104 lx
gal Gal 10−2 m . s−2
 2.5 Prefixes to SI units 17

Table 2.9. Examples of other non-SI units (BIPM, 2006)

Name Symbol Value in SI units

curie Ci 3.7 × 1010 Bq

jansky Jy 10−26 W . m−2 . Hz−1
fermi 10−15 m
standard atmosphere atm 101 325 Pa
micron μ 10−6 m

in progress due to a lack of suitable theories), up to ∼1041 kg for the mass of the
Milky Way galaxy. In astronomy, the range of official SI prefixes is insufficient to
cover the range of numbers, which has lead to the imaginative invention of many
others (not yet approved by the CGPM).

2.5.1 CGPM-approved prefixes for SI units

In order to cope with very large or very small numbers in various branches of
science, engineering and everyday life, the CGPM approved the use of a set of
prefixes that range from 10−24 up to 1024 . They are listed with their multiplication
factors, their names and their symbols in Table 2.10.
It is evident that even this range of prefixes is insufficient to meet the needs of
astronomy.

2.5.2 Unofficial prefixes for SI units

An extension to the CGPM-approved prefixes was suggested by Mayes (1994) to
cover even larger numbers from 1027 to 1048 and smaller numbers from 10−33
to 10−27 . Their names, symbols and multiplying factors are given in Table 2.11.
Examples of possible uses of the Mayes’ prefixes are given by Atkin (2007).
An obvious difficulty with a system of 31 prefixes derived from several different
languages is remembering which is which, though, due mainly to usage in computer
science, the names of the larger number prefixes are becoming more familiar with
increasing computer storage capacity. The following two systems offer alternative
solutions.

2.5.3 Powers of 1000

Languages used (Mayes, 1994) in deriving the official SI prefixes are: Latin, Greek,
Danish and Italian, with the unofficial prefixes adding: Portuguese, French, Span-
ish, Russian, Malay – Indonesian, Chinese, Sanskrit, Arabic, Hindi and Maori.
18 An introduction to SI units

Table 2.10. SI approved prefixes

Factor Name Symbol Factor Name Symbol

10−24 yocto y 101 deca da

10−21 zepto z 102 hecto h
10−18 atto a 103 kilo k
10−15 femto f 106 mega M
10−12 pico p 109 giga G
10−9 nano n 1012 tera T
10−6 micro μ 1015 peta P
10−3 milli m 1018 exa E
10−2 centi c 1021 zetta Z
10−1 deci d 1024 yotta Y

Table 2.11. Mayes unofficial prefixes

Factor Name Symbol Factor Name Symbol

10−33 weto w 1027 nava N or nv

10−30 vindo v 1030 sansa S or sa
10−27 tiso t 1033 besa B or be
1036 vela V or ve
1039 astra A or at
1042 cata C or ca
1045 quinsa Q or qu
1048 ultra U or ut

Whilst being a worthy attempt to produce at least one prefix derived from a lan-
guage originating from each of the six continents there is no obvious connection
between the prefix and the multiplying factor it is meant to represent. A possible
way around this problem is simply to use the straightforward and commonly used
names (e.g., million, billion, trillion, quadrillion etc.) assigned to the numbers. For
numbers smaller than 1 adding ‘th’ after the name (e.g., millionth, billionth, tril-
lionth, quadrillionth etc.) suffices. Table 2.12 sets out a list of multipliers for SI
units that run from 10−48 (one quindecillionth) to 1048 (one quindecillion).
Note that there is a simple relationship between the power, n, to which 1000 is
raised and the Latin prefix of the ending ‘-illion’. In Table 2.12, each line, starting
at the top-left-hand side of the table, is 1000 times smaller than the values in the
next line in the table. Unfortunately, the prefixes tend to become cumbersome and
 2.5 Prefixes to SI units 19

Table 2.12. Powers of 1000

Factor (1000n ) Name n Factor (1000n ) Name n

10−48 quindecillionth −16 103 thousand 1

10−45 quattuordecillionth −15 106 million 2
10−42 tredecillionth −14 109 billion 3
10−39 duodecillionth −13 1012 trillion 4
10−36 undecillionth −12 1015 quadrillion 5
10−33 decillionth −11 1018 quintillion 6
10−30 nonillionth −10 1021 sextillion 7
10−27 octillionth −9 1024 septillion 8
10−24 septillionth −8 1027 octillion 9
10−21 sextillionth −7 1030 nonillion 10
10−18 quintillionth −6 1033 decillion 11
10−15 quadrillionth −5 1036 undecillion 12
10−12 trillionth −4 1039 duodecillion 13
10−9 billionth −3 1042 tredecillion 14
10−6 millionth −2 1045 quattuordecillion 15
10−3 thousandth −1 1048 quindecillion 16
100 one 0

the displacement of the value of n by one relative to the Latin prefix offers the
opportunity for mistakes to be made.

2.5.4 Some astronomical examples

The values of astronomical quantities used in this section were taken from Cox
(2000).

Planck time  5.4 × 10−44 s (second) (SI base unit)

or 5.4 × 10 −20 ys (yoctosecond) (official SI prefix)
or 5.4 × 10−11 ws (wetosecond) (Mayes’ unofficial prefix)
or 5.4 quattuordecillionths s (powers of 1000)
wavelength of γ - radiation  1 × 10−14 m (metre) (SI base unit)
or 10 fm (femtometre) (official SI prefix)
or 10 quadrillionths m (powers of 1000)
1 astronomical unit  1.496 × 1011 m (metre) (SI base unit)
or 149.6 Gm (gigametre) (official SI prefix)
or 149.6 billion m (powers of 1000)
20 An introduction to SI units

1 parsec  3.086 × 1016 m (metre) (SI base unit)

or 30.86 Pm (petametre) (official SI prefix)
or 30.86 quadrillion m (powers of 1000)
1 solar mass  1.989 × 1030 kg (kilogram) (SI base unit)
or 1.989 × 109 Yg (yottagram) (official SI prefix)7
or 1.989 Bg (besagram) (Mayes’ unofficial prefix)
or 1.989 nonillion kg (powers of 1000)
mass of Milky Way Galaxy  1.89 × 1041 kg (kilogram) (SI base unit)
or 1.89 × 1020 Yg (yottagram) (official SI prefix)
or 189 Cg (catagram) (Mayes’ unofficial prefix)
or 189 duodecillion kg (powers of 1000)
luminous intensity of an MV = 0 star outside the Earth’s atmosphere  2.45 ×
1029 cd (candela) (SI base unit)
or 2.45 × 105 Ycd (yottacandela) (official SI prefix)
or 245 Ncd (navacandela) (Mayes’ unofficial prefix)
or 245 octillion cd (powers of 1000)

2.5.5 Other methods of denoting very large or very small numbers

It is evident that none of the proposed modifiers to the name of the unit is satisfactory
for astronomers. A simpler way of writing and speaking powers of ten would appear
to be the answer, e.g., instead of writing 2 × 1030 kg the expression 2 d 30 kg could
be used, where ‘d’, standing for deca is both the Greek word δκα for ten (Liddell
& Scott, 1996), and the SI prefix for 10, which would assume the meaning ‘10 to
the power’. So, in the example given, instead of saying ‘two times ten to the power
thirty’, the shorter ‘two d thirty’ would be used. Examples of the d notation are
given in many of the tables throughout this book.
Allen (1951) proposed the use of ‘dex’ for the logarithm to base 10 of a number
so that 2 × 1030 would be written as 30.301 dex.
Urry (1988), tackling problems associated with objects of galactic mass, used
the compact shorthand M8 instead of 108 M with Mn representing a mass of 10n
solar masses.

2.6 IAU recommendations regarding SI units

The International Astronomical Union in its style manual (Wilkins, 1989)8 states
that: ‘The international system (SI) of units, prefixes and symbols should be used for
all physical quantities except that certain special units, may be used in astronomy,

7 remember that 1000 g = 1 kg.

8 See also www.iau.org/science/publications/proceedings_rules/units/
 2.6 IAU recommendations regarding SI units 21

Table 2.13. Non-SI units recognized for use in astronomy

Name Symbol Value in SI units

Julian year a 3.155 76 × 107 s

cycle c 2 π rad
astronomical unit au 149.598 × 109 m
parsec pc 30.857 × 1015 m
solar mass M 1.9891 × 1030 kg
atomic mass unit u 1.660 539 × 10−27 kg

Table 2.14. Obsolete units that should not be used

Name Symbol Value in SI units

angstrom Å 10−10 m
micron μ 10−6 m
fermi 10−15 m
barn b 10−28 m2
cubic centimetre cc 10−6 m3
dyne dyn 10−5 N
erg erg 10−7 J
calorie cal 4.1868 J
bar bar 105 Pa
standard atmosphere atm 101.325 kPa
gal Gal 10−2 m . s−2
eotvos E 10−9 s−2
gauss G 10−4 T
gamma 10−9 T
oersted Oe (1000/4π) A . m−1

without risk of confusion or ambiguity, in order to provide a better representation

of the phenomena concerned.’
In addition to the non-SI units listed in Table 2.5 and Table 2.6, the non-SI units
given in Table 2.13 are recognized for use in astronomy:
The IAU recommendations for prefixes are in line with those of BIPM and are
as given in Table 2.10.
Non-SI units such as British Imperial, American or other national systems of
units should not be used. The cgs and obsolete units given in Table 2.14 should
not be used, though some are still currently accepted for use with the International
System (BIPM, 2006).
22 An introduction to SI units

2.6.1 Angle
Currently a sexagesimal-based system of units is used to specify the positions of
celestial bodies in astronomy. Declination values are typically given as degrees,
minutes and seconds of arc north or south of the celestial equator, and right ascen-
sion as hours, minutes and seconds of time increasing eastwards from zero at the
intersection point of the celestial equator and the ecliptic, marking the position
of the Sun at the Vernal Equinox (where the declination of the Sun moves from
negative values south of the equator to positive north of the equator). This use of a
unit of time to represent an angle is a possible source of confusion and should be
avoided. Detailed relationships between the SI unit of angle (the radian) and the
various sexagesimal systems are given in Chapter 4. If, for some reason, the radian
is not considered a suitable unit in a particular circumstance, then the degree with
decimal subdivision should be used.
The use of the ‘mas’ meaning milliarcsecond for angular resolution or angular
separation of astronomical objects should be replaced by a more appropriate SI
unit, such as the nrad (nanoradian).

2.6.2 Time
Other than the base SI unit of time, the second, astronomers use longer lengths of
time, such as the minute (60 seconds), the hour (3600 seconds), the day (86 400
seconds) and the Julian year consisting of 365.25 d or 31.557 6 × 106 s. There are,
however, several different kinds of day and year that relate to particular problems
in astronomy (see Chapter 5).
The variability of the Earth’s rotation rate means that time based on that rate
varies with respect to the SI second. Hence sidereal, solar and Universal Time
should be considered as measurements of hour angle expressed in time measure
and not suitable for precise measures of time intervals.

2.6.3 Distance and mass

The IAU accepts the use of a special set of length, mass and time units for the
study of motions in the Solar System – they are related to one another through the
adopted value of the Gaussian gravitational constant k ( = 0.017 202 098 95) when
it is expressed in these units.
The astronomical unit (the unit of distance) is the radius of a circular orbit in
which a body of negligible mass, and free from perturbations, would revolve around
the Sun in 2 π /k days. This distance is slightly less than the semimajor axis of the
Earth’s orbit (∼1.495 978 × 1011 m).
 2.7 Summary and recommendations 23

The parsec is a distance (∼3.086 × 1016 m) equal to that at which the astronom-
ical unit subtends an angle of 1 arcsec (π /(180 × 3600) rad).
The light year is the distance (∼9.461 × 1015 m) travelled by light, in vacuo, in
one Julian year.
The astronomical unit of mass is the solar mass (∼1.989 1 × 1030 kg) denoted
by M .

2.6.4 Wavenumber
The reciprocal wavelength or wavenumber is used mainly by infrared astronomers
and is normally based on the cm−1 . If used it should be in the SI unit form of m−1 ,
but in either case the unit must be given as it is not dimensionless.

2.6.5 Magnitude
Magnitude may be defined as the ratio of the logarithm of the signal strength of
the celestial object of interest to that of a standard star or object. As such it is a
dimensionless quantity. Some magnitude scales have been calibrated in terms of SI
units (see Chapter 8).

2.7 Summary and recommendations

2.7.1 Summary
The Système International d’Unités consists of two classes of units: the base units
of time, length, mass, luminous intensity, thermodynamic temperature, electric
current and amount of substance; and a set of derived units obtained by dividing or
multiplying base units. The SI units are decimal, with larger and smaller multiples
of the base units assigned prefixes. The current set of official prefixes is insufficient
to meet the needs of astronomers.
The International Astronomical Union has produced a set of recommendations
concerning the use of SI units in astronomy and the continued use of specialized
non-SI astronomical units.

2.7.2 Recommendations
Astronomers should follow the recommendations proposed by the IAU with par-
ticular reference to dropping the use of the cgs-based units so prevalent at present.
Given that the current set of official SI prefixes is inadequate, consistency may be
maintained by quantities being presented as the product of the measurement times
the basic unit (i.e., the unit without any prefix). For example, the astronomical unit
should be given as 1.496 × 1011 m or 1.496 d 11 m and not 149.6 Gm. Examples of
all three forms are given in different tables throughout the book.
 3
Dimensional analysis

3.1 Definition of dimensional analysis

Dimensional analysis is a technique for studying the dimensions of physical
quantities. It may be used to:
1. Reduce the physical properties of derived SI units into those of the more
fundamental SI base units.
2. Assist in converting quantities expressed in non-SI units to SI units.
3. Verify the correctness of an equation in terms of dimensional and unitary
consistency.
4. Determine the dimension and unit of a variable in an equation.
5. Provide a means for selecting relevant data and how best to present it.

3.1.1 The dimensions of the SI base units

The seven base SI units provide seven independent dimensions with which to
describe any derived SI unit. The symbols for each of the base quantity dimensions
are given in Table 3.1.

3.1.2 Dimensions of some of the SI derived units

The dimension of a derived unit, X, is the product of the base unit dimensions, Bi ,
where Bi = T , L, M, J ,
, I , and N, such that:
7

dim(X) = dimαi (Bi ) (3.1)
i=1

By way of example, the derived quantity force, P , has the derived SI unit
kg . m . s−2

dim(P ) = [T ]−2 . [L]1 . [M]1 . [J ]0 . [

]0 . [I ]0 . [N]0 (3.2)

24
 3.2 Dimensional equations 25

Table 3.1. Symbols for SI base quantity dimensions

(BIPM, 2006)

Base quantity Dimension symbol

time [T ]
length [L]
mass [M]
luminous intensity [J ]
thermodynamic temperature [
]
electric current [I ]
amount of substance [N ]

which simplifies to:

dim(P ) = [L] . [M] . [T ]−2 (3.3)

Table 3.2 gives examples of the dimensions of some derived SI units. Note that
both the radian and steradian are examples of dimensionless units, as in both cases
the value of the dimension is 1. Some derived quantities have the same dimension,
e.g., the hertz and the bequerel both have the dimension [T ]−1 .

3.2 Dimensional equations

For an equation describing a physical situation to be true it is necessary that both
sides of the equation have the same dimensions, that is, the equation must be
dimensionally homogeneous, e.g.:

v = d/t (3.4)

where v = velocity (dimension [L] . [T ]−1 ), d = distance (dimension [L]) and t =

time (dimension [T ]) may be written in dimensional terms as:

dim(v) = [L] . [T ]−1 = [L]/[T ] = dim(d/t) (3.5)

so the equation is homogeneous. Note that the requirement is for the dimensions
to be consistent, so that any set of consistent units within a particular dimen-
sion may be used and converted to any other by means of a constant factor. If
physical quantities have the same dimensions, they may only be combined by
addition or subtraction. For example, [L] + [L] or [L] − [L] but not [L] × [L]
(which is a measure of area), nor [L]/[L] (which is a ratio and a dimensionless
number).
26 Dimensional analysis

Table 3.2. Dimensions of some derived SI units

Derived quantity Name Dimension

plane angle radian [L] . [L]−1

solid angle steradian [L]2 . [L]−2
frequency hertz [T ]−1
force newton [L] . [M] . [T ]−2
pressure, stress pascal [L]−1 . [M] . [T ]−2
energy, work, joule [L]2 . [M] . [T ]−2
quantity of heat
power, radiant flux watt [L]2 . [M] . [T ]−3
electric charge, quantity of coulomb [T ] . [I ]
electricity
potential difference, volt [L]2 . [M] . [T ]−3 . [I ]−1
electromotive force
capacitance farad [L]−2 . [M]−1 . [T ]4 . [I ]2
electrical resistance ohm [L]2 . [M] . [T ]−3 . [I ]−2
magnetic flux weber [L]2 . [M] . [T ]−2 . [I ]−1
magnetic flux density tesla [M] . [T ]−2 . [I ]−1
inductance henry [L]2 . [M] . [T ]−2 . [I ]−2
luminous flux lumen [J ]
illuminance lux [L]−2 . [J ]
activity bequerel [T ]−1
area square metre [L]2
volume cubic metre [L]3
speed, velocity metre per second [L] . [T ]−1
acceleration metre per second squared [L] . [T ]−2
wavenumber reciprocal metre [L]−1
density, mass density kilogram per cubic metre [M] . [L]−3
specific volume cubic metre per kilogram [L]3 . [M]−1
current density ampere per square metre [I ] . [L]−2
magnetic field strength ampere per metre [I ] . [L]−1
concentration (of amount mole per cubic metre [N] . [L]−3
of substance)
luminance candela per square metre [J ] . [L]−2
rate of cooling kelvin per second [
] . [T ]−1

A dimensional equation that is to be used for converting one set of units within a
dimension to another (e.g., inches to metres) must also include a conversion factor,
k (to convert inches to metres, k = 0.0254). So Equation (3.1) above would be
rewritten in the more general form:
7

dim(X) = [ki dim(Bi )]αi (3.6)
i=1
 3.2 Dimensional equations 27

where the ki relate to the conversion factors required to convert the units in which
the measurements were carried out to base units (e.g., the multiplicative conversion
factor from inches to metres (Kaye & Laby, 1959) is 1/39.370 147 = 0.025 399 96).

Worked example: convert quantities expressed in non-SI units to SI units

The orbital speed of the Earth about the Sun was commonly given as 66 600 mph
(miles per hour) in older English-language popular astronomy books (e.g., Evans,
1954). What does this speed equate to in the SI base units metres and seconds?
Using Equation (3.4) above, the following dimensional equation may be written
where the left-hand side is the dimension of speed:
dim(v) = [kL . L]1 . [kT . T ]−1 (3.7)
kL is the length conversion factor from miles to metres and kT is the time conversion
factor from hours to seconds.
1 mile = 1609.344 m and 1 hour = 3600 s, so
dim(v) = [1609.344 L]1 .[3600 T ]−1 = 0.447 04 [L]1 .[T ]−1 (3.8)
The combined conversion factor, 0.447 04, is then multiplied by the Earth’s orbital
speed in mph (66 600) to give the speed in m . s−1 .
Hence, 66 600 mph ≡ 2.977 × 104 m . s−1 ≡ 29.77 km . s−1 .

Example: compare the consistency between two units of monochromatic

flux density
The two units to be compared are the jansky (10−26 W . m−2 . Hz−1 ) and the
(erg . cm−2 . s−1 . Å−1 ) used in a Catalogue of Stellar Ultraviolet Fluxes (Thompson
et al., 1978).
The component parts of the jansky are the watt, W, which is a derived unit of
power equal to 1 J . s−1 (J = joule); the joule is a derived unit of energy or work
and is equal to 1 m2 . kg . s−2 ; the hertz is a unit of frequency and is measured in
inverse seconds, s−1 . Combining these components, the jansky in SI base units is:
(m2 . kg . s−2 ) . s−1 . m−2 . (s−1 )−1 , which reduces to kg . s−2 , or in dimensional
terms:
dim(fν (J y) ) = [M]1 . [T ]−2 (3.9)
The ultraviolet unit of Thompson et al. (1978), hereafter called the ‘TD1’ unit,
is a cgs unit with the following named components. The erg is the cgs unit of
energy equal to 1 cm2 . g . s−2 or 10−7 J and the angstrom is a unit of length equal
to 10−10 m. Combining all the components, the TD1 unit in cgs base units is
(cm2 . g . s−2 ) . cm−2 . s−1 . cm−1 , which reduces to g . s−3 . cm−1 or, in dimensional
terms:
dim(fλ(TD1) ) = [M]1 . [T ]−3 . [L]−1 (3.10)
28 Dimensional analysis

Quite evidently, even the application of multiplicative scaling factors would not
make the two monochromatic flux units compatible in a dimensional sense. The
difference in derived dimensions in this case is due to the different variable used to
specify the bandwidth of the observation. For the jansky, the bandwidth is measured
in hertz ([T ]−1 ) and for the TD1 unit the bandwidth is measured in angstroms ([L]).
To compare the measures directly, the wavelength bandwidth in angstroms needs to
be converted to a frequency bandwidth in hertz. The algebraic relationship between
frequency (ν) and wavelength (λ) is given by:
c
λ= (3.11)
ν
where c is the velocity of light. The relationship between the bandwidths in wave-
length form ( λ) and frequency form ( ν) is obtained by differentiating Equation
(3.11) to give:
c
λ = − 2 ν (3.12)
ν
The dimensional form of (c/ν 2 ) is ([L] . [T ]−1 /[T ]−2 ) or [L] .[T ], so multiply
dim(fλ(TD1) ) by [L] . [T ] to obtain:

dim(fλ(TD1) ) . [L]1 . [T ]1 = [M]1 . [T ]−3 . [L]−1 . [L]1 . [T ]1 (3.13)

= [M]1 . [T ]−2 (3.14)
= dim(fν(Jy) ) (3.15)

The negative sign in Equation (3.12) is cancelled because the value for the fre-
quency decreases with increasing wavelength, so if λ is positive then ν must
be negative. Now all that has to be done is evaluate the multiplying factor to convert
the TD1 unit to janskys.

Example: how to convert TD1 cgs flux units to janskys (derived SI unit) for
the B5V star HD74071
The ESRO satellite TD1 measured ultraviolet flux of HD74071 is fλ = 24.73 ×
10−11 erg . s−1 . cm−2 . Å−1 at a wavelength of λ = 1565 Å with a bandwidth of
λ = 330 Å. The frequency ν, corresponding to the wavelength 1565Å derived by
using equation (3.11) is 1.916 × 1015 Hz, with a frequency bandwidth derived from
Equation (3.12) of 4.043 × 1014 Hz.
The TD1 flux unit may be converted to SI base units, with appropriate factors, as:
−1
TD1 flux unit = erg . s−1 . cm−2 . Å
≡ (10−7 ) W . (10−2 )−2 . m−2 . (10−10 )−1 . m−1
= 107 . W . m−2 . m−1 (3.16)
 3.3 Summary and recommendations 29

Note that the powers to which the conversion factors are raised match those of
the SI units. These individual conversion factors are multiplied together to form a
single conversion factor for the entire derived SI unit. Now convert the bandpass
from wavelength, λ, to frequency, ν, to give the TD1 units in janskys:
λ
fν(TD1) = fλ(TD1) . . 1026 (3.17)
ν
2.998 × 108
= 107 . (24.73 × 10−11 ) . . 1026
(1.916 × 1015 )2
= 20.2 Jy
So the measured monochromatic flux density from the star HD74071 may be
given as either 24.73 × 10−11 erg . cm−2 s−1 . Å−1 or 20.20 Jy.

Example: how to transform the catalogue measurement error in TD1

catalogue units to janskys
Thompson et al. (1978) give the measurement error in their ultraviolet fluxes in
wavelength space as δ fλ (TD1) . To transform this value to frequency space, δ fν (TD1) ,
Equation (3.17) is rewritten as:
λ
δ fν (TD1) = δ fλ (TD1) . . 1026 (3.18)
ν
Substitute values for the variables from above, including the combined conversion
factor of 107 , in Equation (3.16) to give:
δ fν(TD1) = ± 0.005 Jy

3.3 Summary and recommendations

3.3.1 Summary
Using the techniques of dimensional analysis to transform commonly used, but
generally non-SI derived units, to SI derived or base units can be of considerable
assistance in the reduction in the number of mistakes, which unfortunately are not
uncommon in such an exercise. Some examples are given of such transformations in
simple dynamics and ultraviolet stellar photometry. Many more worked examples
are given in the chapters relating to each of the SI base units.

3.3.2 Recommendations
No matter how much care is taken in the transformation of units, mistakes do occur.
Particular problems are associated with the powers to which the base units are raised
and the signs and values of powers of ten. So always check any unit transformation
that has been carried out and when satisfied that it is correct, check it again!
 4
Unit of angular measure (radian)

4.1 SI definition of the radian

The radian is the angle subtended at the centre of a circle by an arc along the
circumference whose length is equal to that of the radius of the circle.
The dimension of angular measure is [L] . [L]−1 = [1], its unit is the radian and
its symbol is rad.
−→
→ = 1, where OA = OB
AB
In Figure 4.1, the angle AOB is equal to one radian if −
OA
−→
is the radius of the circle and AB is the distance along the circumference of the
circle from A to B. Given that the circumference of a circle of radius r is 2πr, then
the number of radians in the circle is simply 2πr/r or 2π .

4.2 Commonly used non-SI units of angular measure

4.2.1 Converting from ( ◦   ) and (h m s) to radians
Positions of astronomical objects are regularly given as angles from a reference
point or plane in degrees (◦ ), minutes ( ) and seconds ( ), or sometimes as a measure
of time from a reference point in hours (h), minutes (m) and seconds (s). There are
60 in 1 and 60 s in 1 m, 60 in 1◦ and 60 m in 1 h and 360◦ or 24 h in a circle. This
sexagesimal system was first used by the Babylonians (Pannekoek, 1961) more
than 2000 years ago and most astronomical catalogues still use it, though some
catalogues do include positions in radians, e.g., the SuperCOSMOS Sky Survey.9
There are 2π rad in 360◦ , which gives the following conversion factors from
sexagesimal measures to radians:
 2π 
1◦ = = 0.017 453 292 520 rad
360

9 See www-wfau.roe.ac.uk/sss

30
 4.2 Commonly used non-SI units of angular measure 31

B
O

Figure 4.1. Definition of a radian.

 2π 
1 = = 0.000 290 888 209 rad
60 × 360
 2π 
1 = = 0.000 004 848 137 rad (4.1)
60 × 60 × 360
Similarly, there are 2π rad in 24 h or:
 2π 
1h= = 0.261 799 387 799 149 rad
24
 2π 
1m= = 0.004 363 323 129 985 rad
60 × 24
 2π 
1s= = 0.000 072 722 052 166 rad (4.2)
60 × 60 × 24

Example: calculate the position of the bright northern star Capella in radians
Capella has the following coordinates in the FK5 catalogue for equinox (J2000.0)
and epoch (J2000.0):

α = 05h 16m 41.36s δ = +45◦ 59 52.8

where α is the right ascension and δ is the declination of Capella (see Section 4.3.2
below). The right ascension of Capella in radians to six decimal places is:

5 × 0.261 799 + 16 × 0.004 363 + 52.8 × 0.000 073

32 Unit of angular measure (radian)

so
α = 1.381818 rad
and the declination of Capella in radians to six decimal places is:
45 × 0.017 453 + 59 × 0.000 290 + 52.8 × 0.000 005
so
δ = 0.802 816 rad
If, instead of measuring declination from the celestial equator (CE) northwards
from 0 rad to + π2 rad and southwards from 0 rad to − π2 rad, we measure it from
the south celestial pole (SCP) as 0 rad to the north celestial pole (NCP) as +π rad
(see Zacharias et al., 2000, in which declination is given as a south polar distance
in units of milliarcseconds (mas)), then the declination in radians becomes:
π
δ(SCP0 ) = δ(CE0 ) + (4.3)
2
Using the south celestial pole zero point (SCP0 ), the revised declination of Capella
becomes 2.373 612 rad. Expressed in milliradians (mrad), the coordinates of Capella
become:
α = 1381.817 893 mrad
δSCP0 = 2373.612 876 mrad
Given that, e.g., The Astronomical Almanac lists its bright star positions to the
nearest tenth of a second of time and nearest whole second of arc, what number of
decimal places would yield a similar precision using milliradians? The answers are,
approximately, three decimal places for right ascension and two for declination. So
for the worked example using Capella above, a catalogue position in milliradians
with similar precision to that given in The Astronomical Almanac would be:
α = 1381.818 mrad
δSCP0 = 2373.61 mrad
By way of another example, the equatorial coordinates listed in the table of 86
bright stars given in the Handbook of the British Astronomical Association for 2008
were converted to milliradians and the V magitudes to janskys and ln(janskys). The
following equation was used to convert from magnitudes to monochromatic flux
densities10 in janskys (10−26 Wm−2 Hz−1 ):
VJ = 3600(10−0.2V ) (4.4)

10 See Chapter 8 for more details.

 4.2 Commonly used non-SI units of angular measure 33

Table 4.1. Equatorial coordinates in (h,m,s), (o ,  ,  ), and milliradians, V

magnitudes and monochromatic flux densities in janskys for 86 bright stars

Star α δ V α δSCP fVν ln(fVν )

name h m s ◦   mag mrad mrad Jy Jy
α And 00 08 49.7 29 08 14 2.06 38.520 2079.33 1394 7.24
β Cas 00 09 38.2 59 11 48 2.27 42.047 2603.97 1265 7.14
α Cas 00 40 59.7 56 35 02 2.23 178.874 2558.37 1289 7.16
β Cet 00 44 00.9 −17 56 24 2.04 192.051 1257.68 1407 7.24
β And 01 10 12.6 35 39 55 2.06 306.348 2193.27 1394 7.24
α Eri 01 38 01.8 −57 11 37 0.46 427.736 572.57 2912 7.97
γ And 02 04 25.5 42 22 13 2.26 542.906 2310.29 1271 7.14
α Ari 02 07 39.3 23 30 08 2.00 557.000 1980.98 1433 7.26
α UMi 02 41 49.8 89 18 03 2.02 706.116 3129.38 1420 7.25
β Per 03 08 43.5 40 59 16 2.10 823.468 2286.16 1368 7.22
α Per 03 24 56.0 49 53 27 1.79 894.190 2441.55 1578 7.36
η Tau 03 47 59.5 24 07 51 2.87 994.801 1991.95 960 6.86
α Tau 04 36 24.6 16 31 33 0.85 1206.066 1859.22 2433 7.79
β Ori 05 14 56.8 −08 11 33 0.12 1374.214 1427.81 3406 8.13
α Aur 05 17 19.1 46 00 21 0.08 1384.562 2373.74 3469 8.15
γ Ori 05 25 35.2 06 21 24 1.64 1420.639 1681.74 1691 7.43
β Tau 05 26 49.8 28 36 50 1.65 1426.064 2070.20 1683 7.42
δ Ori 05 32 26.5 −00 17 36 2.23 1450.550 1565.67 1289 7.16
 Ori 05 36 38.7 −01 11 50 1.70 1468.890 1549.90 1645 7.40
ζ Ori 05 41 11.2 −01 56 19 1.77 1488.707 1536.96 1593 7.37
κ Ori 05 48 09.6 −09 40 02 2.06 1519.134 1402.07 1394 7.24
α Ori 05 55 37.9 07 24 29 0.40 1551.735 1700.09 2994 8.00
β Aur 06 00 09.1 44 56 51 1.90 1571.458 2355.27 1500 7.31
β CMa 06 23 04.4 −17 57 38 1.98 1671.472 1257.32 1446 7.27
α Car 06 24 08.5 −52 42 02 −0.72 1676.134 650.99 5015 8.52
γ Gem 06 38 12.2 16 23 29 1.93 1737.489 1856.88 1480 7.29
α CMa 06 45 31.3 −16 43 42 −1.46 1769.422 1278.83 7051 8.86
 CMa 06 58 57.6 −28 59 03 1.50 1828.057 1064.92 1804 7.49
δ CMa 07 08 44.2 −26 24 26 1.86 1870.716 1109.90 1528 7.33
α Gem 07 35 08.5 31 52 09 1.95 1985.930 2127.01 1466 7.29
α CMi 07 39 44.8 05 12 10 0.38 2006.023 1661.60 3022 8.00
β Gem 07 45 50.1 28 00 18 1.14 2032.588 2059.57 2129 7.66
ζ Pup 08 03 53.0 −40 01 39 2.25 2111.339 872.18 1277 7.15
γ Vel 08 09 47.7 −47 21 43 1.78 2137.133 744.17 1585 7.36
 Car 08 22 41.3 −59 32 14 1.86 2193.391 531.67 1528 7.33
δ Vel 08 44 56.3 −54 44 23 1.96 2290.475 615.40 1459 7.28
λ Vel 09 08 18.6 −43 28 02 2.21 2392.453 812.15 1301 7.17
β Car 09 13 17.4 −69 45 08 1.68 2414.183 353.39 1660 7.41
ι Car 09 17 19.0 −59 18 40 2.25 2431.752 535.62 1277 7.15
α Hya 09 28 00.3 −08 41 45 1.98 2478.389 1419.02 1446 7.27
α Leo 10 08 49.4 11 55 31 1.35 2656.492 1778.93 1933 7.56
γ Leo 10 20 26.4 19 47 54 1.99 2707.180 1916.34 1439 7.27
34 Unit of angular measure (radian)

Table 4.1. (cont.)

Star α δ V α δSCP fVν ln(fVν )

name h m s ◦   mag mrad mrad Jy Jy
β UMa 11 02 21.0 56 20 12 2.37 2890.047 2554.05 1208 7.10
α UMa 11 04 14.8 61 42 17 1.79 2898.322 2647.74 1578 7.36
β Leo 11 49 29.6 14 31 28 2.14 3095.748 1824.29 1343 7.20
α Cru 12 27 04.6 −63 08 46 1.33 3259.736 468.68 1951 7.57
γ Cru 12 31 38.5 −57 09 38 1.63 3279.655 573.15 1699 7.43
γ Cen 12 41 59.3 −49 00 22 2.17 3324.801 715.47 1325 7.18
β Cru 12 48 13.4 −59 44 06 1.25 3352.006 528.22 2024 7.61
 UMa 12 54 24.1 55 54 50 1.77 3378.964 2546.67 1593 7.37
ζ UMa 13 24 16.0 54 52 52 2.27 3509.275 2528.65 1265 7.14
α Vir 13 25 38.5 −11 12 20 0.98 3515.274 1375.22 2292 7.73
 Cen 13 40 25.9 −53 30 33 2.30 3579.808 636.88 1248 7.12
η UMa 13 47 52.5 49 16 16 1.86 3612.286 2430.73 1528 7.33
β Cen 14 04 25.8 −60 24 49 0.61 3684.520 516.37 2718 7.90
θ Cen 14 07 11.1 −36 24 41 2.06 3696.541 935.29 1394 7.24
α Boo 14 16 03.0 19 08 18 −0.04 3735.222 1904.82 3666 8.20
η Cen 14 36 03.0 −42 11 41 2.31 3822.489 834.35 1242 7.12
α Cen 14 40 10.9 −60 52 12 0.00 3840.517 508.41 3600 8.18
α Lup 14 42 29.9 −47 25 27 2.30 3850.625 743.08 1248 7.12
 Boo 14 45 21.4 27 02 19 2.40 3863.097 2042.70 1192 7.08
β UMi 14 50 41.3 74 07 15 2.08 3886.361 2864.44 1381 7.23
α CrB 15 35 02.9 26 41 11 2.23 4079.918 2036.56 1289 7.16
δ Sco 16 00 50.2 −22 38 43 2.32 4192.440 1175.56 1236 7.12
α Sco 16 29 55.8 −26 27 01 0.96 4319.384 1109.15 2313 7.74
α TrA 16 49 34.3 −69 02 32 1.92 4405.087 365.78 1486 7.30
 Sco 16 50 42.9 −34 18 29 2.29 4410.076 972.00 1254 7.13
λ Sco 17 34 11.2 −37 06 33 1.63 4599.757 923.11 1699 7.43
α Oph 17 35 19.8 −12 33 16 2.08 4604.745 1351.68 1381 7.23
θ Sco 17 37 55.8 −43 00 09 1.87 4616.090 820.26 1521 7.32
γ Dra 17 56 48.2 51 29 18 2.23 4698.440 2469.43 1289 7.16
 Sgr 18 24 44.2 −34 22 48 1.85 4820.323 970.75 1535 7.33
α Lyr 18 37 13.6 38 47 31 0.03 4874.820 2247.84 3550 8.17
σ Sgr 18 55 47.5 −26 17 08 2.02 4955.826 1112.02 1420 7.25
β Cyg 19 31 03.9 27 58 41 3.08 5109.735 2059.10 871 6.77
α Aql 19 51 11.9 08 53 29 0.77 5197.583 1725.98 2525 7.83
γ Cyg 20 22 32.0 40 17 03 2.20 5334.307 2273.88 1307 7.17
α Pav 20 26 18.9 −56 42 26 1.94 5350.808 581.06 1473 7.29
α Cyg 20 41 43.3 45 18 40 1.25 5418.032 2361.62 2024 7.61
α Cep 21 18 46.9 62 37 18 2.44 5579.737 2663.75 1170 7.06
 Peg 21 44 36.2 09 54 51 2.39 5692.405 1743.83 1197 7.08
α Gru 22 08 45.9 −46 55 10 1.74 5797.831 751.89 1615 7.38
β Gru 22 43 10.3 −46 50 24 2.11 5947.958 753.28 1362 7.21
α PsA 22 58 07.1 −29 34 37 1.16 6013.175 1054.58 2110 7.65
β Peg 23 04 11.3 28 07 45 2.42 6039.660 2061.74 1181 7.07
α Peg 23 05 11.1 15 15 04 2.49 6044.009 1836.97 1143 7.04
 4.2 Commonly used non-SI units of angular measure 35

where the conversion factor of 3600 is from an absolute calibration of a zero mag-
nitude star in the Johnson V band given by Bessell (1992), V is the listed value
of the apparent V magnitude and VJ is the apparent monochromatic flux density
of the star in janskys. A plot of the stars in Table 4.1 is shown in Figure 4.2 using
Mollweide’s projection.11
Obvious advantages of the SI radians angular measurement system are that only
two numbers are required to specify the location of the celestial body, rather than
six, and that all the declination values are positive when the declination of the South
Celestial Pole is set equal to 0 mrad.

NCP

3 Polaris

URSA MAJOR

2
Arcturus

Altair ORION
equator
Υ

Sirius
6 5 4 3 2 1
1

Canopus

CRUX

SCP

Figure 4.2. A star chart for the 86 bright stars in Table 4.1, plotted in radians,
with declination measured from 0 at the south celestial pole (SCP) northwards
to π rad at the north celestial pole (NCP) and right ascension increasing in an
easterly direction from 0 at the First Point of Aries, ϒ, to 2π rad. Right ascensions
in radians from 1 rad to 6 rad are printed to the left of the relevant circle and
declinations are printed from 1 rad to 3 rad immediately adjacent to the relevant
circle. The dashed declination lines are at the 0.5-rad points. The size of the star
images are proportional to ln(fVν ) in janskys.

11 See, e.g., http://www.astron.nl/aips++/docs/memos/107/node38.html

36 Unit of angular measure (radian)

4.3 Spherical astronomy

Spherical astronomy is the study of the directions with respect to some fixed point,
line or plane in which celestial bodies (e.g., planets, comets, stars, galaxies) appear
to lie. As celestial bodies are, generally speaking, extremely distant, they seem to
lie on the surface of a sphere, known as the celestial sphere.

4.3.1 Spherical triangles

Some definitions
A sphere is a surface formed by rotating a semicircle about its diameter through
2π radians. A straight line drawn from the mid-point of the diameter (the centre of
the sphere) to the surface of the sphere is the radius of the sphere.
A great circle is the intersection of the sphere with any plane passing through
the centre of the sphere. The Earth’s equator is an example of a great circle.
A small circle is formed by any other intersecting surface that does not pass
through the centre of the sphere. A circle of latitude on the Earth’s surface, other
than the equator, is a small circle.
The axis of a great circle is the diameter of the sphere at right angles to that
great circle. The points of intersection of the axis and the sphere are known as the
poles of the great circle. (e.g., the poles of the great circle on the Earth known as
the equator are the north and south geographic poles).
A spherical triangle is that part of the surface of a sphere bounded by the arcs
of three great circles. Combinations of small circles and great circles do not make
spherical triangles.
Each spherical triangle, like its plane counterpart, may be described by six num-
bers: the lengths of the three sides (a, b, c in Figure 4.3 measured along the arcs)
and the three included angles at points A, B and C. The relationships between these
arcs and angles are given by:
cos a = cos b cos c + sin b sin c cos A
sin a sin B = sin b sin A (4.5)
sin a cos B = cos b sin c − sin b cos c cos A

4.3.2 Coordinate systems in astronomy

The terrestrial coordinate system
The Earth is a good approximation to a sphere. The axis of rotation of the Earth,
termed the polar axis, intersects the surface of the Earth at the north and south
geographic poles. The great circle whose plane is at right angles to the polar axis
 4.3 Spherical astronomy 37
A

a
C

Figure 4.3. A spherical triangle.

is called the terrestrial equator. Other great circles that are at right angles to the
terrestrial equator and pass through the north and south poles are called terrestrial
meridians.
Geographical longitude is the angle between a terrestrial meridian circle and
an arbitrarily chosen reference meridian circle called the prime meridian. This
great circle passes through a telescope known as the Airy transit circle located
at the Royal Observatory, Greenwich, in the United Kingdom. All longitudes are
commonly measured from this prime meridian, either east or west from 0◦ to 180◦ .
Geographical latitude is the angular distance normally measured north or south
of the terrestrial equator from 0◦ to 90◦ along the meridian circle through the point
on the Earth’s surface of interest (see Figure 4.4).
Using the radian equivalents given in Section 4.2.1 above, it is a trivial exer-
cise to convert latitudes (φ) and longitudes (λ) from degrees, minutes and seconds
to radians. Retaining the prime meridian in Greenwich but measuring longitude,
increasing eastwards, from 0 to 2π rad, and measuring latitude from 0 at the south
pole to π rad at the north pole, we may rewrite the longitude and latitude of Mauna
Kea Observatory in the Hawaiian Islands (remembering that its given sexagesi-
mal longitude is in the western hemisphere, so λ = 360 − λ) from 155◦ 28 18W;
19◦ 49 36 N to:
λ = 3586.157 103 mrad φ = 1916.836 940 mrad
Note that three decimal places corresponds to locating the observatory on the surface
of the Earth to approximately ±7 m if the Earth were a perfect sphere.12

12 The length of an arc of 1 radian is equal to the radius of the Earth, 6738.136 km at the Equator according to
Cox (2000), so an arc length of 1 mrad = 6.738 136 km and 0.001 mrad = 6.738 m.
38 Unit of angular measure (radian)
NP

n
idia
mer
P

p r i me
φ

G λ R
equat
or
φSP

SP

Figure 4.4. The geographical coordinates of the point P are longitude λ (the thick
grey arc GR) and latitude, either φ (the thick black arc PR), when measured north
or south from the equator, or φSP (the thick black arc SPRP) when measured from
the south pole northwards. The labelled points are: NP (north geographic pole),
SP (south geographic pole), G (the intersection of the prime meridian with the
equator), and R (the point of intersection of the great circle through the poles and
the point P with the equator).

The horizontal or altitude–azimuth coordinate system

The great circles used in the horizontal coordinate system are the horizon, the pro-
jection of the terrestrial horizon of the observer to the celestial sphere and the
vertical circle, which is at right angles to the horizon and passes through the celes-
tial body of interest, and the zenith, the point in the sky directly above the observer.
The vertical circle that passes through the north and south points of the horizon
and the zenith is known as the prime meridian. The vertical circle at right angles to
the prime meridian defines the east and west points where it intersects the horizon
(see Figure 4.5).
The location of the celestial body is given by its azimuth, the angle measured
along the horizon from north through east, south and west from 0 to 2000π mrad
to the point of intersection of the vertical circle passing through the body, and its
altitude (or sometimes the complement of altitude, the zenith distance), the angle
 4.3 Spherical astronomy 39
Z

W
z
O
h
S N
A

horizon
H E

Figure 4.5. Altazimuth coordinates: azimuth angle, A is the angle NOH measured
(dark grey arc) from north (N) through east (E) to the intersection point on the
horizon (H) of the great circle through the zenith point (Z) and the star (P). The
altitude, h, is the angle POH (thick black arc). The zenith distance, z, is the angle
ZOP (light grey arc).

between the horizon and the body measured along the vertical circle through the
zenith point. Hence altitude and zenith distance would lie between 0 and 500π mrad.
The time and date of the observation and the location at which the observation was
made must also be given, as both altitude and azimuth vary with time, date and
geographical position.

The equatorial coordinate system

The equatorial system of coordinates in astronomy uses the projection of the Earth’s
equator on to the celestial sphere as the fundamental circle, which is known as the
celestial equator. The projection of the Earth’s north pole on to the celestial sphere
is known as the north celestial pole (NCP in Figure 4.6) and the projection of the
Earth’s south pole as the south celestial pole (SCP in Figure 4.6). The zero point
of right ascension is defined to be the intersection of the celestial equator with the
ecliptic (see below) at the position where the Sun moves from south of the celestial
equator to north. This point, known as the Vernal Equinox or the First Point of
Aries, is point ϒ in Figure 4.6. Right ascension is measured eastwards from this
point along the celestial equator.
40 Unit of angular measure (radian)
NCP

equa Υ
tor
α B

δ SCP

SCP

Figure 4.6. Equatorial coordinates: right ascension, α, measured from the first
point of Aries (ϒ) to point B in an easterly direction (thick grey arc ϒB), decli-
nation, δ, measured from either the celestial equator at point B (thick black arc
BA) or δSCP , measured from the south celestial pole northwards through B to
A along the great circle through the celestial poles and point A (thick black arc
SCPBA).

The ecliptic coordinate system

The ecliptic may be defined as the great circle traced out by the apparent path of the
Sun across the sky during the course of one sidereal year. The ecliptic intersects the
celestial equator at two points, termed the First Point of Aries and the First Point
of Libra (collectively the equinoctial points). The angle between the celestial
equator and the ecliptic is called the obliquity of the ecliptic and was equal to
approximately 409.092 6 mrad at the standard epoch J2000.0.
Coordinates measured in the ecliptic system are celestial longitude (λ) and
celestial latitude (β) (see Figure 4.7). Celestial longitude is measured eastwards
from the first point of Aries along the ecliptic. Celestial latitude is measured north-
wards from the ecliptic (0 mrad) to the north ecliptic pole (+500π mrad) and
southwards from the ecliptic to the south ecliptic pole (−500π mrad) or alterna-
tively, as βSEP , it may be measured from 0 mrad at the south ecliptic pole (SEP)
through +500π mrad at the ecliptic and on to +1000π mrad at the north ecliptic
pole (NEP).
 4.3 Spherical astronomy 41
NEP

β
D
equ
ato λ
r β SEP
ϒ
ε E
ecliptic

SEP

Figure 4.7. Ecliptic coordinates: celestial longitude, λ, measured from the first
point of aries (ϒ) eastwards along the ecliptic (thick grey arc ϒD) and celestial
latitude, β, measured either north or south of the ecliptic along a great circle
including the north and south ecliptic poles and the celestial object of interest, S,
(thick black arc DS) or northwards from the south ecliptic pole (thick black arc
SEPEDS). The angle DϒE,  is the obliquity of the ecliptic – the angle between
the plane of the Earth’s equator and its orbit about the Sun.

Disadvantages of the equatorial and ecliptic coordinate systems

The major problem with the equatorial and ecliptic coordinate systems is that their
fundamental reference framework is not fixed but changes with time. Gravitational
interactions between the bulge of the Earth’s equator and the Sun and the Moon
cause the axis of the Earth to describe a circle of radius equal to the obliquity of
the ecliptic on the celestial sphere in a period of approximately 26 000 years. This
effect is known as lunisolar precession. Perturbations from the major planets on
the orbit of the Earth cause a smaller effect, known as planetary precession. The
combination of the two is known as general precession. The consequence of pre-
cession is slow westward movement of the Vernal Equinox along the ecliptic at an
annual rate of approximately 0.243 8 mrad (or 243.82 μrad). Superimposed on the
general precessional motion is an extra movement (nutation) caused by the plane
of the Moon’s orbit rotating with respect to the ecliptic in a period of 18.6 years.
When observations made over a period of years are to be compared, it is necessary
for them all to be reduced to a common date or epoch. To facilitate intercomparison
42 Unit of angular measure (radian)

of observations a fundamental epoch (the precise moment in time for which the
observations are specified) is adopted, to which all appropriate observations are
reduced. The resulting coordinates are referred to as mean coordinates. Star posi-
tions are listed in catalogues and plotted on star charts and atlases reduced to a
fundamental epoch. It is usual for a particular fundamental epoch to be used for
around 50 years; that currently in use (2011) is 2000.0.

The galactic coordinate system

The projection of the mean plane of the Milky Way galaxy on to the celestial
sphere defines the great circle termed the galactic equator. The precise location
of the galactic equator is mainly determined using radio frequency observations of
the distribution of neutral hydrogen gas. The same technique is used to assign the
position of the galactic centre, which is taken as the zero point of the coordinate
galactic longitude l, (measured eastwards from the centre along the galactic equa-
tor). Galactic latitude b, may be measured northwards from 0 mrad at the galactic
equator to the north galactic pole at +500π mrad and southwards from the galactic
equator to the south galactic pole at −500π mrad or, alternatively, as bSGP , it may
be measured from 0 mrad at the south galactic pole through +500 π mrad on the
galactic equator to +1000 π mrad at the north galactic pole (see Figure 4.8).
The equatorial coordinates of the galactic centre (αGC , δGC ) and the north galactic
pole (αNGP , δNGP ) at epoch J2000.0 are given by Murray (1989) as:

αGC = 4649.644 328 mrad δGC = 1065.764 84 mrad

αNGP = 3366.032 942 mrad δNGP = 2044.273 63 mrad

These equatorial coordinates were calculated from the original epoch 1950.0
values agreed to by the International Astronomical Union when setting up the new
system of galactic coordinates in 1958 (Blauuw et al., 1960).
The values of the right ascension (α0 ) and galactic longitude (l0 ) of the ascending
node of the galactic plane on the J2000 equator and the inclination of the galactic
equator to the celestial equator (γ ), derived from Cox (2000), are:

α0 = 4936.829 mrad
l 0 = 574.770 mrad
γ = 1097.319 mrad

Although the coordinate frame in the galactic coordinate system does change
with time, the effect is extremely small in comparison with either the equatorial or
ecliptic coordinate systems. The galactic coordinate system is of particular value
in galactic structure studies.
 4.3 Spherical astronomy 43

S
NGP

r
ato
b

qu
ator
equ

ic e
bSGP

t
lac
l

ga
γ

l0
galactic SGP
O
centre

Figure 4.8. Galactic coordinates: galactic longitude, l, measured along the galactic
equator from the galactic centre (dark grey arc OLE), galactic latitude, b, measured
either north or south of the galactic equator (thick black arc ES) or from the south
galactic pole northwards, bSGP , (thick black arc SGPES). The longitude of the
ascending node of the galactic plane, l0 , is the dark grey arc OL and the inclination
of the galactic equator to the celestial equator is the angle γ .

4.3.3 Relationships between astronomical coordinate systems

Altazimuth to equatorial and vice versa

sin z sin A = − cos δ sin H

sin z sin A = sin δ cos φ − cos δ cos H sin φ
cos z = sin δ sin φ + cos δ cos H cos φ (4.6)
cos δ sin H = cos z cos φ − sin z cos A sin φ
sin δ = sin φ cos z + cos φ sin z cos A
where φ is the latitude of the observer, z is the object’s zenith distance (the com-
plement of its altitude) and H is its local hour angle related to the object’s right
ascension and the local sidereal time (LST) of the observation by:
H = LST − α. (4.7)
where LST may be defined as the circle of right ascension transiting the prime
meridian at the time of observation.
44 Unit of angular measure (radian)

Equatorial to ecliptic
To convert the equatorial coordinates (α, δSCP ), measured in radians, using the south
celestial pole as the zero point of the declination measures, to ecliptic coordinates
(λ, β), also measured in radians, with the zero point of celestial latitude being the
ecliptic plane and  being the obliquity of the ecliptic, the following expressions
may be used:
π
δ = δSCP −
2
sin β = sin δ cos  − cos δ sin α sin 
cos δ sin α cos  + sin δ sin 
sin λ = (4.8)
cos β
cos δ cos α
cos λ =
cos β
cos δ sin α cos  + sin δ sin 
tan λ =
cos δ cos α
π
βSEP = β +
2
To assign the correct value for λ in four quadrants, the following conditions for
the sign of each of the values of sine, cosine and tangent determine which expression
is to be used to calculate the value of λ. (∧ is the Boolean symbol for ‘and’.)


 sin−1 λ if sin λ ≥ 0 ∧ cos λ ≥ 0 ∧ tan λ ≥ 0
 π − sin−1 λ if sin λ ≥ 0 ∧ cos λ < 0 ∧ tan λ < 0
λ=  π − sin−1 λ if sin λ < 0 ∧ cos λ < 0 ∧ tan λ ≥ 0 (4.9)

 2π + sin−1 λ if sin λ < 0 ∧ cos λ ≥ 0 ∧ tan λ < 0

Using as an example, the equatorial coordinates of the star Capella given above
in radians, (α, δSCP ) = (1.381 817, 2.373 61), may be transformed to ecliptic coor-
dinates using equations (4.8) and conditions (4.9) with  = 0.409 092 6 (value at
J2000.0) to yield: (λ, β) = (1.428 690, 0.399 06), where the celestial latitude is given
with reference to the ecliptic; relative to the south ecliptic pole, the value would be
βSEP = 1.979 889, obtained simply by adding π /2 to β.
There is perhaps a stronger case with ecliptic latitude than with equatorial dec-
lination to present the value of the coordinate with reference to the ecliptic plane,
as this is of physical importance when considering Solar System bodies.

Transformation of equatorial to galactic coordinates

To convert the equatorial coordinates (α, δSCP ), measured in radians, using the south
celestial pole as the zero point of the declination measures, to galactic coordinates
 4.3 Spherical astronomy 45

( l , b ), also measured in radians, with the zero point of galactic latitude being the
galactic equator and γ being the angle of inclination between the galactic plane
and the equatorial plane, the following expressions may be used:
π
δ = δSCP −
2
sin b = sin δ cos γ − cos δ sin( α − α0 ) sin γ
cos δ sin( α − α0 ) cos γ + sin δ sin γ
sin (l − l0 ) = (4.10)
cos b
cos δ cos ( α − α0 )
cos (l − l0 ) =
cos b
cos δ sin( α − α0 ) cos γ + sin δ sin γ
tan (l − l0 ) =
cos δ cos( α − α0 )
π
bSGP = b +
2
where α0 is the right ascension of the galactic centre at J2000.0 and l0 is the
galactic longitude of the point of intersection of the celestial equator and the galactic
equator, known as the longitude of the ascending node (i.e., where b = 0 mrad or
bSGP = 1570.796 (= π/2) mrad).
To assign the correct value for l in four quadrants, the following conditions for
the sign of each of the values of sine, cosine and tangent determine which expression
is to be used to calculate the value of l. (∧ is the Boolean symbol for ‘and’.)


 sin−1 (l − l0 ) if sin (l − l0 ) ≥ 0 ∧ cos (l − l0 ) ≥ 0 ∧ tan (l − l0 ) ≥ 0
 π − sin−1 (l − l0 ) if sin (l − l0 ) ≥ 0 ∧ cos (l − l0 ) < 0 ∧ tan (l − l0 ) < 0
l − l0 = 

 π − sin−1 (l − l0 ) if sin (l − l0 ) < 0 ∧ cos (l − l0 ) < 0 ∧ tan (l − l0 ) ≥ 0
 2π + sin−1 (l − l0 ) if sin (l − l0 ) < 0 ∧ cos (l − l0 ) ≥ 0 ∧ tan (l − l0 ) < 0
(4.11)

thence
l = (l − l0 ) + l0 (4.12)
Equation (4.12) can produce galactic longitudes greater than 2π , which conditional
statement (4.13) removes.

 l if l < 2π
l=  l − 2π if l ≥ 2π (4.13)

Using once again as an example the equatorial coordinates of the star Capella
given above as (α, δ) = (1.381 817 rad, 2.373 61 rad), which may be transformed
to galactic coordinates using Equations (4.10) and conditions (4.11) with γ =
1.097 319 rad (value at J2000.0) to yield: (l, b) = (2.837 704 rad, 0.079 699 rad),
46 Unit of angular measure (radian)

where the galactic latitude is given with reference to the galactic equator. Relative
to the south galactic pole the value would be bSGP = 1.650 496 rad, obtained simply
by adding π/2 to b.
The use of galactic coordinates is of greatest value when studying the distribution
of galactic objects with reference to the galactic plane and centre.

4.4 Angular distances and diameters

The distance AB, between the two celestial bodies A and B shown in Figure 4.9
located at (αA , δA ) and (αB , δB ) is given in radians by:

−→
AB = cos−1 (sin δA sin δB + cos δA cos δB cos (αB − αA )) (4.14)

If the positions of A and B are given in ecliptic or galactic coordinates, then

Equation (4.12) may simply be rewritten with (λA , βA ) and (λB , βB ) or (lA , bA ) and
(lB , bB ) substituted for (αA , δA ) and (αB , δB ).

4.4.1 Distances between pairs of astronomical objects

Select the positions of the pairs of stars α Ori and β Ori, and β Cas and β
Car from Table 4.1, with declination measured from the celestial equator (i.e.,
δ = δSCP − 500π ). Transform each number pair to radian measure by dividing by
1000, then substitute into Equation (4.14) above. The distance between α Ori and
β Ori is 324.737 mrad. For the pair β Cas and β Car, the measured separation is
2773.878 mrad.

P
α π/2 − δ B
Β −α
Α

π/2 − δA
B

A r
equato

Figure 4.9. Measurement of angular distances.

 4.4 Angular distances and diameters 47

Table 4.2. Further examples of radian measures: apparent sizes and separations

Sexagesimal Radian
Celestial object system system Notes, Reference
Comet Tebbutt (1861 II) >100◦ >1.75 rad length of tail Chambers (1889)
LMC 10◦ 45 .7 187.83 mrad galaxy Cox (2000)
M31 3◦ 10 .5 55.41 mrad galaxy Cox (2000)
Pleiades 2◦ 34.91 mrad galactic cluster Cox (2000)
Sun 31 59 .26 9.30 mrad star Cox (2000)
Moon 31 05 .2 9.04 mrad satellite Cox (2000)
Helix nebula 12 3.49 mrad planetary nebula Cox (2000)
ω Cen 8 .36 2.43 mrad globular cluster Cox (2000)
Jupiter 48 .9 237.07 μrad planet White (2008)
Saturn’s rings 44 .9 217.68 μrad planetary rings White (2008)
α Cen 7 .53 36.51 μrad double star separation White
(2008)
Mira 0 .060 290.89 nrad stellar diameter Karovska &
Sasselov (2001)
Aldebaran 0 .020 96.96 nrad stellar diameter Harris et al.
(1963)
 Aur 0 .00227 11.01 nrad stellar diameter Stencel et al.
(2008)

The angular diameters or separations of a selection of celestial bodies (comet,

galaxies, globular and galactic clusters, planetary nebulae, stellar diameters (includ-
ing the Sun), double star separations, planets and Saturn’s rings) are given in
Table 4.2 in sexagesimal and radian (or subdivisions thereof) measures.

4.4.2 Field or plate scale determination

The plate scale of a photographic plate or a CCD is a measure of the angular
distance on the celestial sphere imaged on the detector per unit linear distance
measured. A common way of expressing plate scale is in arcseconds per millimetre
(arcsec mm−1 ). The equivalent SI expression is μrad mm−1 .

Example: determine the plate scale of the AAO Schmidt telescope (formally
known as the UK Schmidt telescope)
The measured distance r, on an AAO Schmidt telescope photographic plate,
between  Ori and ζ Ori is:

r = 72.654 mm
48 Unit of angular measure (radian)

The angular distance a between the two stars is calculated using Equation (4.14),
substituting for the spherical coordinates, in radians:

a = cos−1 [sin δ sin δ0 + cos δ cos δ0 cos (α − α0 )] = 0.023 661 rad

Hence the plate scale (= a/r) is:

a
= 3.257 × 10−4 rad mm−1
r
or 0.326 mrad . mm−1 or 326 μrad . mm−1

4.5 Steradian
The steradian is the derived SI unit of solid angle that has its apex at the centre of
a sphere and subtends an area at the surface of the sphere equal to the square of the
radius of the sphere.
It is a dimensionless quantity with the symbol sr, which may be expressed in
terms of SI base units as ([L]2 . [L]−2 ).

The number of steradians on the entire sky is given by the following expression
(4.15), where α is right ascension measured from 0 rad at the first point of Aries
increasing eastwards and δSCP is declination measured northwards from 0 rad from
the south celestial pole:
 2π  π
A= sin δSCP d δSCP d α = 4π (4.15)
0 0

Hence the number of steradians on a sphere is given by:

4π = 12.566 371

4.5.1 Conversions between sexagesimal and steradian measures

The units of solid angle currently in common use in astronomy are the square
degree, square arcminute and square arcsecond. To convert to SI units and vice
versa, the following factors or their inverses may be used.
The number of square degrees in a sphere:
 180 2
4π = 41 252.961 249
π
the number of square degrees in a steradian:
 180 2
= 3282.806 35
π
 4.5 Steradian 49

one square degree in steradians:

 π 2
= 0.000 304 617 sr = 304.617 μsr
180
one square arcminute in steradians:
 π 2
= 8.461 595 × 10−8 sr = 84.615 95 nsr
180 × 60
one square arcsecond in steradians:
 π 2
= 2.350 443 × 10−11 sr = 23.504 43 psr
180 × 3600

4.5.2 Area of the constellation Crux

The IAU boundaries of the constellation Crux are defined by two lines of constant
declination and two of constant right ascension.
The approximate equatorial coordinates in radians of the northernmost
(δmax ) = 0.598, southernmost (δmin ) = 0.441, easternmost (αmax ) = 3.393 and west-
ernmost (αmin ) = 3.128 points were determined from the digital sky atlas The
SkyTM .
The area of the constellation Crux is then derived by evaluating the integral:
 3.393  0.598
ACrux = sin δ d δ d α = 0.020 6 sr (4.16)
3.128 0.441

In sexagesimal units, this computed area is 67.7 square degrees, in tolerable

agreement with the value of 68 square degrees given by Moore (2001).
For constellations with boundaries that follow more than two circles of right
ascension and/or declination, a more general formulation for the total area Atot
would be:
n  αmax  δmax
i i
Atot = sin δ d δ d α (4.17)
i=0 αmini δmini

4.5.3 Further examples of angular area measurement in astronomy

Example: determine the angular area of the Sun and Moon
From Table 4.2 above, the mean apparent angular radius of the Sun is 4.65 mrad.
Assuming the Sun to be spherical, its mean apparent angular area A is

A = π(4.65)2
= 67.93 mrad 2 or 67.93 μsr (4.18)
50 Unit of angular measure (radian)

For the Moon, the mean apparent angular radius derived from Table 4.2 is 4.57
mrad and hence its area is

AMoon = 65.61 mrad 2 or 65.61 μsr (4.19)

Example: determine the visible angular area of the planet Venus on 2009
February 28
On 2009 February 28, the angular radius of the planet Venus was 107.6 μrad and
the percentage, p, of the disc illuminated by the Sun was 20%. Hence, the angular
area of the illuminated portion of the disc was:

AVenus = p π(107.6)2
= 7278 μrad 2 or 7 278 psr (4.20)

Example: determine the angular area of the disc of the F supergiant star 
Aurigae
The recently measured radius (Stencel et al., 2008) of the F-type supergiant pri-
mary of the eclipsing binary star  Aurigae determined using the Palomar Testbed
Interferometer was 5.5 nrad. Assuming a uniform stellar disc, the corresponding
apparent angular area of the  Aur primary is:

AAur = π (5.5)2
= 95 nrad 2 = 95 asr (attosteradians) (4.21)

Example: determine the angular area of the Local Group Galaxy M31
The measured angular diameter of M31 is given (Cox, 2000) as 8.7 arcmin or
522 arcsec within the boundary set by the B = 25 mag . arcsec−2 isophote, and the
axial ratio ab as 0.32. The area (πab) of a fitted ellipse with semi-major axis a and
semi-minor axis b is:

A = π ab
A = 6.848 × 104 arcsec2
= 1.610 × 10−6 sr (4.22)
= 1.610 μsr
= 1.610 mrad 2

To convert the isophote unit from magnitudes per arcsecond squared to janskys
per steradian, the following method may be used: the B = 25 mag . arcsec−2 is
10−10 times fainter than the B = 0 mag . arcsec−2 isophote. B = 0 is equivalent to a
 4.5 Steradian 51

monochromatic flux density of 4000 Jy (Bessell, 1992), so B = 25 is equivalent

to a monochromatic flux density of 400 nJy (4 × 10−7 Jy). Using the conver-
sions between arcsec2 and sr given above, the monochromatic flux density per
steradian is:

4 × 10−7
= 17 018 Jy . sr −1
2.350 443 × 10−11
= 1.701 8 × 10−8 Jy . psr −1 (4.23)
= 1.701 8 Jy . μrad −2

So the apparent angular area of M31 to the limiting isophote of 17018 Jy . sr−1 is
1.610 × 10−6 sr.

Number densities of galaxies

Counting the numbers of objects in cells of similar shape and size has been used to
determine the large-scale distribution of astronomical objects for more than half a
century. The general luminosity function of stars (number of stars per unit brightness
interval per unit area) was first studied in detail by Bok (1937). Some 20 years later,
Shane & Wirtanen (1954) counted the number of galaxies on photographic plates
in unit areas to identify clusters of galaxies.
As an example of counting galaxies in cells, an area of 256 μsr (16 mrad ×
16 mrad), centred on the cluster of galaxies A3266 was measured using the COS-
MOS automated photographic-plate-measuring machine at the Royal Observatory
Edinburgh (Longair, 1989). The equatorial coordinates for the centre of A3266 at
J2000.0 are approximately:

α = 1183 mrad δSCP = 498 mrad

For each image detected by COSMOS, 13 parameters relating to the image centroid,
its size, brightness, orientation and shape were measured, which enabled separate
lists of galaxy images, star images and non-standard images to be compiled. In total,
4328 images were measured, of which 1256 were classified as galaxies and three
as bright galaxies. The field was divided into a grid of 256 (16 × 16) equal-sized
(1 mrad × 1 mrad = 1 μsr) areas, and the number of galaxies in each counted (see
Table 4.3). These numbers were processed by a contour-plotting routine, the output
of which is shown in Figure 4.10.
The mean of the cell count numbers and the standard deviation about that mean
are 4.3 ± 3.9 galaxies per cell, the median cell count is 3.5 galaxies per cell and the
mode is 2 galaxies per cell.
52 Unit of angular measure (radian)

Table 4.3. Cell counts (per mrad2 ) for galaxies in the neighbouring field and
cluster of galaxies A3266

0 4 1 1 0 3 4 6 2 0 3 3 6 3 2 6
5 4 6 4 0 3 2 2 4 2 2 1 8 6 4 2
3 1 4 2 0 1 1 5 4 2 0 8 4 11 17 3
2 2 2 2 3 2 1 4 2 1 3 4 4 4 6 18
1 1 1 7 3 3 5 8 4 4 2 12 10 3 7 1
2 3 3 1 6 2 7 5 8 8 11 5 9 8 2 0
4 3 7 2 4 7 7 16 7 5 7 2 6 8 3 2
2 0 3 5 1 9 8 27 20 8 6 7 13 9 0 1
3 3 3 5 3 1 5 13 27 12 5 4 6 3 6 4
4 3 0 2 4 7 8 8 17 7 17 2 2 6 4 4
1 4 1 2 2 2 6 5 9 5 5 7 9 5 2 3
5 2 3 3 0 3 3 3 7 9 6 4 3 4 2 4
2 1 1 4 1 5 2 3 2 8 1 6 5 6 1 5
5 3 2 1 4 5 0 5 6 4 3 2 4 1 2 1
4 3 6 4 1 4 2 1 2 4 2 3 1 1 3 1
4 4 2 0 2 5 0 5 5 4 0 2 4 3 2 5

10
.8
5
8
10.
5.4

10.
8
.8
10

5.4
Δδ mrad

.8
10 16
.6

.2
21

0
21.6

4
5.

5.4

–5 5. 5.4
5.4

–5 0 5
Δα mrad

Figure 4.10. Number of galaxies per microsteradian, from data obtained using the
COSMOS measuring machine, of the cluster of galaxies A3266 and its surrounding
area from which all the star images have been removed.

Figure 4.10 clearly shows that A3266 is in fact two adjacent clusters of galaxies
and not just one, which Henriksen & Tittley (2002), using the X-ray space-based
CHANDRA Observatory, have recently shown to be in the process of merging.
 4.6 Summary and recommendations 53

4.6 Summary and recommendations

4.6.1 Summary
The SI unit of angular measure is the radian. The units commonly used by
astronomers to define the positions of celestial objects are hours, minutes and sec-
onds for the right ascension coordinate and degrees, minutes and seconds for the
declination coordinate. Transformations from these units to radians and recom-
mended submultiples of radians are given. The common types of astronomical
coordinate systems are described and the relationships linking them given. The
coordinates of different types of astronomical bodies, their angular distances apart
and their angular sizes and areas are presented in tables and by way of worked
examples.

4.6.2 Recommendations
The ease with which the location of points on the terrestrial and celestial spheres
may be made is mentioned, with a pair of numbers replacing the current six. The step
from sexadecimal to radian measure is considered too large by some astronomers
so, as a first step in this process, using decimal degrees rather than degrees, minutes
and seconds may make the final changeover to radians easier.
 5
Unit of time (second)

5.1 SI definition of the second

The second is the duration of 9 192 631 770 periods of the radiation corresponding
to the transition between the two hyperfine levels of the ground state of the caesium
133 atom.
This definition refers to a caesium atom in its ground state at a temperature of 0 K.
The dimension of time is [T], its unit is the second and its symbol is s.

5.2 Definition of time

Funk & Wagnalls New Standard Dictionary of the English Language (Funk et al.,
1946) defines time, inter alia, as:

The general idea, relation, or fact of continuous or successive existence; or the abstract con-
ception of duration as limitless, capable of division into measurable portions, and essentially
comprising the relations of present, past and future.

A system of reckoning or measuring duration; as solar time; sidereal time;

mean time.

5.3 Systems of time or time scales

There are two major systems of time: those based on the Earth’s rotation and the
orbital motions of the Earth, Moon and planets, known as dynamical time; and
those based on atomic clocks13 and known as atomic time.
A time system or scale may be specified by two numbers, the origin from which
the time intervals are to be measured and the number of predefined unit scale
intervals measured since the time of origin (Leschiutta, 2001).

13 See, e.g., http://tycho.usno.navy.mil/clockdev/cesium.html

54
 5.3 Systems of time or time scales 55

5.3.1 Dynamical time

Dynamical time may be thought of as the independent variable in the equations
that describe the motions of the bodies in the Solar System. It was developed from
the natural system of timekeeping based on the apparent motion of the Sun across
the celestial sphere (see, e.g., Lang (2006), Cox (2000) and the USNO website14
on which the greater part of this section is based).

5.3.2 Atomic time

By the middle of the twentieth century the relative accuracy of the unit dynamical
time interval was, at best, around 1 part in 108 , which was proving inadequate for
particular applications in astronomy, spectroscopy and telecommunications. This
accuracy limit was imposed by the irregularities in the period of the Earth’s rotation.
The continued developments of atomic clocks in the 1950s led to an improvement in
the accuracy of time measurement to 1 part in 109 , which in turn led to the discovery
of the variability in the Earth’s rotation period. Currently, the most accurate atomic
clock is a type known as a caesium fountain atomic clock, with a measurement
uncertainty of around 5 parts in 1016 .15

5.3.3 Time systems currently in use

Atomic time is defined above in terms of the duration of a specified number of
cycles of radiation corresponding to the transition of two hyperfine levels of the
ground state of 133 Cs at absolute zero.

TAI, International Atomic Time, is calculated from a statistical analysis of individ-

ual frequency standards and time scales based on atomic clocks situated throughout
the world.

UT, Universal Time, based on the mean solar day, is the time system used for all civil
timekeeping. UT0 is rotational time for a particular location on the Earth, uncor-
rected for shifts in longitude due to polar motion. UT1 is the rotational time corrected
for such shifts, though it is still non-uniform, being subject to the irregularities in
the Earth’s rotation.

UTC, Coordinated Universal Time, is the time scale distributed by means of radio
time signals, satellites, radio and TV broadcasts. UTC differs from TAI by an

14 www.usno.navy.mil/USNO/time/master-clock/systems-of-time
15 See http://tycho.usno.navy.mil/clockdev/cesium.html
56 Unit of time (second)

integral number of seconds and is kept to within 0.9 s of UT1 by means of the
irregular introduction of integer leap seconds. UTC has replaced GMT (Greenwich
Mean Time) as an international time standard.

TT or TDT, Terrestrial Dynamic Time, is the idealized time based on the rotation
period of the geoid. The time unit is 1 day consisting of 86 400 s in SI units. It is
approximately related to TAI by the expression:

TDT = TAI + 32.184 s (5.1)

TDB, Barycentric Dynamical Time, is the relativistically transformed time for

referring equations of motion to the barycentre of the Solar System. It is defined to
differ from TDT solely by periodic variations such that:

TDB = TDT + 0s .001 658 sin g + 0s .000 014 sin 2g (5.2)

g = 357◦ .53 + 0◦ .985 600 28 (JD − 2 451 545.0) (5.3)
= 6.240 075 675 + 0.017 201 969 (JD − 2 451 545.0) rad (5.4)

TCG, Geocentric Coordinate Time, is a coordinate time related to the centre of

mass of the Earth as its spatial origin. It is related to TDT by:

TCG = TDT + 6.969 290 4 × 10−10 (JD − 2 443 144.5) × 86 400 s (5.5)

where JD is the julian day number equal to the number of days that have elapsed
since Greenwich Mean Noon on 1 January 4713 BCE, Julian proleptic calendar
(the Julian Date is the julian day plus the elapsed time since the preceding noon).

TCB, Barycentric Coordinate Time, is a coordinate time related to the barycentre

of the Solar System as its spatial origin. TCB differs from TDB in rate and is related
to it by:

TCB = TDB + 1.550 506 × 10−8 (JD − 2 443 144.5) × 86 400 s (5.6)

ST, Sidereal Time, is defined to be the hour angle of the First Point of Aries, ϒ,
the point of intersection of the celestial equator and the ecliptic (see Chapter 4). In
other words, the right ascension in transit over the prime meridian (the north–south
line) at a given terrestrial longitude (λ). The relationship between the local sidereal
time (LST) and the Greenwich sidereal time (GST) is:

LST = GST + λ (5.7)

 5.3 Systems of time or time scales 57

where λ is measured eastwards from Greenwich. The local hour angle (LHA) of a
celestial object S at right ascension, αS , is:

LHAS = LST − αS (5.8)

The date of any such measurement of the local hour angle has to be given, as the
location of ϒ is not fixed due to the effects of precession and nutation.

MJD, Modified Julian Date, which begins and ends at midnight, is defined in terms
of the julian day number (see above) as:

MJD = JD − 2 400 000.5 (5.9)

For precise applications the time scale should be specified.

5.3.4 Multiples of the second

Multiples of the second in common use are as follows:

1 minute = 60 s
1 hour = 60 m = 3 600 s
1 day = 24 h = 1 440 m = 86 400 s

5.3.5 Leap second

In order to ensure that the difference between time determined using atomic clocks
and that determined from the Earth’s rotation does not exceed 0.9 s, civil time (UTC)
is occasionally adjusted in one second increments.
An example of the application of the leap second is the sequence of dates created
at the end of December 2008:16

2008 December 31, 23 h 59 m 59 s

2008 December 31, 23 h 59 m 60 s
2009 January 01, 00 h 00 m 00 s

The difference (UTC−TAI) applied to TAI was:

−33 s from 2006 January 01, 0 h UTC to 2009 January 01, 0 h UTC;
−34 s from 2009 January 01, 0 h UTC until further notice.

16 See http://tycho.usno.navy.mil/bulletinc2008.html
58 Unit of time (second)

5.3.6 Relationships between mean solar time and mean sidereal time
The mean sidereal rotation period of the Earth is equivalent to (The Astronomical
Almanac, 1995):
86 164.090 54 s (SI units) or
23 h 56 m 04.090 53 s (mean solar time) or
0.997 269 566 33 d (mean solar time)
One mean solar rotation period of the Earth is equivalent to:
1.002 737 909 35 d (mean sidereal time) or
24 h 03 m 56.555 37 s (mean sidereal time)
These values relate to the year 1995.

5.3.7 The month

The length of the month depends on the reference point chosen to define the start
and end of the period. Table 5.1 gives the lengths of the mean months for 1995.0
(The Astronomical Almanac, 1995).

5.3.8 The year

The Julian year is defined to be 365.25 d precisely and forms the basis for the Julian
calendar. The length of the Julian year may also be written as:

1 y = 365.25 d = 8 766 h = 525 960 m = 31 557 600 s (SI unit).

The lengths of some other types of year are given in Table 5.2.

5.3.9 ISO8601 standard on dates and time

The International Organization for Standardization (ISO) has prepared a standard,
ISO8601, which sets out a method for numerically representing the following:
Date: Time of the day: Coordinated Universal Time (UTC): Local time with
offset to UTC.
Date and time: Time intervals: Recurring time intervals.
Formats of particular use in astronomy are:

1. Calendar date: YYYY–MM–DD, where YYYY is the Gregorian calendar year,

MM is the month of the year (01 = January, 02 = February, 03 = March . . . 12 =
December) and DD is the day of the month from 01 to 28, 29, 30 or 31.
For example, 1944–12–08 is the 8th December 1944.
 5.3 Systems of time or time scales 59

Table 5.1. Name, definition and length of types of month

Month type length length length

Reference point d d h m s s
Draconian 27.212 221 27 05 05 35.9 2 351 135.9
Node
Tropical 27.321 582 27 07 43 04.7 2 360 584.7
Equinox
Sidereal 27.321 662 27 07 43 11.6 2 360 591.6
Fixed star
Anomalistic 27.554 550 27 13 18 33.1 2 380 713.1
Perigee
Synodic 29.530 589 29 12 44 02.9 2 551 442.9
New Moon

Table 5.2. Name, definition and length of named years at 1995.0

Year type, length length length

Reference point d d h m s s
Eclipse 346.620 074 346 14 52 54.4 29 947 974.4
Node
Tropical 365.242 190 365 05 48 45.2 31 556 925.2
Equinox
Gregorian 365.242 5 365 05 49 12 31 556 952
Gregorian calendar
Julian 365.25 365 06 0 0 31 557 600
Julian calendar
Sidereal 365.256 36 365 06 09 10 31 558 150
Fixed star
Gaussian 365.256 90 365 06 09 56 31 558 196
Kepler’s law
for a = 1
Anomalistic 365.259 64 365 06 13 53 31 558 433
Perihelion

2. Time of the day uses the 24-hour system in the form hh:mm:ss starting with 00 for
the first hour after midnight and ending with 23 as the last before the following
midnight. Hence, 21:56:16 is equivalent to 9 h 56 m 16 s p.m. or 3 m 44 s before
10 p.m.
3. The representation of both date and time as a single composite number uses the
capital letter T to separate the date component from the time component.
60 Unit of time (second)

For example, 2003–09–03 T 05:21:37 is 21 m 37 s after 5 on the morning of 3rd

September 2003.17

5.4 The hertz: unit of frequency

The derived SI unit of frequency is the hertz. The dimension of frequency is [T ]−1 ,
its unit is the hertz and its symbol is Hz. The unit was previously known as the
‘cycle per second’, c . s−1 .

5.5 Angular motion

5.5.1 Angular velocity and acceleration
The SI derived unit of angular velocity has dimension [1] . [T ]−1 , its unit is the
radian per second and its symbol is rad . s−1 .
The angular velocity, ω, may be expressed as the first derivative with respect to
time of the angle θ , thus:
dθ
ω= = θ̇ (5.10)
dt
Angular acceleration, ω̇, is the second derivative with respect to time of the angle θ ,
thus:
d2 θ
ω̇ = 2 = θ̈ (5.11)
dt

The SI derived unit of angular acceleration has a dimension of [1] . [T ]−2 , its unit
is the radian per second per second and its symbol is rad . s−2 .

5.5.2 Rotation period and period of revolution

The difference between rotation and revolution is defined as (Funk et al., 1946)
‘To rotate is said of a body that has a circular motion about its own centre or axis;
to revolve is said of a body that moves about a centre outside of itself.’

Some examples of angular velocities, rotation periods and periods

of revolution for Solar System bodies
For planets and other bodies in the Solar System, the mean orbital motion is often
quoted in units of degrees per day. In SI units, rad . s−1 is used. The conversion
from degrees per day to rad . s−1 is carried out as follows:
π
1◦ . d −1 = rad . s−1 = 2.020 057 005 × 10−7 rad . s−1 (5.12)
180 × 86 400
17 For further information about ISO8601, see the ISO website at: www.iso.org/iso/support/faqs/
 5.5 Angular motion 61

Table 5.3. Rotation and revolution periods and rates for the planets and some
dwarf planets in the Solar System

Mean orbital Rotational

Sidereal period angular velocity Sidereal rotation angular velocity
Planet Julian years nrad .s−1 rate days μrad .s−1
Mercury 0.240 844 826.683 495 58.646 225 1.240 013
Venus 0.615 182 323.647 219 −243.019 99 −0.299 243
Earth 0.999 978 199.106 385 0.997 269 72.921 154
Mars 1.880 711 105.865 348 1.025 957 70.882 181
Jupiter 11.856 525 16.792 621 0.413 538 175.853 234
Saturn 29.423 519 6.766 769 0.444 009 163.784 990
Uranus 83.747 407 2.377 413 −0.718 333 −101.237 196
Neptune 163.723 204 1.216 090 0.671 250 108.338 253
Dwarf planet
Ceres 4.60 43.283 071 0.378 125 192.322 783
Pluto 248.020 8 0.802 764 −6.387 246 −11.385 510
Eris 560.89 0.354 974 ≥ 0.583 ? ≤ 124.666 ?

Table 5.3 gives the sidereal period of revolution in julian years, the sidereal rotation
period in days, the mean orbital angular velocity in nrad . s−1 and the rotational
angular velocity in μrad . s−1 .
The negative values for the mean rotational periods and angular motions of
the planets Venus and Uranus and the dwarf planet Pluto are due to their retro-
grade motions. The sources used for Table 5.3 were: Cox (2000); The Astronomical
Almanac (1995); Duffard et al. (2008) and the JPL Small-Body Database Browser.18

5.5.3 Proper motions

The proper motion of a celestial body may be defined as the change in direction of
that object across the celestial sphere over a period of time, as seen by an observer
located at the Sun.
This total proper motion is the sum of the actual motion of the celestial body (its
peculiar motion) and the parallactic motion, which is the reflex motion of the Sun
about the galactic centre.
In Figure 5.1 the celestial body is initially located at S0 , with equatorial coordi-
nates (α0 , δ0 ) and declination measured from the celestial equator, north positive
and south negative. The body has a proper motion of magnitude μ, and direction
of motion θ, measured in a direction from the north through east. After a time
interval t, measured in julian years, the body moves to a new position S1 with

18 See http://ssd.jpl.nasa.gov/sbdb.cgi
62 Unit of time (second)
S1
NCP
μ
θ
S0
μδ μα cos δ

equato ϒ
r B
α
δSCP

SCP

Figure 5.1. The proper motion, over a period of time t years, of the celestial body
at S0 , with equatorial coordinates (α0 , δ0 ), takes it to position S1 with coordinates
(α1 , δ1 ).

equatorial coordinates (α1 , δ1 ) relative to the Sun. The components of the annual
proper motion in the equatorial coordinate system are (μα , μδ ) where:
α1 − α0
μα =
t
and
δ1 − δ 0
μδ = (5.13)
t
and the magnitude, μ, is given by:

μ = μ2α cos2 δ + μ2δ (5.14)
the factor cos δ being due to the radius of the small circle of declination being
a function of δ. The position angle θ, as shown in Figure 5.1, is related to the
proper-motion components by:
μ cos θ = μδ
μ sin θ = μα cos δ (5.15)

Proper motions in SI units

The SI unit of proper motion has a dimension of [1] . [T ]−1 , its unit is the radian
per second and its symbol is rad . s−1 .
Units in common use include the arcsecond per Julian year (arcsec . y−1 ); the
arcsecond per Julian century (arcsec . (century)−1 ); the milliarcsecond per Julian
 5.5 Angular motion 63

year (mas . y−1 ) and by using radio interferometry, the microarcsecond per Julian
year (μas . y−1 ).

Converting commonly used proper-motion units into SI units

π
1 arcsec . y−1 = rad . s−1
180 × 3600 × 31.5576 × 106
= 1.536 281 85 × 10−13 rad . s−1 (5.16)
1 mas . y−1 = 1.536 281 85 × 10−16 rad . s−1 (5.17)
1 μas . y−1 = 1.536 281 85 × 10−19 rad . s−1 (5.18)

π
1 arcsec . y−1 = rad . y−1
180 × 3 600
= 4.848 136 811 × 10−6 rad . y−1 (5.19)
1 arcsec . (century)−1 = 4.848 136 811 × 10−6 rad . (century)−1 (5.20)

Table 5.4 gives the proper motions of the ten brightest stars in the night sky,
taken from values obtained by the HIPPARCOS astrometric satellite and listed on
the ESO website database.19 The basic SI unit of angular velocity (column 3 in Table
5.4) could be rendered in alternative forms as, e.g., −83.88 frad . s−1 (femtoradians
per second) or −83.88 quadrillionths rad . s−1 or −8.388 d 14 rad . s−1 , which is
equivalent to −8.388 × 10−14 rad . s−1 .

5.5.4 Proper-motion catalogues

The introduction of high-speed measuring engines, astrometric satellites, orbit-
ing observatories, large ground-based telescopes with adaptive optics and smaller
telescopes dedicated to a specific project (not necessarily astrometric) has led to
a veritable explosion of astrometric and photometric data available online. Three
examples of such catalogues with sample entries in both the original data format
and SI format follow.
The HIPPARCOS/TYCHO satellite provided a large body of positional, distance
and proper-motion data without the restrictions placed by the Earth’s atmosphere
on ground-based observations. The entire database was recently revised by van
Leeuwen (2007). The database is restricted to stars whose monochromatic flux
density in the B waveband exceeds approximately 40 mJy (B magnitude ∼12.5
in the catalogue). The median precision of the proper motions is 12 nrad . y−1
(2.5 mas . y−1 ). The database has been of great importance in establishing the

19 See http://archive.eso.org/skycat/servers/ASTROM
64 Unit of time (second)

Table 5.4. Proper motions of the ten brightest stars in the night sky

Star name μα cos δ μδ mas . y−1 μα cos δ μδ rad . s−1 μα cos δ μδ nrad . y−1

Sirius −546.01 −8.388×10−14 −2 647.13

−1 223.08 −1.879×10−13 −5 929.66
Canopus 19.99 3.071×10−15 96.91
23.67 3.636×10−15 114.76
α Cen A −3 678.19 −5.651×10−13 17 832.37
481.84 7.402×10−14 2 336.03
Arcturus −1 093.45 −1.680×10−13 −5 301.20
−1 999.40 −3.072×10−13 −9 693.36
Vega 201.02 3.088×10−14 974.57
287.46 4.416×10−14 1 393.65
Capella 75.52 1.160×10−14 366.13
−427.13 −6.562×10−14 −2 070.78
Rigel 1.87 2.873×10−16 9.07
−0.56 −8.603×10−17 −2.71
Procyon −716.57 −1.101×10−13 −3 474.03
−1 034.58 −1.589×10−13 −5 015.79
Betelgeuse 27.33 4.199×10−15 132.50
10.86 1.668×10−15 52.65
Achernar 88.02 1.352×10−14 426.73
−40.08 −6.157×10−15 −194.31

Table 5.5. HD73904, a member of the galactic cluster IC2391; data from the
revised HIPPARCOS/TYCHO catalogue by van Leeuwen (2007)

α δ μα μδ
Unit HD HIP h m s ◦   mas . y−1 mas . y−1
ID number number mrad mrad nrad . y−1 nrad . y−1
IAU 73904 42374 08 38 23.94 −53 43 18.6 −23.49 21.86
SI 2261.942 −937.623 −113.88 105.98

International Celestial Reference Frame (McCarthy & Petit, 2004). Examples of

converting the catalogue proper motions in mas . y−1 to SI units are given in
Tables 5.4. and 5.5.
In Table 5.5, the unit ID ‘IAU’ gives the positions and proper motions in what are
the most commonly used and accepted forms at present. It should be noted that in
the van Leeuwen (2007) revision of the HIPPARCOS/TYCHO catalogue both right
 5.5 Angular motion 65

Table 5.6. A partial entry for the star id13222 from the OGLE-II proper-motion
catalogue by Sumi et al. (2004)

Unit α δ μα μδ μl μb

◦ ◦ mas y−1 mas y−1 mas y−1 mas y−1

IAU 271.010 58 −29.241 83 −3.98 −6.50 −7.62 0.31
mrad mrad nrad . y−1 nrad . y−1 nrad . y−1 nrad . y−1
SI 4730.027 −510.366 −19.30 −31.51 −36.94 1.50

ascension and declination are presented in radians, though not the proper-motion
components.
On a much larger scale is the USNO B1.0 all sky catalogue of magnitudes,
colours, positions and proper motions for more than one billion celestial objects
to a limiting monochromatic flux density of 14.3 μJy in the V waveband (V mag-
nitude ∼21).20 This catalogue (Monet et al., 2003) was produced from automated
measuring-machine scans of photographic plates taken for the POSS (Palomar
Observatory Sky Survey) and SERC-I, the UK Science and Engineering Research
Council’s first Southern Sky Survey. The proper motions are given in units of
arcsec . y−1 with a precision of 0.002 arcsec . y−1 .
By way of comparison, the equatorial coordinates of HD73904 given in the
USNO-B catalogue are (08h 38m 23s .91; −53◦ 43 18 .39) and its proper-motion
coordinates (−26, 22) mas . y−1 , which are the equivalent in SI units of (2261.940,
−937.622) mrad and (−126.05, 106.66) nrad . y−1 .
An example of a proper-motion catalogue derived from sequential observa-
tions over a period of three years, obtained primarily to search for stellar and
exoplanetary microlensing, is that published by Sumi et al. (2004). Their cata-
logue, covering 49 selected fields, totalling approximately 0.003 4 sr (11 square
degrees) of the galactic bulge surrounding the galactic centre, lists over 5 million
stars with I-band monochromatic flux densities between ∼100 mJy and ∼150 μJy
(11 < I < 18). Each entry in the catalogue includes an internal identity number, mean
proper motion components in mas . y−1 for both equatorial and galactic coordinates,
with standard deviations about the mean values, right ascension and declination
(J2000.0) in decimal degrees, V magnitudes and (V − I) colours.
A partial entry from the proper-motion catalogue with angular measures as
published and as converted to SI units is given in Table 5.6.

20 See http://www.nofs.navy.mil/data/fchpix/
66 Unit of time (second)

5.6 The determination of the ages of celestial bodies

There are many different ways of determining the ages of celestial bodies, both
in the universe and of the universe itself. In this chapter, only those methods that
depend on the direct measurement of time or angular velocity will be considered,
with examples of the use of radioactive decay (nucleocosmochronology) and the
spin-down rate of pulsars.

5.6.1 Nucleocosmochronology
Nucleocosmochronology is defined by Schramm (1990) as ‘. . . the use of the abun-
dance and production ratios of radioactive nucleides coupled with information
on the chemical evolution of the Galaxy to obtain information about the time
scales over which the solar system elements were formed’. The basic method for
determining ages using this method was first set out by Rutherford (1929).
It is well known that some atomic nuclei are stable and some are not. Those that
are not exist for a certain average length of time and then eject particles sponta-
neously (radioactivity). The new nucleus so formed, known as the daughter nucleus,
may or may not be stable. For each radioactive nucleus there exists a transition prob-
ability λ, such that the nucleus will decay into its daughter nucleus within the next
small interval of time dt. The inverse of this transformation probability is the aver-
age lifetime of the parent nucleus and is measured in seconds (base SI unit) or
Julian years. λ is measured in bequerels (a derived SI unit of dimension [T ]−1 and
symbol Bq), equal to the number of radioactive decays per second. A non-SI unit
that is sometimes used to express the activity of a radioactive sample is the curie,
equivalent to the mean rate of decay of 1 g of radium (∼3.7 × 1010 Bq).
Radioactivity has no ‘memory’, so it is not possible to predict when an individual
nucleus will decay, but it is possible to make predictions about the average behaviour
of large numbers of nuclei.
If a large number, N, of radioactive nuclei of a particular element exists at time t,
then the number, dN, that will decay in the next small time interval dt, is given by:
dN = − N λ dt (5.21)
or
dN
= − λ dt
N
Integrating with respect to t from t0 to (t + t0 ) yields the law of exponential decay:
N(t) = N(t0 ) e−λ t (5.22)
Figure 5.2 illustrates the shape of a typical radioactive decay curve.
 5.6 The determination of the ages of celestial bodies 67

1.0 N(t0)

0.8

0.6 N(t1 )
2
N(t)

0.5
0.4

0.2

0.0
0.139

0.05 0.30 0.55 0.80 1.05 1.30 1.55 1.80

t

Figure 5.2. The radioactive decay of a large number of parent atomic nuclei into
daughter atomic nuclei follows an exponentially shaped curve with time.

The half-life, t 1 , of a radioactive nucleus is the time it takes, on average, for half
2
of the original number of nuclei to decay into daughter nuclei:
ln 2
t1 = (5.23)
2 λ
The mean lifetime of a state τ is defined by:
1
τ= (5.24)
λ
After a time interval τ has elapsed, on average, approximately (1 − 1/e)% of
the original nuclei have decayed into daughter nuclei. The half-life and the mean
lifetime of the state are related by:

t 1 = τ ln 2 (5.25)
2

If at time t = 0, the number of parent nuclei is Np0 and the number of daughter
nuclei Nd0 then, at time t, the number of daughter nuclei is:
Ndt
= eλ t − 1 (5.26)
Npt
From this equation, if λ is known and the ratio Ndt : Npt can be measured observa-
tionally or experimentally, then a value can be determined for the age of the sample
being investigated.
68 Unit of time (second)

5.6.2 Pulsars
Pulsars, named by contraction from pulsating stars, were discovered in 1967 by
Bell and Hewish using a newly designed and built radio telescope in Cambridge
(Thorsett, 2001). Radio signals from pulsars take the general form of a series of
extremely regular short bursts of radiation. It was shown that pulsars are neutron
stars, a late stage of evolution of stars whose masses, from model stellar computa-
tions, are expected to lie between 0.2 and 2 solar masses (4 × 1029 − 4 × 1030 kg).
Pulsars have high rotational inertia that leads to an extremely stable rotation. The
individual pulses are stable to around 1 part in 1014 , which is not far removed from
the current stability of caesium fountain atomic clocks at 5 parts in 1016 . Compar-
ison between the pulse arrival time and the reference atomic standard time can be
used to determine very precise orbital parameters of binary pulsars. Observations
of the pulsation period and the rate of change of the period can also provide data on
the magnetic field strength and age of the pulsar (Burke & Graham-Smith, 2002).
The slowing of the rotation period provides the energy for the pulses.
If the original period of pulsation at the time the pulsar was formed is P0 , the
present pulsation period is P , and the rate of increase in the period is Ṗ , then the
current age of the pulsar, t, is:

n − 1 
P P0
t= 1− (5.27)
(n − 1) Ṗ P

If it is assumed that P P0 and n = 3 (for braking by magnetic spin-down radiation)

then the characteristic age, τ , is defined as:
P
τ= (5.28)
2 Ṗ
An actual value for n can only be directly determined if a value of the second
derivative of P with respect to time can be measured.
The mean monthly signal-frequency data for the Crab pulsar used to plot
Figure 5.3 covering the period from 1982 February 15 to 2002 September 15 was
downloaded from the NVO (National Virtual Observatory of the USA) website.21
The CRABTIME database was sourced from Lyne and collaborators at the Jodrell
Bank Observatory,22 with the addition of values for P and Ṗ (period of pulsation
and its derivative with respect to time) calculated by HEASARC (the High Energy
Astrophysics Science Archive Research Centre of NASA).23

21 See http://nvo.stsci.edu/vor10/getRecord.aspx?id=ivo://nasa.heasarc/crabtime
22 http://www.jb.man.ac.uk/pulsar/crab.html (provided by A.G. Lyne, C.A. Jordan & M.E. Roberts in Jodrell Bank
Crab Pulsar Timing Results, Monthly Ephemeris).
23 See http://heasarc.gsfc.nasa.gov/
 5.6 The determination of the ages of celestial bodies 69

30.05

30.00

29.95
ν Hz

29.90

29.85

29.80

46 000 48 000 50 000 52 000

mjd

Figure 5.3. Crab pulsar rotation-rate slowdown from 1982 to 2002. The abscissa
is given as modified julian day numbers and the ordinate as frequency in hertz.

Over the time covered in this data set the mean pulsation period, P, was
0.033 271 ± 0.002 106 s and the mean rate of change of the pulsation period, Ṗ ,
was 4.192 877 × 10−13 ± 0.265 206 × 10−13 s . s−1 . Substituting these values into
Equation (5.28) and using an observational value of n = 2.51 ± 0.01 for the braking
index determined by Lyne et al. (1993), the age of the Crab pulsar is found to be
1665 y. This may be compared with the known age of 956 y at the present time
(2010 CE). Burke & Graham-Smith (2002) warn that, at best, ‘present day slow-
down rates are only indications of actual age and not infallible measures of it’. One
possible cause of this discrepancy, the effect of a large transverse motion on the
observed period derivatives, was first pointed out by Shklovskii (1970). Applying
a suitable correction for this (see Manchester et al., 2005) produces a better value
for the pulsar slow-down rate.
A catalogue of pulsar data compiled by Taylor et al. (1993) forms the basis of
an online database, the ATNF Pulsar Catalog.24 Some 1640 pulsars in the database
have had their characteristic ages computed (using a braking index of 3), which

24 See http://www.atnf.csiro.au/research/pulsar/psrcat/
70 Unit of time (second)

0.3

0.2

0.1

0.0
2 3 4 5 6 7 8 9 10
log (age)

Figure 5.4. Histogram/frequency plot of the distribution of the logarithm of the

characteristic ages of pulsars in Julian years listed in the ATFN Pulsar Catalog.

vary from the youngest at 218 y (PSR J1808-2024) to the oldest at 6.75 × 1010 y
(PSR J0514-4002A)(see warning above!). An histogram of the logarithms of these
pulsar ages is shown in Figure 5.4. The median age from the database is 5.47 × 106 y.

5.7 Summary and recommendations

5.7.1 Summary
Time was initially measured with reference to the rotation of the Earth and its period
of revolution about the Sun (dynamical time). More recently, the development of
very stable and accurate atomic clocks has led to the introduction of atomic time.
The relationships between these two major time systems is given. Rotation periods
and periods of revolution may be measured in derived SI units of rad . s−1 , some
examples of which are given. A similar unit is used for the proper motions of the
stars across the celestial sphere.
It is possible to determine the ages of some astronomical bodies by using the
abundance and production ratios of radioactive isotopes. Decay rates from parent
to daughter populations may be measured in bequerels, a unit of inverse time. The
 5.7 Summary and recommendations 71

ages of pulsars may be found from a detailed study of the rate at which their rotation
rate slows; an example of such a study is shown for the Crab pulsar.

5.7.2 Recommendations
The IAU recommends using the TAI system when precise measures of time intervals
are required.
The commonly used unit for angular rotation, degrees per day, should be replaced
by the SI unit, rad . s−1 . Similarly, stellar proper motion should use the same unit,
though, given that its numerical value will be very small, rad . y−1 or submultiples
thereof (e.g., mrad . y−1 , μrad . y−1 or nrad . y−1 ) could be used. For very distant
or slow moving bodies it is usual, at present, to quote the proper motion as angular
measure per julian century.
The inverse unit of time can be either the hertz, when used as a derived SI unit
of frequency, or the bequerel, when used as the number of radioactive decays per
second. In the case of the latter, since the age of the celestial body is the unknown
being sought, then the unit of interest becomes the inverse of the bequerel, which
is the SI unit, the second.
 6
Unit of length (metre)

6.1 SI definition of the metre

The metre is the length of the path travelled by light in vacuum during a time
interval of 1/299 792 458 of a second.
The dimension of length is [L], its unit is the metre and its symbol m.

The original definition of the metre was one ten millionth of the distance from
the Earth’s north pole to its equator, determined along a meridian arc that ran
from Dunkirk in the north to Barcelona in the south. Observations were begun in
1792 by J. B. J. Delambre, who worked from Paris northwards and P. F. A. Méchain
who made measurements from Paris to Barcelona. They completed their task in
seven years and the metre thus determined was modelled in pure platinum as a
one-metre-long bar (Alder, 2004).

6.2 Linear astronomical distances and diameters

The sizes of and the distances between astronomical bodies is generally extremely
large by everyday terrestrial standards. This has resulted in astronomers inventing
units such as the light year, the astronomical unit and the parsec, which are, at first
sight, better able to deal with very large distances. The SI unit of length, the metre,
used in conjunction with common prefixes is normally only used for measurements
within the Solar System.

6.2.1 Size of the Earth

Were the Earth a perfect sphere it would follow from the original definition of the
metre that its diameter would be 4×107 /π m. Unfortunately, however, the rotation
of the Earth causes a flattening of the perfect sphere into an oblate spheroid, with
the equatorial radius a being greater than the polar radius c. The amount of this

72
 6.2 Linear astronomical distances and diameters 73

flattening, f , or ellipticity (= 1/f ), is given in McCarthy & Petit (2004)25 as:

(a − c) 1
f= = = 0.003 352 819 70 (6.1)
a 298.256 42
where a = 6.378 136 60 × 106 m (±0.10 m) and c = 6.356 751 86 × 106 m
(McCarthy & Petit, 2004).
A uniform reference spheroid, known as the geoid, is the imaginary surface
that the Earth would have at mean sea level if an ellipse of semimajor axis a,
and eccentricity e, were rotated about its minor axis
(the polar axis in the case of
the Earth). The eccentricity is derived from e = (2f − f 2 ), and the radius r, at
geodetic latitude φ,26 from:

r = a (1 − f . sin2 φ) (6.2)

Example: derive the geocentric radius at the location of the Paris

Observatory
The geodetic latitude of the Paris Observatory is (see The Astronomical Almanac,
1995).

φ = 852.360 629 mrad (≡ 48◦ 50 .2)

where φ is measured northwards from the geodetic equator. Now substitute for φ,
and for a and f (from McCarthy & Petit, 2004) in Equation (6.2) to determine the
radius at mean sea level, rmsl :

rmsl = 6.366 016 5 × 106 m

To this value add the height of the Paris Observatory above mean sea level (= 67 m)
to produce a final result for the geocentric radius at the location of the Paris
Observatory of:

rParis Observatory = 6.366 083 5 × 106 m

The mean radius of the Earth

To obtain the mean radius, ρ, of the oblate spheroid, consider the volume of a sphere
of radius ρ with the same volume as the oblate spheroid so that:
4 4
π ρ 3 = π (a 2 c) (6.3)
3 3

25 McCarthy & Petit cite the 1999 report of E. Groten to the IAG Special Commission SC3, Fundamental
Constants, XXII IAG General Assembly.
26 Woolard & Clemence (1966) define geodetic latitude as ‘the angle between the geodetic vertical and the plane
of the geodetic equator’.
74 Unit of length (metre)

or
1
ρ = (a 2 c) 3 (6.4)
Substituting the above values for a and c in Equation (6.4) produces a value for
the mean radius of the geoid of ρ = 6.371 000 37×106 m.

6.2.2 Distance to the Moon

The distance from the centre of the Earth to the Moon (see Figure 6.1) was ini-
tially determined, via trigonometry, from measurements of selected stars and the
Moon, made at two observatories, one north and one south of the equator, which
lay at approximately the same terrestrial longitude (e.g., the Royal Observatory,
Greenwich, in London and the South African Astronomical Observatory in Cape
Town).27 The baseline, the straight line distance (i.e., through the Earth, not along
the surface) between the two observatories, is computed knowing the latitude, lon-
gitude and height above mean sea level of each (now readily found from Global
Positioning System (GPS) Earth satellite data).
This method has been superseded by using Earth-based laser ranging techniques.
In the 1960s and 1970s, the Apollo astronauts placed optical retro-reflectors on the
surface of the Moon. By transmitting short pulses of coherent light from a telescope
on Earth to the Moon and recording the time of travel, t, of the beam from the Earth
to the Moon and back, the distance, r, from the observer to the retro-reflector on
the lunar surface is simply r = c t/2.
Given that the value of c is defined and that, by using atomic clocks, the value
of t may be measured to better than 1 part in 1015 , this way of finding the distance
to the Moon is far more accurate than the earlier trigonometric method. Recent
measurements at the Apache Point Lunar Laser-ranging Operation have a median
nightly range uncertainty of ±1.8 mm (Battat et al., 2009).
The mean distance from the centre of the Earth to the centre of the Moon is
3.844 01×108 m ± 1000 m (Cox, 2000).

π M
C

Figure 6.1. Horizontal parallax: point C is the centre of the Earth, angle MCT is a
right angle, CT is the equatorial radius of the Earth, CM is the geocentric distance
of the Moon (or Sun; for solar parallax, see Section 6.2.4) and π is the parallax
angle.

27 For details of the method, see Barlow & Bryan (1956).

 6.2 Linear astronomical distances and diameters 75

The horizontal parallax of the Moon πMoon , is the angle subtended by the equato-
rial radius of the Earth at the distance of the Moon. The mean value of the equatorial
horizontal parallax is 16.593 271 8 mrad ( = 3422  . 608).

6.2.3 Astronomical unit

The definition of the astronomical unit in most textbooks on astronomy and astro-
physics (e.g., Kutner, 2003; Burke & Graham-Smith, 2002; Lang, 2006 and even
Cox, 2000) is generally something like: ‘The astronomical unit is the mean distance
from the Earth to the Sun and is equal to approximately 150 million kilometers or
about 93 million miles.’
The conversion of the astronomical unit into SI units involves direct measures
of the distances of planets and other Solar System bodies, which in the past relied
on classical astrometric methods, such as trigonometric parallax, radial velocities
and annual aberration with uncertainties of about 1 part in 104 . The value of the
gravitational constant, G (= k 2 ) (see Equation 6.8 below), is also only known to
about 1 part in 105 . These have led to computations in celestial mechanics being
carried out using the IAU set of astronomical units: the day as the unit of time;
the mass of the Sun as the unit of mass; and the astronomical unit as the unit
of length. This has allowed a far greater accuracy in the preparation of planetary
ephemerides. However, this situation is changing with the far higher accuracy of
Earth-based radar-determined planetary distances aided by atomic clocks and the
use of space probes (see, e.g., Muhleman et al., 1962, and Pitjeva, 2005).

Definition of the astronomical unit, version 1, Kepler’s third law

In 1618, Kepler published what became known as his third law of planetary motion,
which states that the squares of the sidereal periods of revolution of any two planets
about the Sun are proportional to the cubes of their mean distances from the Sun, i.e.,

P12 a13
2
= (6.5)
P2 a23

If one of the planets is the Earth, with a period of revolution about the Sun of one
sidereal year and a mean solar distance of one astronomical unit, then Equation
(6.5) may be rewritten as:
P2
=1 (6.6)
a3
where P is the period of revolution about the Sun of a gravitationally bound astro-
nomical body other than the Earth and a is its mean distance from the Sun in
astronomical units.
76 Unit of length (metre)

For any other planet with sidereal period P years Equation (6.5) may be
rewritten as:
2
a =P 3 (6.7)
Newton expressed Kepler’s third law in a more general form, which involved
the constant of gravitation k, the mass of the planet m, and the mass of the Sun M :
 k 2 P 2 (M + m)  1
 3
a= 2
(6.8)
4π
or when expressed in IAU astronomical units (see below, version 3):
 k 2 P 2 (1 + m)  1
3
a= (6.9)
4π2

Definition of the astronomical unit, version 2, IAU(1976)

In 1976, the IAU28 redefined the astronomical unit to be equal to the distance from
the centre of the Sun at which a particle of negligible mass, in an unperturbed
circular orbit, would have an orbital period of 365.256 898 3 days where 1 day is
defined to be 86 400 SI seconds.

Definition of the astronomical unit, version 3, AAO(2010)

The Astronomical Almanac Online29 provides the following formal definition of
the astronomical unit of length:
The astronomical unit of length is that length (A) for which the Gaussian gravitational
constant (k) takes the value 0.017 202 098 95 when the units of measurement are the astro-
nomical units of length, mass and time. The dimensions of k 2 are those of the constant of
gravitation (G).

That is, A3 . S −1 . D −2 , (dimIAU ([A]3 . [S]−1 . [D]−2 ) ≡ dimSI ([L]3 . [M]−1 .

[T ]−2 )), where S is the astronomical unit of mass equal to the mass of the Sun
and D is the astronomical unit of time equal to one day of 86 400 SI seconds.

Definition of the astronomical unit, future versions?

All the versions of the definition of the astronomical unit given above are based on
Newtonian dynamics. So an obvious development in redefining the astronomical
unit is to use the framework of general relativity. Examples of this approach are
given by, e.g., Huang et al. (1995), Klioner (2007) and in a summary by Capitaine
& Guinot (2008).

28 See http://www.iau.org/public-press/themes/measuring/
29 See http://asa.usno.mil/index.html
 6.2 Linear astronomical distances and diameters 77

Also, given the precision with which the astronomical unit can now be defined,
the assumption that the astronomical unit of mass (one solar mass) is a constant is
no longer acceptable. It is known that the Sun converts mass into energy resulting
in a loss of mass equal to 4.3×109 kg . s−1 , as well as further particulate mass loss
through the solar wind of approximately 1.3×109 kg . s−1 . A total of some 5.5 mil-
lion tonnes per second, or, in IAU astronomical unit terms: 2.405×10−16 M . d−1 ,
or in units of solar masses per century: 8.797×10−12 M . century−1 . According
to Noerdlinger (2008), such a loss will result in the orbits of the planets expanding
at the same relative rate and their periods of revolution to increase at twice that
relative rate, causing the planet Mercury to drift away from its predicted position
by more than 5 km in an interval of 200 years.
Another way in which the astronomical unit may be increased is via the total
conservation of angular momentum law, proposed by Miura et al. (2009).
It is apparent that further revision of the definition of the astronomical unit
of length is not only desirable but necessary. However, given that such revisions
may take many years to carry out, the only presently available solution to convert
astronomical units to metres is to use the currently accepted best conversion value.

Recent determinations of the astronomical unit

By combining more than 300 000 positional observations, both optical (transit
circle, photographic and CCD) and radio, of the planets and spacecraft made
from 1913 to 2003, Pitjeva (2005) determined the astronomical unit to be:
1.495 978 706 960×1011 m ± 0.1 m.
This value was updated by Pitjeva & Standish (2009) using improved ephemeris
computational methods at JPL in Pasadena and at IAA RAS in St Petersburg to:
1.495 978 707 00×1011 m ± 0.3 m.
McCarthy & Petit (2004) in the IERS Conventions (2003) list a scale fac-
tor (km/au) of 149 597 870.691 as an auxiliary constant from the JPL Planetary
and Lunar Ephemerides (DE405/LE405). In metres, this conversion factor is:
1 au = 1.495 978 706 91×1011 m (TDB) or 1 au = 1.495 978 714 64 × 1011 m (SI).
This is the scaling factor that will be used in subsequent computations.
Table 6.1 gives the distances to all the planets and some of the dwarf planets in
both astronomical units and the SI unit of length, the metre. As can be seen from
columns 3 and 4 in Table 6.1, distances within the Solar System may be readily
expressed in either IAU astronomical units or prefixed values of the SI metre.

6.2.4 Solar parallax

The Sun’s equatorial horizontal parallax, π , was defined as the angle at the cen-
tre of the Sun subtended by the Earth’s equatorial radius, a⊕ , at a distance of 1
78 Unit of length (metre)

Table 6.1. Sidereal period in days and distances in IAU astronomical units and SI
metres for the planets and some dwarf planets in the Solar System

Sidereal period Semimajor axis Semimajor axis

Planet IAU d astronomical units ×1012 m = Tm
Mercury 87.968 0.387 098 93 0.057 909 17
Venus 224.695 0.723 331 99 0.108 208 92
Earth 365.242 1.000 000 11 0.149 597 88
Mars 686.930 1.523 662 31 0.279 366 38
Jupiter 4330.596 5.203 363 01 0.778 412 03
Saturn 10 746.940 9.537 070 32 1.426 725 42
Uranus 30 588.740 19.191 263 93 2.870 972 23
Neptune 59 799.900 30.068 963 48 4.498 252 93
Dwarf Planet
Ceres 1 680.150 2.767 000 00 0.413 937 31
Pluto 90 656.073 39.481 686 77 5.906 376 30
Eris 204 865.073 67.959 000 00 10.166 521 70

astronomical unit, when the astronomical unit itself was defined to be the mean
Earth–Sun distance:
a⊕
π = (6.10)
1 au

From Section 6.2.1, the equatorial radius of the Earth is 6.378 136 60×106 m and
from the previous section the value of the astronomical unit in metres is given, so
substituting these numerical values into Equation (6.10) gives the solar parallax
to be: 42.635 209 56 μrad (equivalent to 8 .794 143 240). As the Earth’s orbit is
elliptical, the value of the solar parallax varies from approximately 43.342 μrad
(8 .94) at perihelion on January 4, when the distance to the Sun is about 0.983 au,
to 41.936 μrad (8 .65) at aphelion on July 4, when the distance to the Sun is about
1.017 au.

6.2.5 Horizontal parallax via gravitational microlensing

Generally, the horizontal parallax of celestial bodies beyond the Solar System is
so small that it would not be measurable from the Earth. However, Gould et al.
(2009), from observations made by several of the worldwide microlensing groups,
have been able to measure the effects of terrestrial parallax. The object observed,
OGLE-2007-BLG-224, was shown to be a very low mass thick-disc brown dwarf at
a distance of 525 ± 40 pc. The equivalent parallax, πorbit , in SI units is 9.235 nrad.
The angle π⊕ , subtended by the semimajor axis, a⊕ , of the Earth at the distance of
 6.2 Linear astronomical distances and diameters 79
Σ

πΣ

E
S

Figure 6.2. Trigonometric parallax: point S is the Sun, SE is the mean radius of
the Earth’s orbit about the Sun, S is the heliocentric distance of the celestial body
 and π is the parallax angle of . NB: this figure is NOT to scale since the
distance S SE.

the microlensing brown dwarf is approximately given by:

a⊕
π⊕ = πorbit . = 3.937 168 × 10−13 rad (6.11)
1 au
or 394 frad (femtoradians).

6.2.6 Trigonometric parallax

The trigonometric or heliocentric parallax of a celestial body is the angle subtended
by the radius of the Earth’s orbit about the Sun at the distance of the celestial body
(see Figure 6.2).
Given that the relative size of the measurement errors in determining the parallax
angle is large in comparison with that of determining the Earth–Sun distance, it may
be assumed that the Earth–Sun distance is a good approximation to 1 astronomical
unit. Weygand et al. (1999) suggest, taking into account the gravitational influences
of the major planets, that the Earth’s orbit may best be described as an ellipse, with
semimajor axis equal to 1.000 000 2 astronomical units. So the difference between
the measured value of the Earth–Sun distance and that determined from version
3 of the astronomical unit definition amounts to 2 parts in 107 . Perryman et al.
(1997) give the median value of the standard deviation of the mean parallax for all
stars measured with a monochromatic flux density greater than approximately 1 Jy
(Hp ≤ 9) as ±4.7 nrad (≡ 0.97 mas). The entry in the HIPPARCOS catalogue30 for
the bright nearby star α Cen A is π = 3 597.9 ± 6.8 nrad (≡ 742.12 ± 1.40 mas).
The standard deviation of the measurements of the parallax to the value of the

30 See http://archive.eso.org/skycat/servers/ASTROM
80 Unit of length (metre)

parallax is the equivalent of approximately 1 part in 530, or some 20 000 times

greater than the differences due to the differing definitions of the astronomical unit.

Example: determine the distance to α Cen A in metres

Given the mean parallax π of α Cen A as 3 597.9 nrad, then its distance, D, in
astronomical units, is simply:
A 1
D= = = 277 940 (6.12)
π 3 597.9 × 10−9
Using the standard deviation of the parallax distance range limits, Dmax and Dmin ,
for the star may be calculated, where:
1
Dmax =
π − σπ
1
Dmin = (6.13)
π + σπ
Substituting the catalogue values for the mean parallax and its standard deviation,
the distance in astronomical units to α Cen A may be given as: 277 940 278 465
277 416 . To
convert these distances into metres, the SI unit of length, simply multiply by the
final conversion factor given in Section 6.2.3, which gives: 4.158×1016 m as the
mean distance within the limits 4.150×1016 m to 4.166×1016 m. The mean distance
may also be expressed as 41.579 Pm (petametres) or 4.158 d 16 m.
In Table 6.2, parallaxes in mas and nrad, and distances in astronomical units, SI
metres, parsecs and light years derived from the SIMBAD31 database are given.
For this group of stars, which relative to the size of the Milky Way galaxy may be
considered as very close, the distance measurements in astronomical units become
unwieldy, and the name of the unit does not lend itself to the addition of prefixes.
Astronomers have accordingly invented two larger measurements of distance: the
parsec and the light year.

6.2.7 Parsec
The parsec is the distance, D, at which one astronomical unit subtends an angle of
one arc second. It is the reciprocal of the parallax π, expressed in arcseconds of a
celestial body, i.e.:
1
D= (6.14)
π

31 See http://simbad.u-strasbg.fr/simbad/sim-fid, which refers to Perryman et al. (1997).

 6.2 Linear astronomical distances and diameters 81

Table 6.2. Parallaxes and distances in astronomical units, metres, parsecs and
light years for the ten brightest stars in the night sky

Parallax Distance
πmas astronomical Distance (SI) Distance parsecs
Star name πnrad units metres ×1015 light years

Sirius 379.21 5.439×105 81.371 2.637

1838.462 8.601
Canopus 10.43 1.978×107 2958.464 95.877
50.566 312.710
α Cen A 742.24 2.779×105 41.572 1.347
3598.481 4.394
Arcturus 88.85 2.321×106 347.291 11.255
430.757 36.709
Vega 128.93 1.600×106 239.330 7.756
625.070 25.297
Capella 77.29 2.669×106 399.234 12.938
374.712 42.199
Rigel 4.22 4.888×107 7312.032 236.967
20.459 772.882
Procyon 285.93 7.214×105 107.917 3.497
1386.228 11.407
Betelgeuse 7.63 2.703×107 4044.138 131.062
36.991 427.466
Achernar 22.68 9.095×106 1360.528 44.092
109.956 143.808

If the parallax is expressed in radians πrad , then Equation (6.14) may be

rewritten as:
1
D= (6.15)
206 264.806 265 × πrad
The dimension of the parsec, being the reciprocal of an angular measure, is
[L] . [L]−1 and its symbol is pc.

Example: determine the distance to the star Achernar

In Table 6.2, Achernar is listed as having a parallax of 109.956 nrad, hence from
Equation (6.15):
1
= 44.092 pc
109.956 × 10−9 × 206 264 . 806 265
For stars in the solar neighbourhood the parsec is a useful unit. When consid-
ering the Galaxy as a whole, the kiloparsec equal to 1000 pc is preferable. For
82 Unit of length (metre)

extragalactic distances, the megaparsec, equal to 1 000 000 pc, is commonly used
with, for very large distances, the gigaparsec, equal to 1 000 000 000 pc.
In SI units, the parsec is equal to, 3.085 677 6×1016 m, and in IAU units is equal
to 206 264.806 265 au.
The parsec is calculated directly from a measurement of the parallax of an
astronomical object.

6.2.8 Light year

The light year may be considered to be a derived, defined unit equal to the
distance that light, travelling at 299 792 458 m s−1 , covers in 1 julian year
of exactly 3.15576×107 (SI) s. Hence, 1 light year is exactly equivalent to
9.460 730 472 580 8×1015 m.
Its dimension is [L] . [T ]−1 . [T ] = [L] and its symbol is ly.
To convert parsecs into light years, multiply by 3.261 563 8, and to convert
astronomical units into light years, divide by 63 241.077 101 522.
The distance in light years is not directly measurable for even the nearest star,
though such measurements may become possible in the far distant future via radar
or laser reflections from bodies orbiting nearby stars, or from signals received from
probes sent to such bodies.
At present, the parsec is the unit of choice for professional astronomers and
the writers of university texts, whilst the light year is the preference for authors
of popular astronomical articles and books. Note that the light year is not even a
non-SI unit recognized for use in astronomy, neither is it a unit recommended by
the IAU (see Wilkins, 1989). A brief history of the development of the parsec and
the light year is given by Beech (2008), in which he points out that the first indirect
reference (Bessel, 1838) to the term ‘light year’ predates the published appearance
of the parsec (Dyson, 1913) by some 75 years.

6.2.9 Some examples of astronomical distances

In Table 6.3, the distance in metres is given both in decimal notation (see
Section 2.5.5) and using SI prefixes (Em = Exametres, Zm = Zettametres and
Ym = Yottametres). The distances in parsecs listed in the table were extracted from
WEBDA,32 the NASA/IPAC Extragalactic Database,33 Cox (2000), The Astronom-
ical Almanac (1995) and van Leeuwen (2007). Note that the distances given for
M104, the Coma cluster of galaxies and the A3266 cluster are galactocentric rather
than heliocentric. A number density plot of the cluster A3266 is given in Figure 4.10.

32 http://www.univie.ac.at/webda/
33 http://nedwww.ipac.caltech.edu/
 6.3 Linear motion 83

Table 6.3. Some examples of distances given in metres, parsecs and light years to
astronomical bodies lying outside the Solar System

Distance (SI) Distance parsecs, Distance light

Object name Object type metres kpc or Mpc years
Hyades galactic 1.389 d 18 45 pc 147
star cluster 1.389 Em
Pleiades galactic 3.765 d 18 122 pc 398
star cluster 3.756 Em
IC2391 galactic 4.536 d 18 147 pc 479
star cluster 4.536 Em
M67 galactic 2.802 d 19 908 pc 2961
star cluster 28.02 Em
NGC6572 planetary 2.006 d 19 650 pc 2120
(Ring Nebula) nebula 20.06 Em
47 Tuc globular 1.419 d 20 4.6 kpc 15 003
star cluster 141.9 Em 1.500 d 4
M3 globular 3.086 d 20 10 kpc 32 616
star cluster 308.6 Em 3.262 d 4
LMC local group 1.697 d 21 55 kpc 179 386
galaxy 1.697 Zm 1.793 d 5
M31 local group 2.237 d 22 725 kpc 2 364 634
galaxy 22.37 Zm 2.364 d 6
M104 edge on 6.171 d 23 20 Mpc 65 231 276
(Sombrero Hat) galaxy 617.1 Zm 6.523 d 7
Coma cluster of 2.934 d 24 95 Mpc 310 174 717
galaxies 2.934 Ym 3.102 d 8
A3266 cluster of 7.387 d 24 239 Mpc 780 818 374
(Abell cluster) galaxies 7.387 Ym 7.808 d 8

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .

The simple prefixes work well for the parsec unit and are easy to remember, and
those for the metre also work well, though are less easy to remember and may be
difficult to associate immediately with the relevant power of 10. The d notation does
not have this problem. Prefixes are not commonly used with light years, but with
numbers of light years not exceeding tens of billions this is not an insurmountable
difficulty (e.g., the Coma cluster of galaxies would be said to be at a galactocentric
distance of approximately 310 million light years).

6.3 Linear motion

The linear velocity of a moving body is its rate of change in position with time in
a particular direction.
The dimension of the SI derived unit of linear velocity is [L] . [T ]−1 its name is
the metre per second and its symbol is m . s−1 .
84 Unit of length (metre)

6.3.1 The speed of electromagnetic radiation

In 1975, the 17th CPGM recommended that the speed of electromagnetic radiation
in vacuo, c, be a physical constant, defined to be exactly 299 792 458 m . s−1 .
A fixed speed of electromagnetic radiation not only relates distance in metres
and time in SI seconds but also frequency ν, in hertz, and wavelength λ, in metres,
via:
dim(c) = [L] . [T ]−1 (6.16)
and
c=λ.ν (6.17)
Since it is possible to measure time more accurately than any other SI unit, the above
relationships lead to a greatly improved realization of the metre that is limited solely
by the accuracy of the time reference.
The value selected for the defined c is very close to that determined by Even-
son et al. (1972) at the US National Institute of Standards and Technology
(NIST) in Boulder, Colorado, in 1972. The experimentally measured velocity was
299 792 456.2 ± 1.1 m . s−1 .
Figure 6.3 illustrates the relationship between frequency and wavelength
throughout the electromagnetic spectrum, from short wavelength, high-energy
and high-frequency γ -rays to long wavelength, low-energy and low-frequency
microwave and radio waves.

20 γ - rays

X - rays
log(frequency Hz)

16 ultraviolet

visible
infrared

12

radio frequencies
ν=c/λ
8

–13 –11 –9 –7 –5 –3 –1 1
log(wavelength m)

Figure 6.3. A plot showing the relationship between log(frequency) measured in

Hz and log(wavelength) measured in m. Named portions of the electromagnetic
spectrum are shown.
 6.3 Linear motion 85

6.3.2 Radial velocity

The radial velocity of an astronomical body is the component of its motion along
the line of sight from the observer to the body. The radial velocity component is at
right angles to the transverse component across the celestial sphere.
It is determined by measuring the shift of spectral lines in the electromagnetic
radiation emitted by the celestial body when compared with a laboratory set of spec-
tral lines, which are at rest relative the observer. If the wavelength of an identified
spectral feature in the spectrum of the astronomical body is λ, and the wavelength
measured in the laboratory is λ0 , then the radial velocity v, of the astronomical
body where v  c, is given by:
(λ − λ0 ) (ν0 − ν)
v=c =c (6.18)
λ0 ν
where c is the speed of electromagnetic radiation. Radial velocity may also be
defined in terms of observed frequency ν, and laboratory frequency ν0 . For bodies
moving towards the observer, the spectral features (lines or groups of lines or bands)
are shifted towards shorter wavelengths (higher frequencies) and are said to be ‘blue
shifted’ with a negative radial velocity, whilst those bodies moving away have a
positive radial velocity and their spectral features are said to be ‘red shifted’. A
positive radial velocity corresponds to lower measured frequencies, whilst negative
radial velocities correspond to higher measured frequencies.
Table 6.4 lists the heliocentric radial velocities of the ten brightest stars in the
night sky, with an associated measurement error, a quality rating from A = best
quality measurement to E = worst quality measurement and a simple description of
the star. This information was extracted from the SIMBAD astronomical database.34
Note that all the radial velocity values are less than 100 km . s−1 , implying that the
km . s−1 is a sensible SI unit to use for radial velocities. In fact, since the largest
radial velocity that could be measured is less than 300 000 km . s−1 and the smallest
that is currently being measured is approximately 1 m . s−1 ,35 all radial velocities
are comfortably covered by the SI unit for linear velocity.

6.3.3 Space motion of stars

The components of a star’s space velocity, v, are its radial velocity vr along the line
of sight to the star, and its transverse velocity vt at right angles to that line of sight.
If the motion of the star makes an angle θ to the line of sight, then:
vr = v cos θ

34 See http://simbad.u-strasbg.fr/simbad/sim-fid
35 In radial velocity searches for exoplanets (planets which orbit stars other than the Sun).
86 Unit of length (metre)

Table 6.4. Radial velocities and brief classifications for the ten brightest stars in
the night sky

Star name Radial velocity km . s−1 Quality rating SIMBAD description

Sirius A −7.6 ± 0.9 A spectroscopic binary
Canopus + 20.5 ± 0.9 A star
α Cen A −24.6 ± 0.9 A double star
Arcturus −5.2 ± 0.9 A variable star
Vega −13.9 ± 0.9 A variable star
Capella +30.2 ± 0.9 A RS Cvn type variable star
Rigel +20.7 ± 0.9 A emission-line star
Procyon −3.2 ± 0.9 A spectroscopic binary
Betelgeuse +21.91 ± 0.51 B semiregular pulsating star
Achernar +16 ± 5 C Be star

vr B
Σ
θ vt
ight
of s v
line
S μ C

Figure 6.4. A plot showing the relationship between the transverse and radial
velocity components and the space velocity of a star. The proper-motion angle is
labelled μ, the Sun S, the star , and the line of sight SB. The angle BC is a
right angle.

and

vt = v sin θ (6.19)

In SI units the radial velocity is measured in m . s−1 . The transverse component is

the product of the distance r, in metres to the star, and its angular proper motion μ,
in rad . s−1 , so:
vt = r . μ (6.20)

Radial and transverse components are combined to yield the space velocity v of the
star, thus:

v = (vt2 + vr2 ) (6.21)

The various components are shown in Figure 6.4.

 6.3 Linear motion 87

Example: determine the space velocity of Vega

Consider the prominent nearby A-type star Vega. From Table 6.2 its dis-
tance is 2.3933×1017 m, from Table 6.4 its radial velocity is −13 900 m . s−1
and from Table 5.4 its equatorial coordinate proper-motion components are
(3.088×10−14 , 4.416×10−14 ) rad . s−1 . Combine the proper motion components
to give μ = 5.389×10−14 rad . s−1 . Now multiply μ by the distance to Vega in
metres to obtain the transverse velocity in m . s−1 , vt = 12 900. Finally, take the
square root of the sum of the squares of the velocity components:

v = (−13 900)2 + (12 900)2 = ± 18 961 m . s−1 (6.22)

or, using an appropriate prefix, v = ± 18.96 km . s−1 .

Since, from its measured radial velocity, it is known that the star Vega is
moving towards the Sun, the correct space velocity is the negative solution, so
v = −18.96 km . s−1 .

6.3.4 Red shifts and look-back times

The definition of radial velocity given in Equation (6.17) is strictly Newtonian and
applies only when v  c. Extragalactic objects that lie beyond the local group of
galaxies have spectra which are shifted towards the red with respect to a comparison
terrestrially based laboratory spectrum. The amount of the shift z is accordingly
known as the red shift and is defined, in both wavelength and frequency terms, as:
λ − λ 0 ν0 − ν
z= = (6.23)
λ0 ν
or, by a simple rearrangement:
λ ν0
z= −1= −1 (6.24)
λ0 ν
Being a ratio of wavelengths or frequencies, the red shift is of dimension 1 and
hence is a dimensionless unit:

dim(z) = [L] . [L]−1 = [T −1 ] . [T −1 ]−1 = 1 (6.25)

For radial velocities, v  c, the relationship between z and v may be written as:

v=c.z (6.26)

Hubble (1929) showed that the velocity v, with which a galaxy is receding (its
recession velocity), is directly proportional to its distance D, i.e.:

v = H0 . D = c . z (6.27)
88 Unit of length (metre)

For very distant galaxies, the recession velocity is no longer very small in compari-
son with the velocity of light and the relationship has to be written in the relativistic
form:
 
(z + 1)2 − 1
v=c. (6.28)
(z + 1)2 + 1
There are three component parts to a red shift: that due to the actual or peculiar
motion of the celestial object itself; that due to the gravitational field of the object
and known as the gravitational red shift; and that due to the expansion of the
Universe and known as the cosmological red shift.
Since, in general, celestial objects are detected by the electromagnetic radiation
they emit, and given that the speed of such radiation is finite, then the further
away an object appears to be, the longer ago the detected radiation was emitted.
This leads to yet another unit in common use the look-back time or sometimes
its complement, the time elapsed since the Big Bang, which occurred some 13.7
billion years ago. Note that the sum of these two numbers is the age of the Universe
at the time the observation was made. The dimension of look-back time and time
elapse since the Big Bang is [T ] and its SI unit is the second and its IAU unit the
Julian year.
Wright (2006) has constructed an online calculator36 which, for a given cos-
mological model and input value of red shift, computes distances in parsecs and
light years, light travel times, the age of the Universe at the departure time of the
electromagnetic radiation and other variables of use in cosmological studies.

6.4 Acceleration
The dimension of the SI derived unit of acceleration is [L] . [T ]−2 , its unit is the
metre per second per second and its symbol is m . s−2 .

The acceleration at the surface of a celestial body due to its gravity is known as
its surface gravity.
In Cox (2000), the values of the surface gravity of the bodies in Table 6.5 are
given in cgs units. To convert to SI units, divide the cgs value by 100, e.g., the
surface gravity of Mars:
Mars(cgs) 371
Mars(SI) = = = 3.71 m . s−2
100 100
Whilst cgs units are still in common use for surface gravity measurements, SI
units are now readily accepted and recommended by the IAU.

36 See http://astro.ucla.edu/∼wright/DlttCalc.html
 6.6 Volume 89

Table 6.5. Surface gravities of some bodies in the Solar System

Surface gravity Surface gravity

Name Description m . s−2 Name Description m . s−2
Mercury planet 3.70 Pluto dwarf planet 0.81
Venus planet 8.87 Moon satellite 1.62
Earth planet 9.81 Sun star 274.0
Mars planet 3.71
Jupiter planet 23.12
Saturn planet 8.96
Uranus planet 8.69
Neptune planet 11.0

6.5 Area
The dimension of the SI unit of area is [L] . [L] = [L]2 , its name is the square metre
and its symbol is m2 .
This unit is in common use in astronomy, as are the prefixed units cm2 and km2 ,
and surface areas of planets relative to that of the Earth and the stars relative to that
of the Sun.
Celestial bodies such as the planets and stars generally have a regular shape
and are mainly either spheres or oblate spheroids.37 The surface area, A of such
bodies, with an equatorial radius 38
 a −c   of √
a, a polar
 radius of c, and values of oblateness
a −c
ω = c and eccentricity e = c are given by:

 4 π a2 if c=a
  
A= 2 (6.29)
 2 π a 2 + πec . ln 11 +
−e
e
if a>c

6.6 Volume
The dimension of the SI unit of volume is [L] . [L] . [L] = [L]3 , its name is the cubic
metre, and its symbol is m3 .

This unit is in common use in astronomy, as are the prefixed SI units cm3 and
km3 , and volumes of planets relative to that of the Earth and the stars relative to
that of the Sun.

37 See http://mathworld.wolfram.com/OblateSpheroid.html
38 Note that the mathematical expressions for oblateness and flattening are identical.
90 Unit of length (metre)

Table 6.6. Radii, surface areas and volumes of planets and one dwarf planet in
the Solar System

Equatorial radius Surface area Volume

Name m km r⊕ m2 A⊕ m3 V⊕
Mercury 2.43976 d 6 0.383 7.480 d 13 0.147 6.083 d 19 0.056
2 439.76
Venus 6.052 30 d 6 0.9488 4.603 d 14 0.902 9.286 d 20 0.857
6 052.30
Earth 6.378 14 d 6 1.0000 5.101 d 14 1.000 1.083 d 21 1.000
6 378.14
Mars 3.397 52 d 6 0.533 1.444 d 14 0.283 1.632 d 20 0.151
3 397.52
Jupiter 7.149 2 d 7 11.209 6.164 d 16 120.838 1.437 d 24 1327
71 492
Saturn 6.026 8 d 7 9.449 4.295 d 16 84.213 8.351 d 23 771
60 286
Uranus 2.555 9 d 7 4.007 8.087 d 15 15.854 6.837 d 22 63.1
25 559
Neptune 2.476 4 d 7 3.883 7.620 d 15 14.939 6.254 d 22 57.739
24 764
Pluto 1.195 d 6 0.187 1.795 d 13 0.035 7.148 d 18 0.0066
1 195

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .

Volumes V of spheres and oblate spheroids are computed from one or other of
the following expressions:

 4 π a3 if c = a

V =  34 (6.30)
 π a c if a > c
2
3

Table 6.6 was compiled by substituting data on the radii (listed in km and Earth
radius units) and oblateness from the JPL D405 ephemeris and The Astronomi-
cal Almanac Online39 into Equations (6.29) and (6.30). The planets Mercury and
Venus, and the dwarf planet Pluto, were treated as spheres, whilst Earth to Neptune
inclusive were treated as oblate spheroids. In the table, column 2 lists the planetary
radii in both m (SI unit) and km; column 3 gives the radii in Earth radius units;
columns 4 and 5 the planetary surface areas in m2 and Earth surface area units,
and columns 6 and 7 planetary volumes in m3 and relative to the volume of the
Earth.

39 http://asa.usno.mil/index.html, page K7, 2010.

 6.7 Summary and recommendations 91

6.7 Summary and recommendations

6.7.1 Summary
The SI unit of length, the metre, was originally determined over 200 years ago to
be one ten millionth part of a meridian arc on the Earth’s surface running from the
equator to the pole. This length was represented by a pure platinum rod housed
under controlled conditions. The precise length of the metre is now fixed by the
definition of the speed of light.
The metre and its prefixed derivative, the kilometre, are normally used for mea-
surements of length on the Earth and in near-Earth space (e.g., the distance to the
Moon). The radius of the Earth is often used as a longer comparative unit in studies
of planetary bodies in the Solar System.
The astronomical unit is used within the Solar System for the computation of
planetary ephemerides. Originally defined in terms of the Earth–Sun distance, a
revised definition is now the subject of vigorous debate. Whilst adequate as a dis-
tance measure in the Solar System, the astronomical unit is too small for interstellar
distances, where the parsec and light year are commonly used.
The linear motion of celestial bodies is measured in either the SI derived unit,
the m . s−1 or, more usually, the km . s−1 . Other astronomical velocity-related units
include red shifts and look-back times.
Acceleration is measured in SI units of m . s−2 or the older cgs unit of cm . s−2 .
Examples of the use of this unit as a measure of the surface gravity of bodies in the
Solar System is given.
Areas and volumes have SI units m2 and m3 , respectively. They are used for
all celestial bodies, though km2 and km3 are common, as is the use of the Earth’s
radius, surface area and volume as a larger unit for comparative measurements. For
larger objects, the Sun’s radius, surface area and volume are used.

6.7.2 Recommendations
Wherever possible the SI units, m, m . s−1 , m . s−2 , m2 and m3 should be used.
Until more accurate distance measurements in metres are available, particularly
for objects in the outer Solar System, it is appropriate and sensible to go on using
astronomical units that depend solely on measures of angle and time. The IAU
recommends the use of the parsec, though the light year may prove easier to define,
given that it does not rely on the astronomical unit but only on the defined speed
of electromagnetic radiation and a defined period of time (one Julian year).
SI units of velocity are routinely used in either the m . s−1 form or, for greater
velocities, the km . s−1 form, and this should continue. For cosmological purposes,
red shifts and look-back times can be of more use than simple distance measures.
 7
Unit of mass (kilogram)

7.1 SI definition of the kilogram

The kilogram is the unit of mass; it is equal to the mass of the international prototype
of the kilogram.
The dimension of mass is [M ], its unit is the kilogram and its symbol is kg.

7.1.1 The International Prototype Kilogram

The original definition (1795) of the kilogram was a mass equal to that of a cubic
decimeter of pure air-free water at the temperature of melting ice (273.15 K). This
was altered four years later to the mass of water in the same volume but at the
temperature at which water has its maximum density (which occurs at 277.13 K;
Kaye & Laby, 1959). An all-platinum prototype with the same mass as the cubic
decimeter of water was manufactured the same year and designated the Kilogramme
des Archives. The current standard kilogram mass, a cylindrical platinum–iridium
alloy, made in 1879 and accepted as the standard since 1889, is known as the
International Prototype Kilogram, (IPK). It is now the only SI standard which is a
manufactured artifact.40 The IPK and six replicas are stored at BIPM in a controlled
environment. Further copies, known as replicas, were manufactured for distribution
to other national metrology laboratories throughout the world.

7.1.2 The stability of the International Prototype Kilogram

As the IPK has, by definition, a mass of one kilogram, it has a zero measurement
error. However, when the mass of the IPK is compared with the masses of the
replicas, the IPK is apparently losing mass relative to all of them. Conversely, it
may be that the replicas are gaining mass relative to the IPK. The apparent loss
in mass of the IPK relative to the official copies over a period of just over 100
40 See http://www.sizes.com/units/BIPM.htm and http://www.sizes.com/units/kilogram.htm

92
 7.1 SI definition of the kilogram 93

years from 1889 (Girard, 1994) has amounted to 30 ± 100 μg or approximately

0.003%.41
Such a time-dependent variation is quite obviously unacceptable in the base unit
for mass and is of great concern to metrologists worldwide. Active consideration is
being given to redefining the unit of mass so that it does not depend on a physical
prototype.

7.1.3 Possible future definitions of the kilogram

The CIPM recommended in 2005 that the kilogram be redefined in terms of a
fundamental natural constant. It is possible that a new definition will be forthcoming
at the 2011 meeting of the CIPM. Some possible ways of arriving at a new definition
of the kilogram include the following.
Definition using atomic or subatomic particles, e.g., the number of atoms of
carbon or silicon or the number of electrons required to equal the mass of, say, the
IPK, at a specified date and time.
Definition by force, using the relationship:

F = m×a (7.1)

where F is the force in newtons (kg . m . s−2 ), m the mass in kg and a the acceleration
in m . s−2 . Applying a force of 1 N to a mass of 1 kg will result in an acceleration
of 1 m . s−2 .
Defined by Planck’s constant, using the relationship between energy E in joules
(m2 . kg . s−2 ), frequency ν, in hertz (s−1 ) and Planck’s constant h, in J . s:

E = hν (7.2)

The energy equivalent of a mass of 1 kg is given by:

E = m c2 (7.3)

Now rearrange Equation (7.2) and substitute the values for h, E, m (= 1 kg) and c
(= 299 792 458 m . s−1 ) to find ν:
E (299 792 458)2
ν= =
h 6.626 068 96 × 10−34
= 1.356 × 1050 Hz (7.4)

It is not practical to measure such a high frequency directly.

41 www.french-metrology.com/en/feature/watt-balance.asp
94 Unit of mass (kilogram)

In a paper on redefining the kilogram, Mills et al. (2005) have proposed six
separate revised definitions, three that fix the value of the Planck constant and three
that fix the value of the Avogadro constant. An example of each type follows.
The kilogram is the mass of a body at rest whose equivalent energy corresponds to
frequency of exactly [(299 792 458)2 / (6 626 069 311)] × 1043 Hz.

The kilogram is the mass of a body at rest such that the value of the Avogadro
constant is exactly 6.022 141 527 × 1023 mol−1 .
Until a final decision is made about the revised formal definition by the CIPM
the definition of the kilogram remains as set out at the beginning of this chapter.

7.2 The constant of gravitation

The gravitational force F , with which two bodies of masses M and m attract
one another is proportional to the product of their individual masses and the
inverse square of their distance r apart. This relationship may be expressed
mathematically as:
Mm
F =G 2 (7.5)
r
where G, the constant of proportionality, is known as the gravitational constant.
The dimension of G is [L]3 . [M ]−1 . [T ]−2 and its SI unit is m3 . kg−1 . s−2 .
The gravitational constant is very difficult to measure and the accuracy of
its measurement has not improved greatly since it was first determined in
the late eighteenth century by Henry Cavendish. A recent determination by
Fixler et al. (2007) using a gravity gradiometer based on atom interferometry
yielded a value of 6.693 × 10−11 m3 . kg−1 . s−2 , with both the standard error
about the mean value of ±0.027 × 10−11 m3 . kg−1 . s−2 and a systematic error
of ±0.021 × 10−11 m3 . kg−1 . s−2 given. The CODATA 2006 value is slightly
different at G = 6.674 28 × 10−11 m3 . kg−1 . s−2 .
The gravitational constant is related to the Gaussian gravitational constant k by:
G = k2 (7.6)
The dimension of k is:
3 1
dim(k) = [L] 2 . [M]− 2 . [T ]−1 (7.7)
and its value is:
3 −1
k = 0.017 202 098 95 (A) 2 . (M ) 2 . (d)−1 (7.8)
when A is distance in astronomical units, M is mass in solar mass units and d is
time measured in days of length 86 400 SI seconds.
 7.2 The constant of gravitation 95

7.2.1 Values of G measured in some other systems of units

The dimension of G is:

dim(G) = [L]3 . [M]−1 . [T ]−2 (7.9)

and its value in SI units is 6.673 × 10−11 m3 . kg−1 . s−2 . To convert G to other
systems of units it is only necessary to determine the appropriate constant of
proportionality between the systems.

G(other units) = [kL . L]3 . [kM . M]−1 . [kT . T ]−2 × GSI (7.10)

Example 1: determine the value of G in cgs units

kL = 100; kM = 1000; kT = 1, where kL is the number of centimetres in a metre,
kM is the number of grams in a kilogram and kT = 1, since the unit of time, the
second, is the same in both systems.

Gcgs = [100]3 . [1000]−1 × GSI

= 1000 . GSI
= 6.693 × 10−8 cm3 . g−1 . s−2 (7.11)

Example 2: determine the value of G in IAU astronomical units

Firstly, compute the number of SI units in each corresponding type of IAU unit.
kL = 1/(1.495 978 71 × 1011 ); kM = 1/(1.989 × 1030 ); kT = 1/86 400 where kL is
the number of astronomical units in a metre, kM is the number of solar masses in a
kilogram and kT is the number of julian days in an SI second, thus:
 1 3  1 −1  1 −2
GIAU = . . . GSI
1.495 978 71 × 1011 1.989 × 1030 86 400
 4.435 × 106 . GSI
= 2.968 × 10−4 (A)3 . (M )−1 . (d)−2 (7.12)

Example 3: G using the parsec as a unit of length

Ggal , which could be of use for dynamical studies in the Milky Way Galaxy, uses
the parsec as the unit of length, the solar mass as the unit of mass and the km . s−1
as the unit of velocity. In this case the dimensional equation is somewhat different:

dim(Ggal ) = [kP . L] . [kM . M]−1 . [kV . L]2 . [T ]−2 (7.13)

96 Unit of mass (kilogram)

where kP = 1/(3.085 677 6 × 1016 ); kM = 1/(1.989 1030 ); kV = 1/1000, with kP

being the number of parsecs in a metre, kM the number of solar masses in a kilo-
gram, kV the number of kilometres in a metre and the second being common to
both systems, then
 1   1 −1  1 2
Ggal = . . . GSI
3.085 677 6 × 1016 1.989 × 1030 1000
 6.446 × 107 . GSI
= 4.314 × 10−3 (pc) . (M )−1 . (km . s−1 )2 (7.14)

7.2.2 Standard gravitational parameter

The standard gravitational parameter μ, is the product of the Newtonian gravita-
tional constant G, and the mass of a given celestial body. If that body is the Sun then
the product G . M , is known as the heliocentric gravitational constant, or if the
body is the Earth, then the product G . M⊕ is known as the geocentric gravitational
constant.
From The Astronomical Almanac Online42 the values for the heliocentric and
geocentric gravitational constants are given, in SI units, as:

μ = G . M = 1.327 124 420 76 × 1020 m3 . s−2 (7.15)

μ⊕ = G . M⊕ = 3.986 004 418 × 1014 m3 . s−2 (7.16)

Note that the accuracy with which the values of μ may be expressed far exceeds that
of the constant of gravitation as determined in the laboratory. The best current values
for G have standard deviations about the mean measured value of approximately
±0.01 % or 1 part in 104 , whereas the heliocentric gravitational constant is known
to approximately 1 part in 1010 . Hence it is customary when computing the values
for semimajor axes of Solar System bodies to use IAU units rather than SI units
via the relationship (Kepler’s third law):
  
3
 G . M . P 2  3
 2
k . M . P 2  3
μ.P 2 
a= = = (7.17)
4.π2 4.π2 4.π2
where a is the semimajor axis of the orbit, P is the orbital period, M is the mass of
the Sun, G is the Newtonian gravitational constant, k is the Gaussian gravitational
constant and μ is the standard gravitational parameter. Planetary ephemerides are
currently computed using k with distances in astronomical units and masses in
inverse mass units M / MP .

42 http://asa.usno.mil/index.html, page K7, 2010.

 7.3 Masses of astronomical bodies 97

The standard gravitational parameter may be transformed from SI units to IAU

units in the following way. The dimension of μ is:

dim(μ) = [L]3 . [T ]−2 (7.18)

so
μIAU = [kL . L]3 . [kT . T ]−2 . μSI (7.19)

where kL is the number of astronomical units in a metre and kT is the number of

julian days in an SI second, hence:
 1 3  1 −2
μIAU = . . μSI
1.495 978 71 × 1011 86400
 2.230 × 10−24 . μSI
= 2.959 × 10−4 (A)3 . (d)−2 (7.20)

7.3 Masses of astronomical bodies

The masses of bodies in the Solar System may be determined using Kepler’s third
law and Newton’s second law, augmented by direct measurements for the Earth
and helioseismology for the Sun. For binary stars, Kepler’s third law is again
invoked. For single stars, the empirically determined relationship between mass
and luminosity may be used, as well as asteroseismology, and for stars which act
as the lensing star in a microlensing event, general relativity allows an estimate
of masses through the observed alignment geometry of the two stars (source and
lens). Galaxies present more of a problem since the use of Kepler’s third law is
inadequate when the possibility of dark matter and even dark energy have to be
taken into account.

7.3.1 The solar mass

By rearranging Equation (7.17), the mass of the Sun, M , may be expressed
approximately in SI units as:

4 . π 2 . a3
M  (7.21)
G.P 2
where a is the semimajor axis of a planetary orbit in metres, P is the period of revolu-
tion in SI seconds, and G is the Newtonian gravitational constant in m3 . kg−1 . s−2 .
Substituting the values for a and P that approximate for the Earth and making the
assumption that the mass of the Earth is small in comparison with that of the Sun
98 Unit of mass (kilogram)

gives:
4 . π 2 . (1.495 978 714 64 × 1011 )3
M 
6.673 × 10−11 . (365.25 × 86 400)2
 1.989 × 1030 kg (7.22)

Note that if IAU units are used, then P = 365.25 d, a = 1 (A), and G = 2.968 ×
10−4 (A)3 . (M )−1 . (d)−2 , so that:
4 . π 2 . 13
M 
2.968 × 10−4 . 365.252
 0.997
1 (7.23)

The mass of the Sun may also be determined from observations of the orbital
velocity and position of a planet at a particular time. If the measured velocity of
the planet is vP , and its radius vector is RP , then the approximate mass of the Sun
may be calculated from:
v 2 . RP
M = P (7.24)
G
where G is the Newtonian constant of gravitation.

Example: determine the approximate mass of the Sun from observations

of the planet Venus
The heliocentric coordinates (x , y , z) in astronomical units and velocity compo-
nents (ẋ , ẏ , ż) in astronomical units per day of Venus at JD = 2 449 720.5 are given
in the The Astronomical Almanac (1995) as:

x = −0.541 079 4 y = +0.417 342 5 z = +0.221 997 3

ẋ = −0.013 358 66 ẏ = −0.014 313 22 ż = −0.005 593 58

The length of the radius vector Sun–Venus, R, is given by:


R = x 2 + y 2 + z2 (7.25)

and the rate of change of the radius vector, v, by:


v = ẋ 2 + ẏ 2 + ż2 (7.26)

Substituting the coordinates and velocity components from above in Equations

(7.25) and (7.26) gives values for the radius vector and its time derivative of
R = 0.718 487 63 (A) and v = 0.020 361 98 (A) . (d)−1 . Using the value of G in
 7.3 Masses of astronomical bodies 99

IAU units plus the values for RP = R and vP = v in Equation (7.24) yields a value
for the mass of the Sun of M  1.004 (in solar mass units).
To determine the mass of the Sun in SI units multiply R by
1.495 978 714 64 × 1011 (the number of metres in an astronomical unit) and divide
by the number of SI seconds in a julian day (86 400). Insert the adjusted numbers
into Equation (7.24) to give M  2.002 × 1030 kg.

The solar mass in general relativity

In general relativity, the mass of the Sun is sometimes given in the form of a unit
of length (M )L or time (M )T as follows:
G . M 1.327 124 420 76 × 1020
(M )L = =
c2 (2.997 924 58 × 108 )2
= 1.476 625 061 × 103 m (7.27)
G . M 1.327 124 420 76 × 1020
(M )T = =
c3 (2.997 924 58 × 108 )3
= 4.925 491 025 × 10−6 s (7.28)
Note that the solar mass units are now quoted in metres and seconds, which may
be verified using dimensional analysis.
dim(G) . dim(M )
dim((M )L ) =
dim(c2 )
[L]3 . [M]−1 . [T ]−2 . [M]
=
[L]2 . [T ]−2
= [L] (7.29)

dim(G) . dim(M )
dim((M )T ) =
dim(c3 )
[L]3 . [M]−1 . [T ]−2 . [M]
=
[L]3 . [T ]−3
= [T ] (7.30)

7.3.2 Mass of the Earth

The earliest attempts at determining the mass of the Earth relied on regularly shaped
mountains. A plumb bob is gravitationally attracted by the mountain so that the
plumb line no longer points directly to the zenith. Measurements of this small
deviation and the estimated mass of the mountain leads to an approximate value of
100 Unit of mass (kilogram)

the mass of the Earth. Laboratory-based determinations later used torsion balances
and gravity balances (Spencer-Jones, 1956).
The advent of artificial satellites in orbit about the Earth, which obey Kepler’s
third law, allow the same dynamical method to be employed in the determination
of the Earth’s mass as is used for the Sun. If, in Equation (7.21), the mass of the
Earth M⊕ , is substituted for the mass of the Sun, and the semimajor axis aS and
the orbital period PS of the artificial satellite for those of the Earth, then:
4 . π 2 . aS3
M⊕ = (7.31)
G . PS2

Example: The International Space Station and the mass of the Earth
By way of an example it is possible to determine approximately the mass of the Earth
using observations of the International Space Station. Measurements of the orbital
period in minutes and the height of the apogee point of the orbit above the surface
of the Earth in kilometres are to be found on the satellite database of astrosat.net.43
The semimajor axis length of the satellite’s orbit is taken to be the sum of the
Earth’s equatorial radius and the satellite’s apogee distance and the listed orbital
period multiplied by 60 to convert to seconds of time. The value GSI is substituted
in Equation (7.31) for G to give:
4 . π 2 . (6.732 × 106 )3
M⊕ =
6.673 × 10−11 . (5.491 × 103 )2
= 5.97 × 1024 kg (7.32)

which, allowing for the very basic analysis performed, is a reasonable approxi-
mation to the mass of the Earth given in The Astronomical Almanac Online as
5.972 198 6 × 1024 kg.44
For a more detailed account of how to determine the mass of the Earth or any
other planet taking into account the oblateness of the planet see, e.g., Blanco &
McCuskey (1961).

7.3.3 Masses of Solar System objects

The masses of all the major planets and some of the dwarf planets in the Solar Sys-
tem may be determined using observations of their oblateness in conjunction with a
detailed study of the orbits of one or more of their satellites (either natural or artifi-
cial). Other than using the SI kilogram, the masses of the Sun and the planets Earth
and Jupiter are also commonly used as units in planetary and exoplanetary studies.

43 See www.satview.com.br/us/track_lista_sat.php
44 http://asa.usno.mil/index.html, page K7, 2010.
 7.3 Masses of astronomical bodies 101

Table 7.1. Masses, in kilograms, solar mass units, reciprocal solar mass units,
Jovian mass units and Earth mass units of the Sun, planets, dwarf planets,
satellites and a comet in the Solar System

Name Mass
Star kg M 1 / M MJ M⊕
Sun 1.9884 d 30 1 1 1.0473 d 3 3.3294 d 5
Planet
Mercury 3.3010 d 23 1.6601 d − 7 6.0236 d 6 1.7387 d − 4 0.0553
Venus 4.8673 d 24 2.4478 d − 6 4.0852 d 5 2.5637 d − 3 0.8150
Earth 5.9721 d 24 3.0035 d − 6 3.3295 d 5 3.1457 d − 3 1
Mars 6.4169 d 23 3.2272 d − 7 3.0987 d 6 3.3800 d − 4 0.1074
Jupiter 1.8985 d 27 9.5479 d − 4 1.0473 d 3 1 317.891
Saturn 5.6846 d 26 2.8589 d − 4 3.4979 d 3 2.9942 d − 1 95.184
Uranus 8.6818 d 25 4.3662 d − 5 2.2903 d 4 4.5730 d − 2 14.5371
Neptune 1.0243 d 26 5.1514 d − 5 1.9412 d 4 5.3953 d − 2 17.1512
Dwarf Planet
Ceres 9.3455 d 20 4.7 d − 10 2.1 d 9 4.9 d − 7 1.6 d − 4
Pluto 1.4707 d 22 7.3964 d − 9 1.3520 d 8 7.7467 d − 6 2.4626 d − 3
Eris 1.9884 d 20 1 d − 10 1 d 10 1.0 d − 7 3.3294 d − 5
Satellite
Moon 7.3458 d 22 3.6943 d − 8 2.7068 d 7 3.8693 d − 5 1.230 d − 2
Ganymede 1.4818 d 23 7.4522 d − 8 1.3419 d 7 7.805 d − 5 2.48 d − 2
Europa 4.7994 d 22 2.4137 d − 8 4.1430 d 7 2.528 d − 5 8.04 d − 3
Comet
Halley 5 d 14 2.4 d − 16 4 d 15 2.6 d − 13 8.4 d − 11

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .

Examples of the masses of the planets, dwarf planets, satellites and Halley’s
comet are given in Table 7.1. The values were computed from data given in Cox
(2000) (for the Moon), Brown & Schaller (2007) (for Eris) and The Astronomical
Almanac Online.
It should be remembered that until the value of the solar mass in kilograms is
greatly improved, the relative masses of the planets in solar mass units are much
more accurate. In Table 7.1, the masses in kilograms of the various objects were
derived by multiplying the mass of the body in solar mass units by one solar mass
in kilograms which, being given to five significant figures, immediately restricts
all the other derived values to being no better than five significant figures, and
generally rather worse. The values given for the reciprocal solar masses of these
same objects range from seven significant figures for Uranus and Neptune up to 12
significant figures for the Earth, which explains why IAU units are to be preferred
for the computations of ephemerides at present.
102 Unit of mass (kilogram)

7.3.4 Stellar masses

The masses of double stars that are gravitationally bound may have their component
masses determined using Kepler’s third law and Newton’s laws of motion.
In a binary star system each component describes an elliptical orbit about the
common centre of gravity (the barycentre), as is illustrated in Figure 7.1. If the
binary star is relatively close to the Sun then each component star may be indi-
vidually observed and plotted on the plane of the sky to construct an orbit. If the
binary is much further away the component stars may not be resolved as individ-
ual stars but may be detected via their composite spectrum, in which the spectral
lines generated by each star will move relative to each other. This type of binary is
known as a double-line spectroscopic binary, a special case of which occurs when
one member star is so much fainter than the other that its spectral lines are invisible
but its presence is still detectable by the effect it has on the spectral lines of the
brighter component, which shift in a regular fashion. Such a binary is known as
a single-line spectroscopic binary. For spectroscopic binaries of either type whose
orbital plane lies along the line of sight to the binary, regular eclipses occur as one
component star passes in front of the other. These objects are known as eclipsing
binaries.
For a binary star, the sum of the masses of the components may be derived from
Kepler’s third law:
4.π 2 .a 3
MA + M B = (7.33)
G.P 2
where a is the semimajor axis of the absolute elliptical orbit in metres, P is the
orbital period of the system in SI seconds and MA and MB are the masses of the
individual stars in kilograms, with G the constant of gravitation in SI units. So, if
the orbit of the binary is well defined, it is a simple matter to substitute appropriate

B⬙

B⬘
A

C A⬘
A⬙

Figure 7.1. The orbits of a pair of stars, A and B, in a binary system about their
centre of mass, C, the barycentre. Star A moves from A through A to A whilst
star B moves from B through B to B . The barycentric distances BC and CA are
equal to rB and rA , respectively.
 7.3 Masses of astronomical bodies 103

values into Equation (7.33) to obtain the total mass of the binary system. To obtain
the individual masses, consider the binary star orbits shown in Figure 7.1. The
distance from star A to the barycentre C, is rA , and likewise star B is distance rB
−→
from the barycentre. Distance AB = r, so r = rA + rB . The total mass of the system
is M, where M = MA + MB and MA . rA = MB . rB . Using Newton’s second law of
motion balancing gravitational and centripetal force gives:45

G . MA . MB MA . ṙA2
= (7.34)
r2 rA
where ṙA is the orbital speed of star A, which may be measured spectroscopically
or simply taken as the mean value defined by:
2 . π .rA
ṙA = (7.35)
P
Combining the various expressions above produces the relationship:
M . (r − rA )
MA = (7.36)
r
The value for the mass of star B is derived from:

MB = M − MA (7.37)

If IAU units are used with mass in solar mass units, distance in astronomical
units and time in days of 86 400 SI seconds, Equation (7.33) may be rewritten as:
3
4 . π 2 . aIAU
MIAU = 2
(7.38)
GIAU . PIAU
which gives the mass of the binary system in solar mass units. Equations (7.34) to
(7.37) may similarly be reworked.

Example: determine the total mass of the Sirius binary system in both IAU
and SI units
(1) IAU units
Parallax: 379 mas
Distance: 206 264.806
0.379
25
= 5.442 343 × 105 au
7.5
Orbital semimajor axis (in radians): 206 264.806 25 rad
7.5×5.442 343×105
Orbital semimajor axis (in astronomical units): 206 264.806 25 = 19.789 au

45 For more detail and a worked example, see

http://outreach.atnf.csiro.au/education/senior/astrophysics/binary_mass.html
104 Unit of mass (kilogram)
−1 −2
Constant of gravitation: 2.968 × 10−4 A3 .M .d
4 . π 2 ×(19.789)3
Total mass of the Sirius system: 2.968×10−4 ×(1.829 537×104 )2
= 3.08 M
(2) SI units
Distance: 8.141 63 × 1016 m
Period: 1.580 72 × 109 s
Orbital semimajor axis: 2.960 38 × 1012 m
Constant of gravitation: 6.673 × 1011 m3 . kg−1 . s−2
Total mass of the Sirius system: 6.143 × 1030 kg ( 3.09 M )
The Sirius binary system has the catalogue number WDS06451-1643 in the
online Washington Double Star Catalog. The orbital and distance data were obtained
from this catalogue.46
It is not uncommon in binary-star research to measure the orbital period of the
system in years rather than days, which alters the approximate value of 4 . π 2 / G
to unity, as follows:
Pd = 365.25 × Py
3
4 . π 2 . aIAU
M=
GIAU . (Py × 365.25)2
3
aIAU
= 0.997 ×
Py2
3
aIAU
 (7.39)
Py2
So when the semi-major axis of the binary-star orbit is measured in astronomical
units, its period is measured in years and the value of 4 . π 2 / G is set equal to 1,
then the total mass of the binary system is measured in solar mass units.

Asteroseismology
Asteroseismology may be defined as the study of the internal structure of individual
stars through the analysis of their pulsation periods and the interpretation of the
derived frequency spectra. Stellar masses and radii can be determined to a much
higher precision than by any other method.47
This branch of astronomy has matured in recent times, with the development of
high-speed photometers and the establishment of networks of observational sites
throughout the world (e.g., the WET (Whole Earth Telescope) collaboration). The

46 See http://ad.usno.navy.mil/wds/
47 See http://kepler.nasa.gov/Science/about/RelatedScience/Asteroseismology/
 7.3 Masses of astronomical bodies 105

analysis of the power spectrum of the light curve (Winget et al., 1994) allows direct
estimations to be made of the masses of individual stars. For the pulsating DB white
dwarf GD358, Bradley (1993) derived a mass of 0.61 ± 0.03 M .
Asteroseismological reduction techniques have also been applied to observations
made of relatively bright main-sequence and subgiant branch stars using spectro-
graphs. Bedding et al. (2006) used such observations to determine the mass of the
metal-poor subgiant star ν Indi to be 0.85 ± 0.04 M . Ferdman et al. (2010) used
radio telescopes in the USA, Australia and France to make extensive observations
of the intermediate-mass binary pulsar PSR J1802-2124, which yielded masses
for the pulsar component of 1.24 ± 0.11 M and for the white-dwarf component
of 0.78 ± 0.04 M . Such measurements and analysis provide important indepen-
dent estimates of stellar masses using a method other than those of Kepler/Newton
dynamics.
Some examples of stellar masses determined from dynamical and asteroseis-
mological techniques are given in Table 7.2, which was compiled, in part, from
material to be found in Cox (2000), Ferdman et al. (2010), Winget et al. (1994)
and Bedding et al. (2006). It is worth noting that currently there is no prefix in SI
that is large enough for stellar masses, even the smallest mass in Table 7.2 would
be given as 3.182 × 108 Yg (3.182 × 108 yottagrams). The additional, unofficial,
prefixes proposed by Mayes (1994) would describe the mass as 318.2 Sg (318.2
sansagrams). (Again, remembering that the prefix is applied to the mass unit gram
and not to the SI unit kilogram.)

7.3.5 Mass loss and gain

The masses of celestial bodies may be either increased or decreased due to various
physical processes, such as nuclear reactions converting mass into energy, stellar
winds, accretion and collisions.
The dimension of mass change is [M] . [T]−1 its unit is the kilogram per second
and its symbol kg . s−1 . Non-SI units include the kg . y−1 , M . y−1 and the cgs
version g . s−1 .

Example of mass gain: According to Love & Brownlee (1993), the infall of dust
and small meteoroids on to the Earth amounts to some 4 × 107 kg . y−1 , equivalent
to approximately 1.3 kg . s−1 .

Examples of mass loss: in the region near the centre of the Sun, energy is generated
by the conversion of hydrogen nuclei into helium nuclei. This reaction results in a
loss of mass amounting to 4.3 × 109 kg . s−1 , with a further 1.3 × 109 kg . s−1 being
lost via the solar wind. A total mass loss of 5.6 × 109 kg . s−1 , equivalent in IAU
units to 2.4 × 10−16 M . d−1 , or 8.8 × 10−14 M . y−1 .
106 Unit of mass (kilogram)

Table 7.2. Masses of binary-star components determined dynamically

and single stars determined using asteroseismological methods

Mass
Star identifier Spectral type kg M
Sirius A A1V 4.535 d 30 2.28
Procyon A F5IV–V 3.361 d 30 1.69
PSR J1802-2124 A Pulsar 2.466 d 30 1.24
α Cen A G2V 2.148 d 30 1.08
SUN G2V 1.989 d 30 1.00
Sirius B DA 1.949 d 30 0.98
70 Oph A K0V 1.790 d 30 0.90
α Cen B K0V 1.750 d 30 0.88
ν Ind G0IV 1.691 d 30 0.85
70 Oph B K4V 1.293 d 30 0.65
GD358 DBV 1.213 d 30 0.61
Procyon B WD 1.193 d 30 0.60
Krüger 60 A dM4 5.370 d 29 0.27
Krüger 60 B dM6 3.182 d 29 0.16

The recently discovered super-Earth exoplanet CoRoT-7b is located very close to

its star and is subject to intense EUV radiation, causing a mass loss from the planet of
approximately 1011 g . s−1 , or 108 kg . s−1 in SI units, according to Valencia et al.48

7.4 Density
Density may be defined as the mass or quantity of matter of a substance per unit
volume.
The dimension of density is [M ] . [L]−3 , its derived unit is the kilogram per
cubic metre and its symbol, kg . m−3 .

Density unit conversion

Units of density other than the kg . m−3 in common use in astronomy include
the gram per cubic centimetre, g . cm−3 , and the solar mass per cubic parsec,
M . pc−3 . Conversion factors between the various units are given below as worked
examples.

48 Reported by D. Valencia, M. Ikoma, T. Guillot and N. Nettlemann at the 41st Lunar and Planetary Sci-
ence Conference (2010) in a paper entitled ‘Composition and fate of short-period super-Earth’s: The case
of CoRoT-7b.’
 7.4 Density 107

Table 7.3. Densities of various celestial objects

Mean density
Celestial object kg . m−3 g . cm−3 M . pc−3 Notes
Sun 1409 1.409 – mean density
Earth 5514.8 5.515 –
Jupiter 1330 1.33 –
Saturn 700 0.70 –
Moon 3341 3.341 – mean density
Asteroids 2250 2.25 – mean of estimated
range of densities
Halley’s comet 650 0.65 – mean of estimated
range of densities
Interstellar matter 2.71 d − 21 2.71 d − 24 0.04
Main sequence stars 3.38 d − 21 3.38 d − 24 0.050 mass density in the
solar neighbourhood

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .

Example 1: cgs unit conversions



−3 1 1 −3 −3
1 g . cm = . kg . .m
1000 100
= 1000 kg . m−3 (7.40)

the inverse of which is:

1 kg . m−3 = 0.001 g . cm−3 (7.41)

Example 2: solar mass per cubic parsec unit conversions

1M . pc−3 = (1.989 × 1030 ) . kg .(3.085 667 6 × 1016 )−3 . m−3

= 6.769 980 × 10−20 kg . m−3 (7.42)

the inverse of which is:

1 kg . m−3 = 1.477 109 × 1019 M . pc−3 (7.43)

Some examples of the densities of astronomical bodies are given in Table 7.3,
with data from Cox (2000) recalculated where necessary.
A form of density unit that is often used is the column mass density, whose unit
is the M . pc−2 . Column mass density is normally accompanied by a range within
which the quoted value is valid. For example, the observed column mass density to
108 Unit of mass (kilogram)

|z| = 1.1 kpc of the interstellar medium is approximately 13 M . pc−2 . In SI units,

this column mass density would be given as:

|z| = 1.1 × 1000 × 3.085 667 6 × 1016

= 3.394 234 × 1019 m (7.44)
13 × 1.989 × 1030
ρ=
(3.085 667 6 × 1016 )2
= 0.027 157 kg . m−2 (7.45)

so in SI units the column mass density of the local interstellar medium would be
0.027 kg . m−2 within |z| = 3.394 × 1019 m.

7.5 Force
Funk et al. (1946) define force as: ‘Any cause that produces, stops, changes or tends
to produce, stop or change the motion of a body.’
The dimension of force is [M ] . [L] . [T ]−2 , its derived SI unit is the newton with
symbol, N, or, kg . m . s−2 .

Another unit still used by astronomers is the dyne, a cgs unit equivalent to
1 g . cm . s−2 . The conversion coefficient between the two systems is derived as
follows:

1 kg . m . s−2 = 1000 . g . 100 . cm . s−2

= 105 . g . cm . s−2 (7.46)

or 1 newton = 105 dynes.

7.5.1 Energy
The energy of a body is its capacity for doing work, where work is defined as the
product of the force that moves the body and the distance through which it is moved.
The SI units for energy and work are the same.
The dimension of energy is [M ] . [L]2 . [T ]−2 , its derived SI unit is the joule with
symbol, J, or N . m or kg . m2 . s−2 .

The cgs unit of energy, the erg, is regularly used by astronomers. Measures in
ergs are converted to joules via 1 erg = 10−7 J.

The potential energy of a system is the energy that the system has due to the
relative positions of its component parts.
 7.6 Moments of inertia and angular momentum 109

The kinetic energy of a moving body or system is the energy it possesses due to
its motion.
The dimensions, units and symbols of both potential and kinetic energy are the
same as those for energy.

7.5.2 Power
Power is the rate at which energy is transferred per unit time.
The dimension of power is [M ] . [L]2 . [T ]−3 , its derived SI unit is the watt with
symbol W, or J . s−1 , or kg . m2 . s−3 .
The cgs unit of power, the erg . s−1 , is still used as, e.g., a measure of luminosity
in model stellar atmosphere calculations (see Table G.10 in Irwin, 2007).

7.5.3 Pressure
Pressure is a force applied perpendicularly to a unit area.
The dimension of pressure is [M ] . [L]−1 . [T ]−2 , its derived SI unit is the pascal
with symbol, Pa, or N . m−2 , or kg . m−1 . s−2 .
Astronomers tend to use the cgs unit of pressure, the dyn . cm−2 , or the bar,
equal to 106 dyn . cm−2 . The conversion factors from these units to the SI unit of
pressure are:
1 bar = 105 Pa 1 Pa = 10−5 bar
1 dyn . cm−2 = 10−6 bar 1 Pa = 10 dyn . cm−2 (7.47)
The atmospheric surface pressures on the Sun, some planets, a dwarf planet and
a satellite, listed by Cox (2000), are converted to SI units using the appropriate
factors from Equations (7.47) and set out in Table 7.4.
The atmospheric pressure given for the Sun refers to a point where the radial
optical depth at a wavelength of 500 nm is equal to 1.

7.6 Moments of inertia and angular momentum

The moment of inertia, I, of a rigid body about a given axis is the sum of the
products m . r 2 , for all the elements composing the body, where m is the mass of
each element and r is its perpendicular distance from the axis concerned.

I= m . r2 (7.48)

The dimension of moment of inertia is [M ] . [L]2 , its unit is the kilogram metre
squared and its symbol is kg . m2 .
110 Unit of mass (kilogram)

Table 7.4. Surface pressures for a selection of various celestial

objects

Surface atmospheric pressure

Celestial object Pa dyn . cm−2 bar
Star
Sun 1.207 d 4 1.207 d 5 1.207 d−1
Planet
Venus 9d6 9d7 90
Earth 1d5 1d6 1
Mars 1d3 1d4 0.01
Jupiter 3d4 3d5 0.3
Saturn 4d4 4d5 0.4
Dwarf planet
Pluto 8 80 8 d−5
Satellite
Titan 1.5 d 5 1.5 d 6 1.5

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .

Moments of inertia are often presented in cgs units as g . cm2 and, for Solar
System objects, as M . R 2 , where M is the mass of the objects in units of the
Earth’s mass and R is the object’s radius in units of the Earth’s radius.
To convert from cgs to SI units: the conversion factor is derived in the following
way:
 1   1 2
ISI = . . Icgs
1000 100
= 10−7 . Icgs (7.49)
In Cox (2000), the moment of inertia of the Sun is given as 5.7 × 1053 g . cm2 ,
which, using the conversion factor in Equation (7.49), becomes 5.7 × 1046 kg . m2
in SI units.
To convert from M⊕ . R⊕ 2 units: the conversion factor is derived by substituting

values for the Earth’s mass in kilograms and the Earth’s radius in metres:
ISI = (5.972 198 6 × 1024 ) . (6.378 136 6 × 106 )2 . IM⊕ . R 2
⊕
38
= 2.429 527 × 10 . IM⊕ . R 2 (7.50)
⊕

2
In Cox (2000), the moment of inertia of the Earth is given as 0.333 5 M⊕ . R⊕
which is readily converted into SI units via Equation (7.50) to equal
8.10 × 1037 kg . m2 .
 7.7 Summary and recommendations 111

The angular momentum, Iω, of a celestial body about its rotation axis is defined
as the product of its moment of inertia and its angular velocity.
The dimension of angular momentum is [M ] . [L]2 . [T ]−1 , its unit is the
kilogram metre squared per second, and its symbol is kg . m2 . s−1 .
In Cox (2000), the angular momentum of the Sun is given in cgs units as
1.63 × 1048 g . cm2 . s−1 . The time unit is the same in both cgs and SI units. The
conversion factor from cgs to SI units is the same as that given in Equation (7.49),
so that in SI units the solar angular momentum is 1.63 × 1041 kg . m2 . s−1 .

7.7 Summary and recommendations

7.7.1 Summary
The SI unit of mass the kilogram is the only remaining base unit that is defined
by a prototype, the International Prototype Kilogram (IPK), a block of platinum–
iridium of mass 1 kg. Comparison of the mass of the IPK with a set of replicas
over more than a century has shown that the IPK is losing mass. Obviously, this is
an unacceptable situation for a base unit standard and considerable effort is being
expended in trying to arrive at a new definition for the kilogram.
In dynamical astronomy the constant of gravitation is of great importance, being
the constant of proportionality relating the magnitude of the gravitational attraction
or force between two bodies and their individual masses and separation. Unfortu-
nately, it has proved extremely difficult to measure, which has resulted in a low
precision value (∼1 part in 104 ) for the constant. However, the standard gravita-
tional parameter, the product of the constant of gravitation and the solar mass, may
be determined with far more precision (∼1 part in 1010 ).
The masses of celestial objects are determined in a variety of ways using dynam-
ics, asteroseismology, a relationship between mass and luminosity and gravitational
microlensing. Astronomical body masses are most commonly given relative to the
mass of the Sun, Earth or other Solar System objects.
The SI units used for density, force, energy, power, pressure, moments of inertia
and angular momentum are all given and are relevant and simple to use, though are
less common than either cgs units, IAU units or versions of IAU units (e.g., Earth
mass and radius units rather than solar mass and radius units).
Worked examples of how to transform the various cgs and IAU units into SI
units are given throughout the chapter.

7.7.2 Recommendations
The probable change from the IPK to a fundamental particle or energy definition
of the SI kilogram should have no effect on the use of the kilogram as the base unit
of mass in astronomy.
112 Unit of mass (kilogram)

At present, the lack of precision in the determination of G means that it is sensible

to continue using IAU units for planetary dynamics, since only relative masses are
available to a high accuracy.
Unconventional units are regularly used to represent mass – including those
based on terrestrial and Jovian masses. Cgs units are mainly used with moments
of inertia and angular momenta. Atmospheric pressures are regularly quoted in
dyn . cm−2 or bars instead of pascals, though since the differences between them
are only powers of ten it is simple to effect the changeover to SI units. However,
should there be a valid reason for not changing, then perhaps an extra entry in the
abstract and summary giving the main result of the research programme in SI units
should not prove too difficult.
 8
Unit of luminous intensity (candela)

8.1 SI definition of the candela

The candela is the luminous intensity, in a given direction, of a source that emits
monochromatic radiation of frequency 540×1012 hertz and that has a radiant
intensity in that direction of 1/683 watt per steradian.
The dimension of luminous intensity is [J ], its unit is the candela and its symbol
is cd.

8.2 Radiometry and photometry

Radiometry is the measurement of electromagnetic radiation from a fre-
quency / wavelength of 3×1011 Hz / 1000 μm to 3×1016 Hz / 10 nm. The unit used
for such measurements is the watt or m2 . kg . s−3 .

Photometry is a subset of radiometry in being the measurement of electro-

magnetic radiation visible to the human eye and weighted by the response of
the human eye, which is approximately Gaussian, having a maximum sensi-
tivity around 5.41×1014 Hz / 555 nm with cut-offs at 3.86×1014 Hz / 770 nm and
7.89×1014 Hz / 380 nm (Cox, 2000).
The radiometric unit that corresponds to the SI base unit, the candela, is the
watt per steradian (W . sr−1 ). They are related through the formal definition of
the candela by:
1
1 cd ≡ 1 lm . sr −1 ≡ W . sr −1 (8.1)
683
1
where the ( 683 ) resulted from the need to maintain continuity of units when an
earlier definition of the candela was changed.
Given that astronomical observations are now made throughout the electromag-
netic spectrum, it would seem appropriate to link astronomical photometry with

113
114 Unit of luminous intensity (candela)

Table 8.1. Photometric, radiometric and astronomical units

Quantity Photometry symbol Radiometry symbol Astronomy symbol

Power lumen total radiant flux total luminosity
lm W mbol
spectral radiant flux monochromatic luminosity
W . Hz−1 mν
Power lux irradiance flux density
per unit area lm . m−2 W . m−2 W . m−2
spectral irradiance monochromatic flux density
W . m−2 . Hz−1 Jy
Power candela radiant intensity
per unit cd or lm . sr−1 W . sr−1 W . sr−1
solid angle spectral radiant intensity
W . sr−1 . Hz−1 W . sr−1 . Hz−1
Power nit radiance intensity
per unit area cd . m−2 W .m−2 . sr−1 W .m−2 . sr−1
per unit lm . m−2 . sr−1 spectral radiance specific intensity
solid angle W . m−2 . sr−1 . Hz−1 W . m−2 . sr−1 . Hz−1

radiometry rather than photometry as defined (it should be noted that astronomical
observations cover an even larger total bandwidth than that of radiometry).
Equivalent photometric, radiometric and astronomical units and the quantities
they represent are set out in Table 8.1.
The jansky is a non-SI unit that is recognized by the IAU for use in astronomy.
Formally:

The dimension of monochromatic flux density or spectral irradiance is

[M ] . [T ]−2 , its unit is the jansky with symbol, Jy, or 10−26 W . m−2 . Hz−1 .

The jansky was introduced by radio astronomers and is also used by infrared
astronomers. Absolute calibrations of the many magnitude systems (see below)
in common use often use the jansky or a similar SI-based or cgs-based unit.
Other than the generally very small measurements encountered in astronomy for
monochromatic flux density there would seem to be no good reason for not adopt-
ing the SI derived radiometric unit for spectral irradiance, the W . m−2 . Hz−1 .
As an example, the monochromatic flux density of a zero-magnitude A0V star
in the V waveband according to Bessell (2001) is 3636 Jy. Using the radiometric
unit for spectral irradiance, this is equivalent to 3.636×10−23 W . m−2 . Hz−1 or
3.636 d −23 W . m−2 . Hz−1 .
 8.2 Radiometry and photometry 115

Monochromatic flux density and total luminosity are the quantities most com-
monly used in astronomy. An in-depth description of these and other radiometric,
photometric and astrophysical terms is given by Sterken & Manfroid (1992).

Example: converting radiometric to photometric units

According to Sackman et al. (1993), the value for the solar constant (the total solar
irradiance just outside the Earth’s atmosphere) f , determined from the combined
observations made by several spacecraft, is:

f = 1 370 ± 2 W . m−2 (8.2)

The surface area of a sphere with a radius r, equal to the mean Sun–Earth
distance, is:

4 π r 2 = 4 π (1.496 0 × 1011 )2 = 2.812 4 × 1023 m2 (8.3)

The total solar radiant flux falling on this sphere is:

4 π r 2 . f = 3.853 0 × 1026 W (8.4)

(equivalent to a radiant intensity of r 2 . f = 3.066 1×1025 W . sr−1 ). To convert

from the radiant flux in watts to the luminous flux in lumens, simply multiply by
683 (the number of lumens in a watt):

683 × 4 π r 2 . f = 2.631 6 × 1029 lm (8.5)

The luminous intensity I , measured in candelas (the SI base unit), is the

luminous flux F per steradian:
F
I = = 2.094 1 × 1028 cd (8.6)
4π
Note that as the Sun is obviously not a point source at the distance of the Earth
and does not radiate uniformly in all directions nor uniformly as a function of time,
the value calculated for the luminous flux is only an approximation.

Example: stellar magnitude to candelas

Consider an A0V star of apparent magnitude mV = 0.0 at a distance of 10 pc
(3.086×1017 m) so that its absolute magnitude MV = 0.0 as well.
The surface area A of a sphere of radius 10 pc in units of square metres is:

A = 4 π r 2 = 4 π (3.086 × 1017 )2 = 1.1967 × 1036 m2 (8.7)

The empirical conversion factor from V magnitudes to janskys is (Bessell, 2001):

fV=0 = 3636 Jy = 3.636 × 10−23 W . m−2 . Hz−1 (8.8)

116 Unit of luminous intensity (candela)

Total monochromatic flux from the star is:

A . fV=0 = 3.636 × 10−23 × 1.1967 × 1036
= 4.3512 × 1013 W . Hz−1 (8.9)
Multiply by the V bandwidth ν , in frequency units (Dodd, 2007):
ν = 8.94 × 1013 Hz (8.10)
to give:
A . ν . fV=0 = 3.8516 × 1027 W (8.11)
In candelas:
3.8516 × 1027 × 683
IV=0 = = 2.1143 × 1029 cd (8.12)
4π

8.2.1 Common astronomical photometric units

It is perhaps a trifle misleading to use the adjective ‘common’ in the title of this
section as on occasion the unit is only ever used by the author of the paper or
catalogue to which reference is made. Some examples of the many different units
used are:
(i) for wavelengths: Å (angstrom unit = 10−10 m), 0.1 nm, nm, μm, mm, cm, and
m;
(ii) for frequency: Hz, kHz, MHz, GHz, THz, and PHz;
(iii) for monochromatic flux densities: erg . s−1 . cm−2 ; erg . s−1 . Å −1 ; erg . s−1 ;
magnitudes (many different systems); Jy; erg . Hz−1 . s−1 ; mW . m−2 .
(0.1 nm)−1 .
The many attempts to produce conversion factors from magnitudes to flux den-
sities have not been of great assistance either, as the quoted flux values generally
have their own units as well, e.g.:
UBV photometry: for a V = 0, A0V star, fV = 3636×10−30 W .cm−2 Hz−1
Vilnius photometry for a U = 0, OV star, fU = 19.22×10−12 W cm−2 μm−1
2MASS infrared photometry for a J = 0 star, fJ = 1592 ± 15.2 Jy
Given the plethora of units in use at present, astronomers must surely benefit
by adopting a single set to be used by all. In this book, SI photometric units are
abandoned in favour of SI derived radiometric units, as well as substituting these
units for magnitudes. Since the IAU approves the use of the jansky, which is essen-
tially a prefixed (by 10−26 ) SI derived unit (W . m−2 . Hz−1 ), all transformations
from other flux units and from magnitudes are made to janskys. A substitution for
Hipparchus/Pogson magnitudes using a ‘jansky’ magnitude ln(fν ), where fν is the
monochromatic flux in janskys, is also introduced.
 8.2 Radiometry and photometry 117

8.2.2 Wavelengths to frequencies

Consider a device (e.g., a glass filter in an optical telescope) that isolates a portion
of the electromagnetic spectrum of bandwidth λ (taken to mean the full width at
half maximum height (FWHM) centred on a wavelength λ). The following equation
is used to compute the frequency ν, equivalent to the wavelength λ, where c is the
velocity of light (= 299 792 458 m . s−1 ) in vacuo:
c
ν= (8.13)
λ
A filter with a bandwidth λ, defined by the FWHM, symmetrically placed
about the mean wavelength λ, does not have a symmetrical bandpass in frequency
space about the frequency ν, corresponding to the wavelength λ. To compute the
bandwidth and locate the upper and lower frequency bounds the following method
may be used:
c
νhfb = (8.14)
λ − 0.5 λ
c
νlfb = (8.15)
λ + 0.5 λ
ν = νhfb − νlfb (8.16)

where νhfb is the cut-off frequency at the high-frequency boundary of the FWHM
location of the filter and νlfb is the cut-off frequency at the low-frequency boundary
of the FWHM location of the filter. Figure 8.2 illustrates the differences between
bandwidths in wavelength and frequency spaces.

Example: convert from wavelength to frequency space for the Johnson V

band
Bessell (2001) gives the effective wavelength and FWHM bandwidth of the
Johnson–Cousins–Glass V band as: λ = 545 nm and λ = 85 nm. In metres, these
values become 5.45×10−7 and 8.8×10−8 . Substituting for λ in Equation (8.13)
gives:
299 792 458
ν= = 5.501 × 1014 Hz (8.17)
5.45 × 10−7
The short-wavelength end of the V-filter FWHM is at:
1
λ − λ = 5.025 × 10−7 m (8.18)
2
and the long-wavelength end is at:
1
λ + λ = 5.875 × 10−7 m (8.19)
2
118 Unit of luminous intensity (candela)

λ ν

SW LW HF LF

0.5 Δλ

Δλ Δν

Figure 8.1. Bandwidths in wavelength and frequency spaces. SW is short

wavelength, LW is long wavelength; HF is high frequency and LF is low frequency.

Compute the corresponding frequencies νhfb and νlfb using Equations (8.14) and
(8.15) to give νhfb = 5.966 0×1014 Hz and νlfb = 5.102 9×1014 Hz. The frequency
bandwidth is simply the difference between the two:
ν = 8.631 × 1013 Hz. (8.20)
Note that the wavelength FWHM is symmetrically placed with respect to the
mean wavelength but the frequency FWHM is not (see Figure 8.2).

8.2.3 The conversion of some commonly used photometric systems

from wavelength to frequency space
Using the method set out in the worked example above, effective wavelengths and
bandwidths in nanometres were converted into frequencies in hertz and listed in
Table 8.2. Also included in the table are the low- and high-frequency bounds of the
FWHM of each filter (or isolated spectral band) in hertz.
Wavelength data for the Johnson–Cousins–Glass photometry, Geneva photom-
etry, Strömgren photometry, Walraven photometry, DDO photometry, Washington
photometry, Thuan–Gunn photometry (all photoelectric systems); SDSS (Sloan
Digital Sky Survey) photometry, UBV photometry, MACHO and EROS photome-
try (gravitational microlensing observational programmes), which are all ground-
based CCD photometric systems; HST and HIPPARCOS/TYCHO photometry
(space-based photometric systems) were obtained from Bessell (2001).
The photographic I-band photometry of the USNO-B astrometric catalogue
was derived by Dodd (2007). The Spitzer infrared space telescope filter spec-
ifications are given in the online MIPS (Multiband Imaging Photometer for
 8.2 Radiometry and photometry 119

Table 8.2. Conversion of effective wavelengths and bandwidths from units of

length to frequency units

λ (nm) ν (Hz)
Filter λ (nm) ν (Hz) νlfb (Hz) νhfb (Hz)
Johnson Cousins Glass
U 367 8.169 d 14 7.495 d 14 8.976 d 14
66 1.481 d 14
B 436 6.876 d 14 6.207 d 14 7.707 d 14
94 1.500 d 14
V 545 5.501 d 14 5.103 d 14 5.966 d 14
85 8.632 d 13
R 638 4.699 d 14 4.175 d 14 5.373 d 14
160 1.197 d 14
I 797 3.762 d 14 3.440 d 14 4.149 d 14
149 7.094 d 13
J 1220 2.457 d 14 2.260 d 14 2.692 d 14
213 4.323 d 13
H 1630 1.839 d 14 1.681 d 14 2.030 d 14
307 3.495 d 13
K 2190 1.369 d 14 1.257 d 14 1.503 d 14
390 2.457 d 13
L 3450 8.690 d 13 8.133 d 13 9.328 d 13
472 1.194 d 13
M 4750 6.311 d 13 6.020 d 13 6.633 d 13
460 6.126 d 12
Geneva
U 350 8.565 d 14 8.027 d 14 9.182 d 14
47 1.155 d 14
B 424 7.071 d 14 6.489 d 14 7.767 d 14
76 1.278 d 14
B1 402 7.458 d 14 7.121 d 14 7.827 d 14
38 7.065 d 13
B2 448 6.692 d 14 6.399 d 14 7.013 d 14
41 6.137 d 13
V 551 5.441 d 14 5.129 d 14 5.793 d 14
67 6.640 d 13
V1 541 5.541 d 14 5.325 d 14 5.776 d 14
44 4.514 d 13
G 578 5.187 d 14 4.984 d 14 5.407 d 14
47 4.225 d 13
Strömgren
u 349 8.590 d 14 8.236 d 14 8.976 d 14
30 7.398 d 13
v 411 7.294 d 14 7.129 d 14 7.467 d 14
19 3.374 d 13
b 467 6.420 d 14 6.298 d 14 6.546 d 14
18 2.475 d 13
120 Unit of luminous intensity (candela)

Table 8.2. (cont.)

λ (nm) ν (Hz)
Filter λ (nm) ν (Hz) νlfb (Hz) νhfb (Hz)
y 547 5.481 d 14 5.368 d 14 5.598 d 14
23 2.306 d 13
βw 489 6.131 d 14 6.038 d 14 6.226 d 14
15 1.881 d 13
βn 486 6.169 d 14 6.150 d 14 6.188 d 14
3 3.808 d 12
Walraven
W 323.3 9.273 d 14 9.057 d 14 9.499 d 14
15.4 4.420 d 13
U 361.6 8.291 d 14 8.037 d 14 8.561 d 14
22.8 5.233 d 13
L 383.5 7.817 d 14 7.600 d 14 8.047 d 14
21.9 4.468 d 13
B 427.7 7.009 d 14 6.630 d 14 7.435 d 14
49.0 8.057 d 13
V 540.6 5.546 d 14 5.207 d 14 5.931 d 14
70.3 7.242 d 13
DDO
35 349.0 8.590 d 14 8.143 d 14 9.089 d 14
38.3 9.455 d 13
38 381.5 7.858 d 14 7.532 d 14 8.213 d 14
33.0 6.810 d 13
41 416.6 7.196 d 14 7.125 d 14 7.269 d 14
8.3 1.434 d 13
42 425.7 7.042 d 14 6.982 d 14 7.103 d 14
7.3 1.208 d 13
45 451.7 6.637 d 14 6.582 d 14 6.693 d 14
7.6 1.117 d 13
48 488.6 6.136 d 14 6.021 d 14 6.255 d 14
18.6 2.337 d 13
Thuan Gunn
u 353 8.493 d 14 8.037 d 14 9.003 d 14
40 9.654 d 13
v 398 7.532 d 14 7.172 d 14 7.931 d 14
40 7.589 d 13
g 493 6.081 d 14 5.678 d 14 6.546 d 14
70 8.678 d 13
r 655 4.577 d 14 4.283 d 14 4.915 d 14
90 6.319 d 13
Vilnius
U 345 8.690 d 14 8.213 d 14 9.224 d 14
40 1.011 d 14
P 374 8.016 d 14 7.747 d 14 8.305 d 14
26 5.579 d 13
 8.2 Radiometry and photometry 121

Table 8.2. (cont.)

λ (nm) ν (Hz)
Filter λ (nm) ν (Hz) νlfb (Hz) νhfb (Hz)
X 405 7.402 d 14 7.207 d 14 7.609 d 14
22 4.024 d 13
Y 466 6.433 d 14 6.259 d 14 6.618 d 14
26 3.592 d 13
Z 516 5.810 d 14 5.694 d 14 5.931 d 14
21 2.365 d 13
V 544 5.511 d 14 5.382 d 14 5.646 d 14
26 2.635 d 13
S 656 4.570 d 14 4.501 d 14 4.641 d 14
20 1.394 d 13
2MASS
J 1235 2.427 d 14 2.278 d 14 2.598 d 14
162 3.198 d 13
H 1662 1.804 d 14 1.677 d 14 1.951 d 14
251 2.740 d 13
K 2159 1.389 d 14 1.309 d 14 1.478 d 14
261 1.685 d 13
SDSS A0 star
u 356 8.421 d 14 7.727 d 14 9.253 d 14
64 1.526 d 14
g 475 6.311 d 14 5.526 d 14 7.357 d 14
135 1.831 d 14
r 620 4.835 d 14 4.354 d 14 5.436 d 14
137 1.082 d 14
i 761 3.939 d 14 3.577 d 14 4.383 d 14
154 8.055 d 13
z 907 3.305 d 14 3.058 d 14 3.597 d 14
147 5.392 d 13
Washington
C 391 7.667 d 14 6.722 d 14 8.922 d 14
110 2.201 d 14
M 509 5.890 d 14 5.339 d 14 6.567 d 14
105 1.228 d 14
T1 633 4.736 d 14 4.455 d 14 5.056 d 14
80 6.010 d 13
T2 805 3.724 d 14 3.407 d 14 4.107 d 14
150 7.000 d 13
ANS
ANS1 154.5 1.940 d 15 1.910 d 15 1.972 d 15
5.0 6.281 d 13
ANS2 154.9 1.935 d 15 1.847 d 15 2.033 d 15
14.9 1.866 d 14
ANS3 179.9 1.666 d 15 1.600 d 15 1.738 d 15
14.9 1.383 d 14
122 Unit of luminous intensity (candela)

Table 8.2. (cont.)

λ (nm) ν (Hz)
Filter λ (nm) ν (Hz) νlfb (Hz) νhfb (Hz)
ANS4 220.0 1.363 d 15 1.303 d 15 1.428 d 15
20.0 1.241 d 14
ANS5 249.3 1.203 d 15 1.167 d 15 1.240 d 15
15.0 7.242 d 13
ANS6 329.4 9.101 d 14 8.965 d 14 9.241 d 14
10.0 2.764 d 13
TD1
TD1 156.5 1.916 d 15 1.733 d 15 2.141 d 15
33 4.085 d 14
TD2 196.5 1.526 d 15 1.407 d 15 1.666 d 15
33 2.580 d 14
TD3 236.5 1.268 d 15 1.185 d 15 1.363 d 15
33 1.777 d 14
TD4 274.0 1.094 d 15 1.036 d 15 1.160 d 15
31 1.242 d 14
IRAS
12 μm 12 000 2.498 d 13 1.934 d 13 3.527 d 13
7 000 1.593 d 13
25 μm 25 000 1.199 d 13 9.805 d 12 1.543 d 13
11 150 5.628 d 12
60 μm 60 000 4.997 d 12 3.932 d 12 6.852 d 12
32 500 2.921 d 12
100 μm 100 000 2.998 d 12 2.590 d 12 3.558 d 12
31 500 9.684 d 11
HST
HST336 334 8.976 d 14 8.386 d 14 9.655 d 14
47 1.269 d 14
HST439 430 6.972 d 14 6.440 d 14 7.599 d 14
71 1.159 d 14
HST450 451 6.647 d 14 5.942 d 14 7.542 d 14
107 1.600 d 14
HST555 532 5.635 d 14 4.951 d 14 6.539 d 14
147 1.587 d 14
HST675 667 4.495 d 14 4.104 d 14 4.968 d 14
127 8.636 d 13
HST814 788 3.804 d 14 3.480 d 14 4.196 d 14
147 7.159 d 13
HIPPARCOS TYCHO
BT 421 7.121 d 14 6.574 d 14 7.767 d 14
70 1.192 d 14
VT 526 5.699 d 14 5.205 d 14 6.298 d 14
100 1.093 d 14
HP 517 5.799 d 14 4.744 d 14 7.458 d 14
230 2.714 d 14
 8.2 Radiometry and photometry 123

Table 8.2. (cont.)

λ (nm) ν (Hz)
Filter λ (nm) ν (Hz) νlfb (Hz) νhfb (Hz)
MACHO
B 519 5.776 d 14 5.073 d 14 6.707 d 14
144 1.634 d 14
R 682 4.396 d 14 3.888 d 14 5.056 d 14
178 1.167 d 14
UBV(CCD)
B 436 6.876 d 14 6.207 d 14 7.707 d 14
94 1.500 d 14
V 545 5.501 d 14 5.103 d 14 5.966 d 14
85 8.632 d 13
R 641 4.677 d 14 4.158 d 14 5.344 d 14
160 1.186 d 14
I 791 3.790 d 14 3.476 d 14 4.167 d 14
143 6.908 d 13
Z 909 3.298 d 14 3.133 d 14 3.482 d 14
96 3.493 d 13
USNO-B(I)
I 807.5 3.713 d 14 3.331 d 14 4.193 d 14
185 8.619 d 13
EROS
BE1 485 6.181 d 14 5.557 d 14 6.964 d 14
109 1.407 d 14
BE2 539 5.562 d 14 4.729 d 14 6.752 d 14
190 2.023 d 14
RE1 657 4.563 d 14 3.984 d 14 5.339 d 14
191 1.355 d 14
RE2 767 3.909 d 14 3.342 d 14 4.706 d 14
260 1.364 d 14
Spitzer
24 μm 23700 1.265 d 13 1.151 d 13 1.404 d 13
4700 2.533 d 12
70 μm 71000 4.222 d 12 3.724 d 12 4.875 d 12
19000 1.151 d 12
160 μm 156000 1.922 d 12 1.728 d 12 2.165 d 12
35000 4.367 d 11
DENIS
i 791 3.790 d 14 3.476 d 14 4.167 d 14
143 6.910 d 13
J 1228 2.441 d 14 2.265 d 14 2.647 d 14
191 3.820 d 13
KS 2145 1.398 d 14 1.306 d 14 1.503 d 14
302 1.970 d 13

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .
124 Unit of luminous intensity (candela)

Spitzer) instrument handbook.49 The ultraviolet photometer and filters used by

the ANS (Astronomical Netherlands Satellite) are described by van Duinen et al.
(1975). A description of the equipment used by the TD1 photometric satellite
is given in the introduction of the Catalogue of Ultraviolet Stellar Fluxes by
Thompson et al. (1978).
Detailed information on the Vilnius seven-colour photometry was obtained from
Straižys (1992). The absolute calibration of the 2MASS infrared photometric system
and the filter parameters are given by Cohen et al. (2003), and finally data on the
Infrared Astronomical Satellite (IRAS) is given on its website.50

8.2.4 Determination of the spectral irradiance and the monochromatic flux

density of stars from a selection of published values in non-SI units
In the introduction to the third edition of a Catalog of Infrared Observations, Gezari
et al. (1993) list 26 different ways in which the infrared stellar monochromatic
flux density of stars are given in the catalogue. Some of the units are cgs based,
some are SI based, some in wavelength space, some in frequency space, some are
magnitudes, some magnitudes per square arcsecond and some are derived from
spectrophotometric measures. Inter-comparison between the different units is not
easy and often prone to mistakes.
Various specific examples are set out below, beginning with monochromatic flux
densities given in non-SI and ending with the conversion of the most common unit
of stellar brightness unit, the magnitude, into janskys and the radiometric spectral
irradiance unit W . m−2 . Hz−1 .

Conversion of apparent monochromatic flux density from wavelength

space to frequency space
Many of the earlier monochromatic flux density units are in wavelength space
and so must be converted into frequency space to produce measures in janskys
or W . m−2 . Hz−1 . If fλ is the apparent monochromatic flux density in a wave-
length band λ wide, centred on wavelength λ, both measured in metres, and ν
is the bandwidth in hertz, then fν , the apparent monochromatic flux density in
W . m−2 . Hz−1 at frequency ν (derived from Equation 8.13), is given by:

λ
fν = fλ . (8.21)
ν

49 See http://www.irsa.ipac.caltech.edu/data/SPITZER/docs/mips/mipsinstrumenthandbook/
50 See http://irsa.ipac.caltech.edu/IRASdocs/
 8.2 Radiometry and photometry 125

where ν is determined using Equation (8.16). In janskys, the expression is simply:

λ
fν = fλ . . 1026 (8.22)
ν

Conversion of magnitudes to monochromatic flux density

The majority of catalogues of the brightness of stars from the ultraviolet to the near
infrared use magnitudes as a measure of that brightness. The scale of magnitude is
logarithmic, with a base of 2.512 = 100.4 (i.e., the fifth root of 100). The magnitude
system is ordinal, so that the brighter the object the smaller the magnitude: first
magnitude objects are brighter than second magnitude objects and so on. The system
was devised by the Greek astronomer Hipparchus over 2000 years ago, and made
more respectable by the English astronomer Pogson in the nineteenth century when
he fixed the brightness ratio of 5 magnitudes to equal a factor of 100.
The apparent magnitude, mλ , of a celestial body at effective wavelength λ,
through a filter of bandwidth λ, may be defined in terms of the monochromatic
flux density as:
mλ = −2.5 log fλ (8.23)
If this Equation (8.23) is inverted and converted to a natural logarithm base, rather
than base 10, then:
fλ = f(mλ =0) e−0.921 mλ (8.24)
where fmλ =0 is the measured monochromatic flux density in a bandwidth centred
on wavelength λ of a star of magnitude mλ =0.

Conversions of some published non-SI flux units to janskys

Example 1
Bessell (2001) gives the absolute calibration of the Johnson–Cousins–Glass
broadband photometry in units of 10−30 . W . cm−2 . Hz−1 . Now

1 Jy = 10−26 W . m−2 . Hz−1

10−30 W . cm−2 . Hz−1 = 10−30 W . (0.01 . m)−2 . Hz−1
= 10−30 . 104 . W . m−2 . Hz−1
= 10−26 W . m−2 . Hz−1
= 1 Jy (8.25)

So the units used by Bessell were essentially janskys expressed in a hybrid cgs and
SI system.
126 Unit of luminous intensity (candela)

Example 2
Convert the Vilnius U-band monochromatic flux density fλ(U =0) (Straižys, 1992)
expressed in units of W . cm−2 . μm−1 into janskys, for a hypothetical OV-type
star.

fλ(U=0) = 19.22 × 10−12 W . cm−2 . μm−1

= 19.22 × 10−12 W . (10−2 . m)−2 . (10−6 . m)−1
= 19.22 × 10−12 × 104 × 106 W . m−2 . m−1
= 19.22 × 10−2 W . m−2 . m−1 (8.26)

From Table 8.2, extract the appropriate values for bandwidth in wavelength ( λU )
and frequency ( νU ) space and substitute these values into Equation (8.22) to
determine the value of the monochromatic flux density in janskys:

λU = 4 × 10−8 m νU = 1.011 × 1014 Hz

4 × 10−8
fν(U) = 19.22 × 10−2 . . 1026
1.011 × 1014
= 7.604 × 103 Jy (8.27)

In radiometric units, the spectral irradiance is 7.604×10−23 W . m−2 . Hz−1 .

Example 3
Convert the TD1-S68 ultraviolet satellite measure at λ = 274 nm (Thompson et al.,
1978) expressed in erg . cm−2 . s−1 . Å−1 into janskys for the A0V star Vega.
−1
f274 = 3.123 × 10−9 erg . cm−2 . s−1 . Å
= 3.123 × 10−9 (10−7 J) . s−1 . (10−2 m)−2 . (10−10 m)−1
= 3.123 × 10−9 × 10−7 × 104 × 1010 J . s−1 . m−2 . m−1
= 3.123 × 10−2 W . m−2 . m−1 (8.28)

From Table 8.2, λ274 = 3.1 × 10−8 and ν274 = 1.242×1014 Hz, so, using
Equation (8.22):

3.1 × 10−8
fν274 = 3.123 × 10−2 . . 1026 Jy
1.242 × 1014
= 779.5 Jy (8.29)

In radiometric units the spectral irradiance is 7.795×10−24 W . m−2 . Hz−1 .

 8.2 Radiometry and photometry 127

Example 4
HST spectrophotometry of the star Vega from 170 nm to 1010 nm. Bohlin &
Gilliland (2004) used observations obtained with the Space Telescope Imaging
Spectrograph (STIS) to determine the flux density in units of mW . m−2 . (0.1 nm)−1
(where 0.1 nm = 1 Å) against wavelength in angstroms. A plot of their data con-
verted to SI units (i.e., wavelength in metres and monochromatic flux densities in
W . m−2 . m−1 ) is shown in Figure 8.2.
The conversion from Bohlin & Gilliland (2004) units to janskys is carried out
for the flux measurement of 3.04×10−9 mW . m−2 . (0.1 nm)−1 determined at a
wavelength of 5800 Å and an FWHM measurement of 11.6 Å:

fλ = 3.04 × 10−9 mW . m−2 . (0.1 nm)−1

= 3.04 × 10−9 . (10−3 . W) . m−2 . (10−10 . m)−1
= 3.04 × 10−9 . (10−3 . 1010 ) . W . m−2 . m−1
= 3.04 × 10−2 W . m−2 . m−1 (8.30)

To convert a wavelength-space monochromatic flux density fλ , at wavelength

λ, into a frequency-space monochromatic flux density fν , at a frequency ν, cor-
responding to λ, Equation (8.22) is used. Bohlin & Gilliland (2004) list values of
the FWHM bandwidth at each central wavelength sampled, which is taken as λ
and ν then calculated using Equations (8.13) to (8.16). Finally, a value for the
monochromatic flux (W . m–2 . m–1)

0.08

0.06

0.04

0.02

0.00

2 3 4 5 6710–6 2 3 4 5 67 –5 2 3 4 5 6710–4 2 3 4
10–7 10
wavelength (m)

Figure 8.2. HST spectrophotometry of Vega in wavelength space. Wavelength

is measured in metres (logarithmic scale) and monochromatic flux density in
radiometric units (W . m−2 . m−1 ).
128 Unit of luminous intensity (candela)

frequency monochromatic flux density is derived from Equation (8.22). For the
numerical example, entry 2436 in the Bohlin & Gilliland (2004) tabulation is used
and firstly converted to SI units:
c = 299 792 458 m . s−1
λ = 5.80 × 10−7 m
λ = 1.16 × 10−9 m (8.31)
Convert from wavelength to frequency:
c
ν = = 5.169 × 1014 Hz
λ
c
νhfb = = 5.174 × 1014 Hz
λ − 0.5 λ
c
νlfb = = 5.164 × 1014 Hz
λ + 0.5 λ
ν = νhfb − νlfb = 1.034 × 1012 Hz
λ
fν = fλ . 1026 = 3415 Jy (8.32)
ν
As an additional example that utilizes the Bohlin & Gilliland (2004) data, a subset
was extracted bounded by the bandwidth of the Johnson V filter. In total, some 157
data points are available within the V bandwidth in the HST Vega data. As above,
(λ, λ, fλ ) were converted to (ν, ν, fν ) and a plot drawn (Figure 8.3) of ν against
fν . The mean monochromatic flux density f(ν=Vν ) , within the V frequency band,
is the sum of the product of the individual monochromatic flux densities multiplied
by the FWHM of each measured frequency sample (ν ν) then divided by the sum
of the FWHM values, thus:
157
(fνk ) . ( νk )
f(ν=Vν ) = k=1 157 = 3646 Jy (8.33)
k=1 ( νk )
For the Bohlin & Gilliland (2004) data, the value of f(ν=Vν ) is 3646 Jy, which
may be compared with the Bessell (2001) value of 3636 Jy for a standard A0V star
of magnitude V = 0.00 (note that Bohlin & Gilliland, 2004 determined a value of
0.026 for the V magnitude of Vega).

Radiometric calibration of selected photometric systems

Figure 8.5 combines the Hubble Space Telescope spectrophotometry of the star
Vega with the photometry of nine different photometric systems, both ground and
space based, converted in each case to janskys. The spectral irradiance in the radio-
metric unit W . m−2 . Hz−1 is obtained from the monochromatic flux density in
 8.2 Radiometry and photometry 129

3900

νeff
monochromatic flux density (Jy)

3700

νlfb V band νhfb

3500

ga
Ve

3300
5.0 5.2 5.4 5.6 5.8 6.0
frequency (x 1014Hz)

Figure 8.3. V band from HST spectrophotometry of Vega in frequency space.

Frequency is measured in hertz and monochromatic flux density in janskys
(10−26 . W . m−2 . Hz−1 ), νeff is the frequency equivalent to the effective wave-
length λeff , the squares mark the extreme ends of the FWHM points of the
bandwidth of the Johnson V filter and are labelled νlfb for the low-frequency
bound and νhfb for the high-frequency bound.

janskys by multiplying by 10−26 . In all, some 47 individual photometric points

are plotted with their frequency FWHM bandwidth shown as a horizontal error
bar. Some of the monochromatic flux densities were taken directly from published
values, some by converting cgs or hybrid units into janskys and some by trans-
forming system magnitudes into janskys. Notes on the methods used in each case
are given below. Table 8.3 sets out for each of the identified filters, the filter set to
which it belongs and its identifier within that set, the effective frequency in hertz,
the frequency bandwidth (FWHM) in hertz and the monochromatic flux density in
janskys. The table is arranged in order of increasing frequency, beginning in the
infrared and ending in the ultraviolet part of the spectrum.

8.2.5 Low-frequency, high-frequency and wavenumber unit conversions

Low frequency: The low-frequency part of the electromagnetic spectrum is the
domain of radio astronomy that covers frequencies from less than 30 MHz to
130 Unit of luminous intensity (candela)

5000

4000
monochromatic flux density (Jy)

3000

2000

1000

67 2 3 4 5 67 2 3 4 5 67 2 3 4 5 67 2 3 4 5 6
1012 1013 1014 1015
frequency (Hz)

Figure 8.4. HST spectrophotometry of Vega in frequency space. Frequency is

measured in hertz (logarithmic scale) and monochromatic flux density in jan-
skys (10−26 . W . m−2 . m−1 ). Note that in this figure the photon energy increases
from left to right in contrast with the wavelength figure, where the highest-energy
photons are at the left-hand side of the plot.

30 GHz over which ground-based observations may be made. Astronomical sources

cannot be observed from the ground at frequencies much below 30 MHz due to the
reflective properties of the ionosphere. At frequencies greater than 30 GHz, molec-
ular absorption by the lower atmosphere restricts observations to discrete windows.
The wavelength range corresponding to these frequencies is approximately 1 cm to
a few tens of metres.
In comparison with optical astronomy, radio astronomy is a relatively new sci-
ence and was established by scientists and engineers with expertise in fields such
as electrical engineering and physics rather than astronomy. As such it developed
independently its own units which, fortunately, have tended to be based on the
physics of the day.
To specify the part of the electromagnetic spectrum that is being observed, radio
astronomers use frequency measured in Hz.
The unit of spectral irradiance or monochromatic flux density (termed specific
flux or flux density by radio astronomers (Burke & Graham-Smith, 2002)) is the
 8.2 Radiometry and photometry 131

Table 8.3. Monochromatic flux densities, either measured observationally for the
A0V star Vega or determined by other means for a standard A0V star

Filter Effective Frequency MFD

identifier frequency (Hz) bandwidth (Hz) FWHM (Jy)
iras 100 μm 2.998 d 12 9.684 d 11 0.39
iras 60 μm 4.997 d 12 2.921 d 12 1.09
iras 25 μm 1.199 d 13 5.625 d 12 6.29
iras 12 μm 2.498 d 13 1.593 d 13 26.97
jcg M 6.311 d 13 6.162 d 12 154.00
jcg L 8.690 d 13 1.194 d 13 285.00
jcg K 1.369 d 14 2.457 d 13 640.00
2mass KS 1.389 d 14 1.690 d 13 666.70
den KS 1.398 d 14 1.970 d 13 665.00
2mass H 1.804 d 14 2.740 d 13 1024.00
jcg H 1.839 d 14 3.495 d 13 1020.00
2mass J 2.427 d 14 3.200 d 13 1594.00
den J 2.441 d 14 3.820 d 13 1595.00
jcg J 2.457 d 14 4.323 d 13 1589.00
jcg I 3.762 d 14 7.094 d 13 2416.00
den i 3.790 d 14 6.910 d 13 2499.00
vil S 4.570 d 14 1.394 d 13 3142.04
jcg R 4.699 d 14 1.197 d 14 3064.00
gen G 5.187 d 14 4.225 d 13 3321.24
gen V 5.441 d 14 6.640 d 13 3588.39
strom y 5.481 d 14 2.306 d 13 3690.37
jcg V 5.501 d 14 8.632 d 13 3636.00
vil V 5.511 d 14 2.635 d 13 3749.53
gen V1 5.541 d 14 4.514 d 13 3548.95
vil Z 5.810 d 14 2.365 d 13 3951.37
strom v 6.420 d 14 2.475 d 13 4225.45
vil Y 6.433 d 14 3.592 d 13 4169.27
gen B2 6.692 d 14 6.137 d 13 4131.43
jcg B 6.876 d 14 1.500 d 14 4063.00
gen B 7.071 d 14 1.278 d 14 3634.96
strom b 7.294 d 14 3.374 d 13 4043.27
vil X 7.402 d 14 4.023 d 13 4105.86
gen B1 7.458 d 14 7.065 d 13 3927.35
vil P 8.016 d 14 5.579 d 13 2111.13
jcg U 8.169 d 14 1.481 d 14 1790.00
gen U 8.565 d 14 1.155 d 14 1317.35
strom u 8.590 d 14 7.398 d 13 1317.92
vil U 8.690 d 14 1.011 d 14 1321.46
ans 6 9.101 d 14 7.242 d 13 1102.00
td1 4 1.094 d 15 1.242 d 14 779.49
ans 5 1.203 d 15 1.241 d 14 720.83
td1 3 1.268 d 15 1.777 d 14 687.11
td1 2 1.526 d 15 2.580 d 14 630.33
132 Unit of luminous intensity (candela)

Table 8.3. (cont.)

Filter Effective Frequency MFD

identifier frequency (Hz) bandwidth (Hz) FWHM (Jy)
ans 3 1.666 d 15 1.384 d 14 598.64
td1 1 1.916 d 15 4.085 d 14 459.58
ans 2 1.935 d 15 1.866 d 14 435.19
ans 1 1.940 d 15 6.381 d 13 447.18

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .
Sources for Table 8.3

1. IRAS (Infrared Astronomical Satellite) photometry (iras) at 12 μm, 25 μm, 60 μm and 100 μm for the star
Vega (Cohen et al., 1992).51
2. Johnson–Cousins–Glass (jcg). Monochromatic flux densities for the nine bands (U, B, V, R, I, J, H, K, L, M)
are those given by Bessell (2001) for an A0V star with magnitude V = 0.00.
3. The 2 Micron All Sky Survey (2MASS) measured monochromatic flux densities in three infrared bands (J,
H, KS ). Cohen et al. (2003) provided the absolute calibration for a zero-magnitude A0V star.
4. DENIS (den). Fouqué et al. (2000) used a synthetic spectrum of Vega to derive monochromatic flux densities
at magnitude zero for the three bands (i, J, KS ).
5. The Vilnius (vil) seven-colour (U, P, X, Y, Z, V, S) absolute monochromatic flux densities given by Straižys
(1992) in W . cm−2 . μm−1 units for an A0V star were converted to janskys using the transformation given
in example 2 above.
6. Geneva (gen) seven-colour (U, B, B1, B2, V, V1, G) photometric bandwidths with their absolute calibration
for Vega were determined by Rufener & Nicolet (1988), who used the relationship:

log(Eν ) = −0.4 (mν − Kν ) + C (8.34)

where Eν is the spectral irradiance at frequency ν in W . m−2 . Hz−1 , mν is the measured magnitude at
frequency ν, Kν is the colour index of the Geneva band at frequency ν relative to that at the effective
frequency of the B filter, and C is the zero-point shift for the B filter.
7. Strömgren (strom) photometry absolute calibrations of monochromatic irradiance are given by Sterken &
Manfroid (1992) in units of 10−11 W . m−2 . nm−1 for a star of spectral type A0V and magnitude V = 0.00.
Example 2 above shows how a similar unit is converted to janskys.
8. Astronomical Netherlands Satellite (ans) photometry. Wesselius et al. (1980) related magnitudes to fluxes in
W . m−2 . nm−1 (fλ ) using:
 
26.1+mλ
−2.5 +9
fλ = 10 (8.35)

which may be converted to janskys using Equation (8.22). From the ultraviolet magnitudes listed by Wes-
selius et al. (1982), the following magnitudes were extracted for Vega; ans1 = −0.491, ans2 = −0.441,
ans3 = −0.462, ans4 = no measurement, ans5 = 0.046, ans6 = 0.191. Substituting these magnitudes into
Equation (8.35) give values for fλ , which may then be converted to fν in janskys via Equation (8.22).
These monochromatic flux density values are shown in Table 8.3.
9. The Thor–Delta (td1) ultraviolet catalogue (Thompson et al., 1978) the wavelength-space monochromatic
flux densities of Vega in units of 10−10 erg . cm−2 . s−1 . Å−1 , which were converted to janskys in Table 8.3
as per example 3 above.

51 See also http://lambda.gsfc.nasa.gov/product/iras/docs/exp.sup/ch6/C2a.html

 8.2 Radiometry and photometry 133

5000

4000
monochromatic flux density (Jy)

3000

2000

1000

67 2 3 4 5 67 2 3 4 5 67 2 3 4 5 67 2 3 4 5 6
1012 1013 1014 1015
frequency (Hz)

Figure 8.5. HST spectrophotometry of Vega overlaid with broad- and

intermediate-band photometry. The points are at the location of the effective fre-
quency of the filter, whose monochromatic flux density is plotted with the error
bars showing the FWHM bandwidth. The low-frequency end of the spectrum
in the infrared shows the location of the IRAS satellite photometric bands, the
near infrared has 2MASS, DENIS and Johnson–Cousins–Glass photometry and
bands, the optical part of the spectrum is overlaid with Vilnius, Geneva, Strömgren,
and Johnson–Cousins–Glass photometric bands and the ultraviolet spectrum has
measures and bandwidths from the TD1 and ANS satellites.

jansky, appropriately named after one of the pioneers of radio astronomy, Karl
Jansky.
The monochromatic flux density per unit solid angle (Jy . rad−1 ) is called the
brightness of the (extended) source.
In short, radio astronomers already use derived SI units that meet with the
approval of the IAU.
By way of an example, an attempt was made by Hollis et al. (1985) to detect Vega
at radio frequencies using the VLA (Very Large Array) of the US National Radio
Astronomy Observatory at a frequency of 4.86 GHz (λ ∼ 6 cm) with a beamwidth
of 40 μrad by 17 μrad (8.5 arcsec × 3.5 arcsec). No signal significantly greater than
the background noise (30 μJy) was detected.
134 Unit of luminous intensity (candela)

Millimetre and submillimetre frequencies: Between the optical infrared frequen-

cies and the high radio astronomical frequencies is to be found the millimetre
and submillimetre region. It covers a band of frequencies from about 1014 Hz
(λ ∼ 3 mm) to about 1015 Hz (λ ∼ 300 μm). This region is transitional between
optical and radio astronomy and, as such, uses units from both disciplines. So
wavelengths are more often used than frequencies, though janskys are generally
preferred to magnitudes.
Example: the star Vega was observed with the James Clerk Maxwell telescope by
Holland et al. (1998). At a mean frequency of approximately 3.5×1011 Hz, a signal
strength averaging around 46 mJy was observed in a beam of diameter 48.5 μrad

(∼10 diameter).
High frequencies, X-rays: The ultraviolet and extreme ultraviolet part of the elec-
tromagnetic spectrum covers the range in frequencies from approximately 1015 Hz
to 5×1016 Hz. At still higher frequencies, between 5×1016 Hz and 2×1020 Hz, the
soft and hard X-rays are to be found and beyond that to 1026 Hz, γ -rays (Culhane
& Sanford, 1981).
It is conventional for both X-ray and γ -ray astronomers to use electron volts Eν ,
rather than frequency ν.
The electron volt is classed as a non-SI unit accepted for use with the international
system, whose value in SI units is obtained by experiment.
The electron volt has the dimension of energy [L]2 . [M ] . [T ]−2 , may be
expressed in terms of base SI units as m2 . kg . s−2 and has symbol eV.
The relationship between electron volts and hertz is linear:
Eν E0
ν= =n (8.36)
h h
where h is Planck’s constant and equal to 6.626 075 5×1034 J . s (Cox, 2000), E0
is the energy equivalent of 1 eV in joules (1.602 177×10−19 J) and n is the number
of electron volts.

Example 1: convert 1 eV to Hz
1.602 177 × 10−19
ν = 1× = 2.418 × 1014 Hz (8.37)
6.626 076 × 10−34
Example 2: convert 1 keV, 1 MeV and 1 Gev to Hz
ν(1 keV) = 2.418 × 1017 Hz; ν(1 MeV) = 2.418 × 1020 Hz;
ν(1 GeV) = 2.418 × 1023 Hz (8.38)
where 1 keV = 1000 eV, 1 MeV = 106 eV and 1 GeV = 109 eV.
 8.2 Radiometry and photometry 135

A typical observational X-ray spectrum has photon energy in keV as the abscissa
plotted against the monochromatic flux density in units of photons . cm−2 . keV−1
as ordinate.

Example 3: convert photons . cm−2 . keV−1 to W . m−2 . Hz−1 and janskys

1 ph . cm−2 . s−1 . keV−1 = 1 ph . (10−2 m)−2 . s−1 . (2.418 × 1017 Hz)−1

= (104 × (2.418 × 1017 )−1 ) ph . m−2 . s−1 . Hz−1
= 4.136 × 10−14 ph . m−2 . s−1 . Hz−1 (8.39)

Now the energy of a 1 keV photon is equivalent to 1.602 177×10−16 J, hence:

1 ph . cm−2 . s−1 . keV−1 = (1.602 177 × 10−16 J)(4.136 × 10−14 ) m−2 . s−1 . Hz−1
= 6.6266 × 10−30 J . s−1 . m−2 . Hz−1
= 6.6266 × 10−30 W . m−2 . Hz−1 (8.40)
= 6.6266 × 10−4 Jy (8.41)

X-ray observations of Vega

An attempt to extend observational measurements of the electromagnetic spectrum
of Vega to frequencies higher than the extreme ultraviolet was made by Pease et al.
(2006) using the high-resolution camera (HRC-1) of the CHANDRA X-ray obser-
vatory. Exposure times of up to 25 900 s produced total counts of around 10 photon
events over the 156-pixel array used. The energy range of the detector extended from
0.08 keV to 10.0 keV.52 The pixel size is given as 6.4×6.4 μm2 , so the total detec-
tor area is 6.390×10−5 cm−2 . The rate of X-ray photons incident upon the detector
was 10/2599 = 3.861×10−4 s−1 and the X-ray bandwidth taken as approximately
10 keV. So the monochromatic flux density, fX , in ph . cm−2 . s−1 . keV−1 is:
3.861 × 10−4
fX =  0.6 ph . cm−2 . s−1 . keV−1 (8.42)
6.39 × 10−5 × 10
which, by using the conversion in Equation (8.41), is equivalent to 4.0×10−4 Jy.
The centre of the measured X-ray band was taken to be 5 keV or 1.209×1018 Hz.
High frequencies: γ -rays are photons with the highest detectable energies.
Ground-based instruments, such as the HESS (High Energy Stereoscopic Sys-
tem) array in Namibia,53 were designed to detect 1 TeV (1012 eV ≡ 2.416×1026 Hz)

52 See http://cxc.harvard.edu/proposer/POG/html/chap7.html page 2 of 19

53 See http://www.saao.ac.za/∼wgssa/as5/steenkamp.html
136 Unit of luminous intensity (candela)

photons with a monochromatic flux density as low as 10−12 erg . s−1 . cm−2
(≡ 10−15 W . m−2 ) indirectly by means of the Čerenkov radiation emitted by the
γ -ray’s passage through the Earth’s atmosphere. The use of several telescopes in the
array allows the direction from which the γ -ray photon originated to be estimated.
For lower-energy γ -ray photons, detectors are commonly used with high-altitude
balloons, rockets and satellites. The most recent project is the Fermi γ -ray space
telescope mission launched by NASA in 2008. The principal instrument on this
satellite is the Large Area Telescope (Atwood et al., 2009), which covers the energy
range from 20 MeV to 300 GeV, corresponding to frequencies from 4.8×1021 Hz
to 7.5×1025 Hz.
In the first catalogue of active galactic nuclei detected by the Fermi Large Area
Telescope, Abdo et al. (2010), in common with most γ -ray astronomers, used
photons . cm−2 . s−1 as a unit for flux density, with the electron volt being used
instead of frequency in hertz. Atwood et al. (2009) used a mixture of units for flux
density, such as photons . cm−2 . s−1 and particles . m−1 . s−1 .
The unit transformations used in the section on X-ray astronomy apply equally
well to the γ -ray units.

Extended flux-density plot for Vega

The extra values, or upper limits, for monochromatic flux densities found for radio,
millimetre, submillimetre, X-rays and γ -rays have been added to the data used to
construct Figure 8.5 to produce Figure 8.6. To accommodate the larger range of
monochromatic flux densities in the enhanced data set, a logarithmic scale is used
for the ordinate.

Wavenumbers of electromagnetic radiation may be defined as the number of

electromagnetic waves per unit length.
The SI unit for wavenumber is the inverse metre with dimension [L]−1 and
symbol m−1 .

In infrared astronomy a cgs wavenumber unit commonly found is the kayser or

inverse centimetre (cm−1 ). Occasionally, the inverse micrometre, μm, is used. The
IAU recommends only the use of the m−1 . The relationship between frequency ν,
wavelength λ, and wavenumber ν, is given by Browning (1969) as:

1 ν
ν= = (8.43)
λ c
 8.3 Magnitudes 137

103

102
monochromatic flux density (Jy)

101

100

10–1

10–2

10–3 submillimetre X-ray

10–4 radio

10–5

10–6

109 1010 1011 1012 1013 1014 1015 1016 1017 1018
frequency (Hz)

Figure 8.6. A log-log plot of HST spectrophotometry of Vega overlaid with broad-
and intermediate-band ultraviolet, optical and infrared photometry, with diamond
symbols for measures or attempted measures in the radio, submillimetre and X-ray
bands. The radio and X-ray plots are upper limits to the flux at that frequency.

Some sample conversions: Consider the rest wavelength, λα = 6562.817 Å

(Moore, 1959) of the hydrogen Hα line of the Balmer series:

6 562.817 Å = 6.562 817 × 10−7 m

= 6.562 817 × 10−5 cm
1
νm = m−1 = 1.523 736 × 106 m−1 (8.44)
6.562 817 × 10−7
1
ν cm = cm−1 = 15 237 cm−1 (8.45)
6.562 817 × 10−5

8.3 Magnitudes
A very brief history (for more detail, see Hearnshaw 1996) of the use of magnitudes
in describing the brightness of stars is given in Chapter 12 and the present chapter
includes numerous examples of different magnitude schemes used over the past
century. Magnitudes are thoroughly engrained in the minds of astronomers, both
138 Unit of luminous intensity (candela)

professional and amateur. What are the advantages and disadvantages of using
magnitudes? Is there a better way that would link magnitudes more strongly and
logically to physical measurements?

The advantages of magnitudes

1. Magnitudes, being logarithmic, are able to cover, with a small range of numbers,
a large range of brightness.
2. Using magnitudes and magnitude differences, simple plots may be constructed
that often provide considerable astrophysical insight to a given problem, e.g.,
the study of star clusters and the interstellar medium.
3. The overwhelming majority of optical astronomers use magnitudes, as they have
for more than 2000 years, making any change extremely difficult.

The disadvantages of magnitudes

1. The magnitude scale was originally conceived as an ordinal scale (1st magnitude,
2nd magnitude, 3rd magnitude, etc.) that was later transformed to a cardinal scale
(magnitude 1, magnitude 2, magnitude 3, etc.) without inversion (i.e., the bigger
the value of the magnitude the fainter the object and vice versa).
2. The scale is logarithmic, which makes simple comparison of the relative flux
densities of two or more astronomical objects
√ difficult.
3. The logarithmic base was selected as the 5 100 or approximately 2.511 886 to tie
the ordinal and cardinal schemes together. The difference in brightness between
a star of the first magnitude and one of the sixth magnitude was√ set equal to a
5
factor of 100, leading to a one magnitude difference equalling 100.
4. The zero point of a magnitude scale is chosen arbitrarily.
Given the uses to which magnitudes are put and their obvious universality it
would seem appropriate to attempt to bridge the gap between radiometry and
astronomical magnitudes.

8.3.1 A proposed magnitude system based on radiometric measures

In proposing a new magnitude scale, attention needs to be paid to the lists of
advantages and disadvantages given above. The current magnitude scale is defined
mathematically as:
m = −2.5 log(f ) + m0 (8.46)
where m is the magnitude and f is a number related to the flux density of the
object being measured. This number may be counts obtained from a photoelectric
photometer or CCD camera, or a measure obtained from a photographic plate (e.g.,
image area), or even a comparative visual estimate (see Chapter 12). m0 is the
 8.3 Magnitudes 139

Table 8.4. Magnitude

  differences for various flux-density
f
differences f0 with varying logarithmic bases

m
f
f0 log2 log10 log √
5
100
ln

2 1.0000 0.3010 0.7500 0.6931

10 3.3219 1.0000 2.5000 2.3026
50 5.6438 1.6990 4.2474 3.9120
100 6.6439 2.0000 5.0000 4.6052
10000 13.2877 4.0000 10.0000 9.2103

zero-point adjustment that transforms the measured magnitude of the detector into
that of a standard magnitude system.
Essentially, the simple revised magnitude system proposed changes Equation
(8.46) to:

m = k (logarithm)base (fjanskys ) (8.47)

The base of the logarithm has to be selected,√plus a constant coefficient, k. Four

possibilities for the logarithm base are, 2, 10, 5 100, and e, where e is defined by
Abramowitz & Stegun (1972) to be:
 1 n
e = lim 1+ = 2.718 281 828 4... (8.48)
n→∞ n

The magnitude changes in each system corresponding to a measured flux density

change of 2×, 10×, 50×,100× and 10 000× are shown √ in Table 8.4.
5
There is an obvious advantage in staying with the 100 system in that the values
of differential magnitude will remain as they are now, log10 is the most commonly
used system of logarithms, log2 is of particular significance in computing and
ln = loge is of significance in many branches of mathematics and physics.
Naturally, any decision on what form, if any, a new magnitude system should
take is for the IAU to decide. In order to provide meaningful worked examples on
how such a revised system might operate, the logarithmic base e has been selected.
An appropriate name might be e-magnitudes, em, which would be defined as:

emν = ln(fν ) (8.49)

where ν is the effective frequency of the electromagnetic radiation being measured

and fν is the monochromatic flux density at frequency ν measured in janskys.
140 Unit of luminous intensity (candela)

Example 1: e-magnitude computation.

Derive the e-magnitude for an A0V star with mV = 0.00 and fν = 3636 Jy (Bessell,
2001).
emV = ln(fν ) = ln(3636) = +8.199 (8.50)

Apparent and absolute magnitudes

The observed, or apparent magnitude, or monochromatic flux density, depends on
the intrinsic brightness of the celestial object, its distance and the amount of inter-
stellar absorption along the line of sight to the object. Interstellar extinction and
reddening is caused by the presence of dust grains that absorb and scatter the radi-
ation from the celestial object. Absolute magnitude is defined to be the magnitude
of an unreddened celestial body at a distance of 10 pc (3.085 677 6×1017 m).
The colour index is defined to be the difference between two magnitudes,
determined from observing a celestial object through two different bandwidths,
e.g., (B–V), (R–I). In radiometric terms, the colour index is the ratio of the
observed monochromatic flux densities in two frequency bands and is thus
dimensionless, e.g.,
fν (Jy)
(colour index)fν = B (8.51)
fνV (Jy)
A list of absolute magnitudes for stars of spectral type (see Chapter 12) O5 to M5
in the Johnson–Cousins–Glass broadband photometry is given in Cox (2000). These
magnitudes were converted to janskys, as shown in Example 2 below, using the
material in Table 8.3 for jcgU, jcgB and jcgV. Subsequently, these monochromatic
flux densities were transformed into e-magnitudes and e-colours.

Example 2: absolute e-magnitude and e-colour computations:

convert the UBV magnitudes for a G0V star to e-magnitudes and e-colours.
For a G0V star, the following absolute magnitudes may be taken or computed
from the values listed in Cox (2000): U = +5.04; B = +4.98; V = +4.40. Convert
these magnitudes into janskys (fU , fB and fV ) using the monochromatic flux den-
sities given by Bessell (2001) for an A0V star with zero magnitude and zero colour
indices (Equation 8.51).
fU0 = 1790 Jy fB0 = 4063 Jy fV0 = 3636 Jy (8.52)
fU = fU0 × 10−0.4 U (8.53)
fB = fB0 × 10−0.4 B (8.54)
fV = fV0 × 10−0.4 V (8.55)
 8.3 Magnitudes 141

For the G0V star, the computed fluxes are:

fU = 17.25 Jy fB = 41.39 Jy fV = 63.19 Jy (8.56)
The e-magnitudes corresponding to these fluxes are:
emU = ln(17.25) = +2.85
emB = ln(41.39) = +3.72
emV = ln(63.19) = +4.15 (8.57)
Note that in the e-magnitude system the bigger the numerical value of the e-
magnitude the brighter the astronomical object.
e-colours derived from the e-magnitudes are:
e(B − V) = 3.72 − 4.15 = −0.43
e(U − B) = 2.85 − 3.72 = −0.87 (8.58)
Plots of the original data from Cox (2000) and Straižys (1992) in the form
of colour–magnitude and colour–colour diagrams are shown in Figures 8.7 and
8.8, whilst similar plots derived from monochromatic flux densities are shown as
e-colour–magnitude and e-colour–colour diagrams in Figures 8.9 and 8.10.

–5

III
mV magnitude

5
V

10

–0.5 0.0 0.5 1.0 1.5 2.0

(B – V) colour

Figure 8.7. Colour–magnitude diagram showing the loci of luminosity class I

(dark grey), III (light grey) and V (black) stars as listed by Cox (2000) and Straižys
(1992).
142 Unit of luminous intensity (candela)

–1

0 V
(B – V) colour

I
III

–0.5 0.0 0.5 1.0 1.5 2.0

(U – B) colour

Figure 8.8. Colour–colour diagram showing the loci of luminosity class I (dark
grey), III (light grey) and V (black) stars as listed in Cox (2000) and Straižys (1992).
The colour of the stars changes from blue to red in the direction of increasing
value of the colour index. In the currently conventional colour–colour diagram,
the ordinate is plotted in the direction of decreasing (B–V) index.

The relationship between magnitudes mν , and e-magnitudes emν is given, in

frequency space, by the simple equation:

emν = ln((fν )0 ) − 0.921 mν (8.59)

where (fν )0 is the value, in janskys, of an A0V star of zero magnitude in the
passband that has an effective frequency of ν. The 0.921 is a conversion factor
relating natural logarithms to base-10 logarithms.

8.4 Summary and recommendations

8.4.1 Summary
The SI unit of luminous intensity is the candela, which relates strongly to the early
days of human-eye-tuned photometry and covers a very restricted wavelength or
frequency range. In parallel with the pure photometric system of units are those of
radiometry and astronomy. The radiometric units are directly related to the mass,
 8.4 Summary and recommendations 143

15 I

10

III
emV (ln(Jy))

–5
–1.5 –1.0 –0.5 0.0 0.5
(B – V) e – colour

Figure 8.9. e-colour–magnitude diagram showing the loci of luminosity class I

(dark grey), III (light grey) and V (black) stars derived using Equations (8.51) to
(8.57). Stars increase in brightness with increasing value of e-magnitude and their
colour changes from red to blue with increasing value of e-colour.

length and time base units of the SI system. Astronomical units are rather more
esoteric. Comparisons between the different photometries are given and worked
examples of converting from one set of units to another given. For many of the
common astronomical photometric systems, the central wavelengths and band-
widths are transformed from wavelength to frequency space. Examples are given
of determining the spectral irradiance (radiometic unit) and the monochromatic flux
density of stars from many different published non-SI unit sources. Magnitudes,
a much-used astronomical photometric unit, are converted to monochromatic flux
densities and a flux versus frequency plot of HST spectrographic observations of
Vega produced.
Some selected commonly used photometric systems from the ultraviolet to the
infrared are calibrated in radiometric terms and in janskys. Very high frequency
astronomical observations in the X-ray and γ -ray regions are also converted to
janskys, with observations from the opposite end of the electromagnetic spectrum
in the submillimetre, millimetre and radio regions. Wavenumbers of three different
144 Unit of luminous intensity (candela)

I
0
e(U – B) colour

V
–1

–2
I III

–1.5 –1.0 –0.5 0.0 0.5

e(B – V) colour

Figure 8.10. e-colour–colour diagram showing the loci of luminosity class I (dark
grey), III (light grey) and V (black) stars as listed by Cox (2000). The stars colour
change from red to blue in the direction of increasing colour index along both axes.

types, μm−1 , cm−1 and m−1 are converted to frequencies in hertz and simple
worked examples given.
A discussion on the advantages and disadvantages of astronomical magnitudes
is given along with a proposed magnitude system based on radiometric measures
in janskys.

8.4.2 Recommendations
The major shortcoming of the defined SI photometric unit, the candela, when applied
to astronomy, is due principally to the restricted frequency band covered. However,
the radiometric units defined in terms of base SI units are easily related to astronom-
ical photometry. The great variety of different units in common use in astronomical
photometry makes inter-comparison between them time consuming and prone to
the making of mistakes.
For observational work, the most logical unit to use for monochromatic flux
densities is the jansky. It has been in use for many decades by both radio and
infrared astronomers and is not difficult to calculate from other similar units, which
 8.4 Summary and recommendations 145

may differ by scale factors or by being tied to wavelength rather than frequency
space. The logarithm (whether natural or to base 10) of the flux density in janskys
makes a very acceptable physically based magnitude that allows for the continued
use of colour–magnitude and colour–colour diagrams. For both theoretical and
observational studies of the total power output of all types of astronomical bodies,
the watt is ideal.
 9
Unit of thermodynamic temperature (kelvin)

9.1 SI definition of the kelvin

The kelvin, unit of thermodynamic temperature, is the fraction 1/273.16 of the
thermodynamic temperature of the triple point of water.
The dimension of thermodynamic temperature is [], its unit is the kelvin and
its symbol is K.

9.1.1 Possible future definition of the kelvin

Under discussion at the present time is the redefining of the unit of temperature in
the following way:

The kelvin is a unit of thermodynamic temperature such that the Boltzmann constant
is exactly 1.380 650 5 × 10−23 J . K−1 (joules per kelvin).

9.1.2 Definition of temperature

The average kinetic energy of the molecules, atoms and ions that comprise an object
may be used as a measure of its thermodynamic temperature.
Given two objects with different temperatures, the one with the higher temper-
ature (the hotter object) will transfer heat to the object with the lower temperature
(the cooler object) by means of radiation, convection or conduction, depending on
the nature of the objects and their location, until the temperatures of the objects are
equalized.

9.1.3 Thermodynamic temperature

The thermodynamic temperature, Tt , of a system may be defined as the inverse of
the rate of change of entropy S, with respect to the internal energy U , assuming
that the volume V of the system and the number of its constituent parts N remain

146
 9.2 Temperature scales 147

constant,54 i.e.
 ∂S −1
Tt = (9.1)
∂U N,V

The entropy S, is a measure of the disorder of the system and increases with
increasing levels of disorder (e.g., steam has a higher entropy than water, which in
turn has a higher entropy than ice).

9.1.4 Kinetic temperature

The kinetic temperature, Tk , may be defined for an ensemble of identical particles
of individual mass m, with root-mean-square particle speed < v2 > as:

m < v2 >
Tk = (9.2)
3k

where the speed distribution f (v), of the particles is Maxwellian,55 i.e.,

 m  32 2 − m v2
f (v) = 4π v e 2k T (9.3)
2π k T

Example: check the dimensional consistency in Equation (9.2)

The dimension of the LHS of Equation (9.2) is [
]. The RHS of Equation (9.2)
involves Boltzmann’s constant k, which has J . K−1 as a unit and may be expressed
in dimensional terms as [M] . [L]2 . [T ]−2 . [
]−1 , hence:

[M] . [L]2 . [T ]−2

dim[RHS] =
[M] . [L]2 . [T ]−2 . [
]−1
= [
]
= dim[LHS]

9.2 Temperature scales

Various temperature scales have been used over the past two hundred years, of
which the Celsius, Fahrenheit, Kelvin and Rankine scales are still in common use.
The differences between such scales are essentially due to different temperature
unit size and different zero points.

54 See http://hyperphysics.phy-astr.gsu.edu/hbase/thermo/temper2.html
55 See http://hyperphysics.phy-astr.gsu.edu/hbase/kinetic/kintem.html
148 Unit of thermodynamic temperature (kelvin)

9.2.1 Converting the Celsius, Fahrenheit and Rankine scales to the Kelvin
temperature scale
Temperatures measured in the Celsius and Fahrenheit scales may be either positive
or negative and both define the freezing point of water with a pre-assigned value for
temperature (0 for celsius and 32 for fahrenheit). The kelvin and rankine have as
their lowest temperature value 0, known as absolute zero, the temperature at which,
in classical mechanical systems, all translational motion of atoms and molecules
ceases. In quantum mechanical systems there remains the possibility of zero-point
energy-induced particle motion.
The temperature intervals in the Kelvin and Celsius scales are the same,
1 K = 1 ◦ C, as are those in the Fahrenheit and Rankine scales, 1 ◦ R = 1 ◦ F. The
Kelvin and Rankine scales are both absolute temperature systems.
The IAU recommends citing temperatures in kelvin. To assist with their rec-
ommendation, the appropriate conversion formulae from one temperature system
to another is given. There are many online converters such as that given in the
Engineering Toolbox.56

To convert Celsius temperatures to Kelvin temperatures

θK = θC + 273.16 (9.4)

where θK is the temperature in kelvin and θC the temperature in degrees Celsius.

To convert Fahrenheit temperatures to Kelvin temperatures

5 (θF − 32)
θK = + 273.16 (9.5)
9
where θF is the temperature in degrees Fahrenheit.

To convert Rankine temperatures to Kelvin temperatures

5 θR
θK = (9.6)
9
where θR is the temperature in degrees Rankine.
In astronomy K, ◦ C and ◦ F are all commonly used. Fahrenheit is used mainly in
popular astronomical publications in countries where that scale is more familiar to
the local readers.

56 See http://www.engineeringtoolbox.com/temperature-d_291.html
 9.3 Some examples of the temperatures of astronomical objects 149

9.3 Some examples of the temperatures of astronomical objects

Solar system temperatures
For the Sun, planets, some dwarf planets, some satellites and Halley’s comet, tem-
peratures have been determined for their surfaces (S) or within their atmospheres
(A) at a point where the atmospheric pressure is equal to 105 Pa. Temperatures are
given in Tables 9.1 and 9.2 in the four systems, kelvin (K), celsius (C), rankine (R),
and fahrenheit (F), defined above in Equations (9.4) to (9.6).
The kelvin scale values were obtained from Cox (2000) for the Sun, planets and
satellites, from Emerich et al. (1988) for Halley’s comet and websites for Ceres57
and Eris.58

More extreme astronomical temperatures

The temperatures in Table 9.2 range from absolute zero through the temperature of
the cosmic microwave background (CMB) radiation (the red-shifted remnant of the
Big Bang event), interstellar molecular gas clouds, cool brown dwarfs, hot stellar
surfaces, the central temperature of a pre-eruptive supernova star to the Planck
temperature that existed a minute fraction of a second after the Big Bang. As with
Table 9.1, temperatures are given in the Kelvin, Celsius, Rankine and Fahrenheit
systems.

Determination of effective temperature

The effective temperature, Teff , of a celestial object is defined by the equation:
 L  14
Teff = (9.7)
4π σ R 2
where L is the luminosity of the object in watts, R is the radius of the object in metres
and σ is the Stefan–Boltzmann constant equal to 5.670 51 × 10−8W . m−2 K−4 . It
is not uncommon in astronomy to find the Stefan–Boltzmann constant given in cgs
units, with the radius and luminosity of the celestial object likewise in cgs units.

Conversion of the Stefan–Boltzmann constant from cgs to SI units

In cgs units the value of the Stefan-Boltzmann constant σ , is:

σ = 5.670 51 × 10−5 erg . s−1 . cm−2 . K −4 (9.8)

= 5.670 51 × 10−5 (10−7 W) . (10−2 m)−2 . K −4

57 See http://www.princeton.edu/∼willman/planetary_systems/Sol/Ceres/
58 See http://starchild.gsfc.nasa.gov/docs/StarChild/solar_system_level2/eris.html
150 Unit of thermodynamic temperature (kelvin)

Table 9.1. Solar System temperatures

Temperature
Object K ◦C ◦R ◦F

Solar centre 1.6 d 7 1.6 d 7 2.88 d 7 2.88 d 7

Sun Teff 5779 5506 10 400 9943
Planets
Mercury (S) 440 167 792 332
Venus (S) 730 457 1314 854
Earth (S) 290 17 522 62
Mars (S) 226 −47 407 −53
Jupiter (A) 165 −108 297 −163
Saturn (A) 134 −139 241 −218
Uranus (A) 76 −197 137 −323
Neptune (A) 73 −200 131 −328
Dwarf planets
Ceres (S) 173 −100 311 −148
Pluto (S) 58 −215 104 −355
Eris (S) 30 −243 54 −406
Satellites
Moon (S) 250 −23 450 −10
Titan (S) 94 −179 169 −290
Triton (S) 38 −235 68 −391
Comet
Halley 375 102 675 215

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .

= 5.670 51 × 10−5 . 10−3 W . m−2 . K −4

= 5.670 51 × 10−8 W . m−2 . K −4 (9.9)

where Equation (9.9) gives the value of the Stefan–Boltzmann constant in SI units.

Example: determine the effective temperature of the Sun

Given the luminosity of the Sun, L = 3.845 × 1026 W and the solar radius to be
R = 6.955 08 × 108 m, then its effective temperature is:
 L  14
Teff = 2
4πσ R
 3.845 × 1026 1
4
= −8 8 2
4π × (5.67051 × 10 ) × (6.95508 × 10 )
= 5779 K (9.10)
 9.4 Blackbody radiation 151

Table 9.2. More extreme astronomical temperatures

Temperature
Object K ◦C ◦R ◦F

Absolute zero 0 −273.2 0 −459.7

Cosmic microwave background 2.7 −270.4 4.9 −454.8
Interstellar molecular cloud 10 −263 18 −442
Cool brown dwarf 450 177 810 350
Teff (O5V star) 42 000 41 730 75 600 75 140
25 M supernova 3.4 d 9 3.4 d 9 6.1 d 9 6.1 d 9
Planck temperature 1 d 32 1 d 32 1.8 d 32 1.8 d 32

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .
Notes:
Cosmic microwave background radiation was discovered by Penzias & Wilson (1965). The value given in Table 9.2
is that obtained from measurements made using the COBE (COsmic Background Explorer) satellite by Mather
et al. (1994).
Ferriere (2001) lists various parameters of the different components of interstellar gas for the region of space
near the Sun. Temperatures are given for the molecular (the value in the table), cold and warm atomic, and warm
and hot ionized gas.
Lucas et al. (2010) reports on the discovery of a cool brown dwarf listed as UGPS J0722-05 in the UKIDSS
Galactic Plane Survey. A provisional spectral type (see Chapter 12) T1O has been assigned, though the possibility
of a new spectral class has been suggested. The temperature in Table 9.2 is the mean value of the range given.
The effective temperature (see below) of the brightest main sequence star (spectral type O5V) is given as
42 000 K in Cox (2000).
Weaver et al. (1978) give the central temperature of a 25 M (= 4.97 × 1031 kg) supernova at the onset of core
ignition for the silicon-to-iron-burning stage.
The Planck temperature is that which existed at the Planck time (5.38 × 10−44 s) after the Big Bang (Lang,
2006).

9.4 Blackbody radiation

A blackbody may be defined as an idealized object that neither reflects nor scatters
electromagnetic radiation incident upon it, but absorbs and re-emits it completely.
A blackbody emits radiation as a continuous spectrum that depends solely upon its
temperature and not upon its constituent parts, shape or internal structure.

Planck’s law
The frequency distribution of the radiation emitted by a blackbody following
Planck’s law is a function of temperature alone. The spectral radiance or specific
intensity, Bν (T ), emitted by a blackbody whose temperature is T K, at a frequency
of ν Hz, is given by:

2 h ν3  1 
Bν (T ) = 2 hν
W . m−2 . Hz−1 . sr −1 (9.11)
c ekT − 1
152 Unit of thermodynamic temperature (kelvin)

where h is Planck’s constant and equal to 6.626 075 × 10−34 J . s, k is Boltzmann’s

constant and equal to 1.380 658 × 10−23 J . K−1 , and c is the speed of light equal to
2.997 924 × 108 m . s−1 .
In terms of wavelength λ, in metres, the expression for the spectral radiance
or specific intensity, Bλ (T ), emitted by a blackbody whose temperature is T K, is
given by:


2hc2 1
Bλ = 5 hc
− 1 W . m−2 . m−1 . sr −1 (9.12)
λ e λkT

Examples of log–log blackbody curves in frequency space for various astronom-

ical objects are given in Figure 9.1 from the very low temperature (T = 2.73 K) of
the cosmic background radiation to the very high temperature (T = 3.4 × 109 K)
near the centre of a 25 M star.

10

9 K
1 0
3.4x
T= e
7 K
0
x1
log(Bν) W . m–2 . sr –1 . Hz

0 1.6
T=

9K d
5 77
–10 T=
K
2 90 c
T=
b
K
2.73 a
–20
T=

8 10 12 14 16 18 20
log(ν) Hz

Figure 9.1. Blackbody curves computed using Equation (9.11). Curve ‘a’ is for a
blackbody at the same temperature as the cosmic background radiation, curve ‘b’
for the surface of the Earth, curve ‘c’ for the surface of the Sun, curve ‘d’ for the
centre of the Sun and curve ‘e’ for the centre of a 25 M star.
 9.4 Blackbody radiation 153

9.4.1 Wein’s displacement law

Wein’s displacement law gives the wavelength λmax , or frequency νmax , of the
maximum value of the specific intensity of a Planck curve. The wavelength or
frequency is a function solely of the temperature, but is different in the sense that
λmax = c / νmax due to the different forms of the Planck function in wavelength
and frequency space. Values for λmax and νmax may be found by differentiating the
appropriate expression for specific intensity with respect to either λ or ν and setting
the result equal to zero.
For λmax the expression is:
b
λmax = (9.13)
T
where T is the blackbody temperature in kelvin and the Wein displacement constant
b = 0.002 897 8 K . m.
For νmax the expression is:

νmax = 5.88 × 1010 T (9.14)

Example: compute λmax and νmax for a blackbody of temperature 2.726 K

0.002 897 8
λmax = = 1.063 × 10−3 m = 1.063 mm (9.15)
2.726
νmax = 5.88 × 1010 × 2.726 = 1.603 × 1011 Hz (9.16)

9.4.2 Stefan–Boltzmann law

The Stefan–Boltzmann law states that the power P , radiated per unit surface area of
a blackbody, is directly proportional to the fourth power of its absolute temperature,
so that:
P =σT4 (9.17)
In SI units, P is measured in W . m−2 and T in K. The total power emitted by a
blackbody of temperature, T , in all directions is simply:

Ptotal = 4πσ T 4 (9.18)

The effective temperature, T , of a star may be written as:

 L  14
T = (9.19)
4πσ R2
or, by rearranging, the luminosity L of the star becomes:

L = 4πσ R2 T4 (9.20)

154 Unit of thermodynamic temperature (kelvin)

9.4.3 Aerial or antenna temperature

In radio astronomy, the aerial or antenna temperature, Ta , is often used as an alter-
native way of expressing the power, Pa , of a celestial radio source incident upon
the receiving element of a radio telescope (Lovell & Clegg, 1952). The antenna
temperature may be related to the source power by:
Pa
Ta = (9.21)
4 k ν
where ν is the frequency bandwidth of the radio telescope and k is Boltzmann’s
constant.

9.4.4 Brightness temperature

The brightness temperature, TB , is the temperature that a blackbody would need
to have in order to match the observed specific intensity or spectral radiance of an
astronomical source at an observed frequency ν, hence:
c2 Bν
TB = (9.22)
2 ν2 k
where the Rayleigh–Jeans approximation to the Planck law, valid for the low
frequencies of radio astronomy, may be used.
If the celestial radio source is not a blackbody, such as a thin plasma, a maser or
emitting synchrotron radiation, then the temperature derived using Equation (9.22)
will not be correct and either too low or, sometimes, far too high.

9.5 Spectral classification as a temperature sequence

In Chapter 12, an outline of the way in which spectral types are assigned to stars
using the relative strengths of the spectral lines produced by various ions, atoms
and molecules is given. A relationship between temperature and spectral type or
photometric colour differences may be determined empirically, which allows a base
SI unit to be used to classify stars.

9.5.1 Energy states in atoms

Unlike, for example, the Solar System where an infinite number of orbits are pos-
sible for the planets, dwarf planets and other bodies such as comets, locations
available to an electron relative to the nucleus of the atom of which it is a part is
restricted by quantum theory to certain energy values. The absorption or emission
of a photon by an electron allows it to move from one energy state to another,
creating an absorption or emission line in the spectrum of the star.
 9.5 Spectral classification as a temperature sequence 155

The energy of a bound electron is given by Irwin (2007) as:

2 π 2 me e4
En = − erg (9.23)
n2 h2
where me is the mass in grams and e is the charge of an electron in electrostatic
units, h is Planck’s constant in cgs units (erg . s) and n is the energy quantum level
(an integer).
Substituting me = 9.1093 × 10−28 g, e = 4.8032×10−10 esu and h = 6.626 075×
10−27 erg . s in Equation (9.23) gives:
−1
En = (2.17987 × 10−11 ) erg
n2
−1
= 2 (2.17987 × 10−18 ) J
n
−1
= 2 (13.61) eV (9.24)
n
Note that Equation (9.24) is given in cgs and esu units, resulting in a measure of
energy also in cgs units, the erg. The transformation to SI units is accomplished by
multiplying by the factor 10−7 since 1 J = 107 erg. The problems associated with
transforming values in one set of electrostatic and electromagnetic units to another
is dealt with in the following chapter (Chapter 10). Equation (9.24) also gives the
transformation of the cgs energy value to one in electron volts.
The electron volt is a non-SI unit of energy accepted for use with the SI whose
values in SI units are obtained experimentally. The electron volt dimension is
[M ] . [L]2 . [T ]−2 and symbol eV.
The electron volt is commonly used in spectroscopy. Its equivalent energy
measurement in SI units is given by:
1 eV = 1.602 177 33 × 10−19 J (9.25)
This numerical value is the same as the SI value for the electrical charge e of
an electron, 1.602 177 33 × 10−19 C (C = coulomb), which in turn is numerically
the same as the esu value of the electron charge divided by 10 c (c = speed of
light = 2.997 924 m . s−1 ), i.e.,
e = 1.602 177 33 × 10−19 C
4.803 206 8 × 10−10 esu
≡ (9.26)
10 × 2.997 924 × 108 m . s−1
The capital C is the symbol for the SI unit of electrical charge, the coulomb (see
Chapter 10).
156 Unit of thermodynamic temperature (kelvin)

9.5.2 Energy levels for the hydrogen atom

Using Equation (9.24) above it is possible to compute the values for the different
energy levels (n = 1, 2, 3 etc. in the equation) for the hydrogen atom in SI units (J)
as well as in eV. The energy difference Ej,k , between any two states j and k is
given by:
1 1
Ej,k = Ej − Ek  −2.18 × 10−18 2 − 2 J (9.27)
nj nk
1 1
 −13.6 2 − 2 eV (9.28)
nj nk
The frequency νj,k , of the emitted or absorbed radiation is given by:
Ej,k
νj,k = (9.29)
h
Example: find the energy, frequency and wavelength of the Hβ spectrum line
The lower level of the Hβ Balmer line transition is at n = 2 and the upper level is
at n = 4, so:
1 1
E2,4 = E2 − E4  −2.18 × 10−18 2 − 2
2 4
−19
= − 4.088 × 10 J (9.30)
The frequency of Hβ radiation is νHβ where:
E2,4
νH β = = 6.17 × 1014 Hz (9.31)
h
and its emission or absorption wavelength λHβ is:
c 2.997 924 × 108
λH β = = = 4.859 × 10−7 m
νHβ 6.17 × 1014
= 485.9 nm (9.32)
In Table 9.3, values for the energy differences, frequencies and wavelengths in
SI units for the hydrogen Lyman, Balmer and Paschen line series are given up to
energy level 6.
Grotrian diagram: Figure 9.2 is known as a Grotrian diagram. The numerical
values for the energy levels in joules, computed above and listed in Table 9.3, have
been plotted in Figure 9.2 as the ordinate. Each level is represented as an arbitrary
2 units along the abscissa. The arrow-headed lines show electron transitions from a
lower to a higher state of energy representing the absorption of a photon; these are
 9.5 Spectral classification as a temperature sequence 157

Table 9.3. Hydrogen spectrum lines

En En nk − n1,2,3 E ν λ
Name n J eV J Hz nm
Lyman series
L∞ ∞ 0 0 ∞−1 2.18 d − 18 3.29 d 15 91
L 6 −6.06 d − 20 −0.82 6−1 2.12 d − 18 3.20 d 15 94
Lδ 5 −8.72 d − 20 −1.19 5−1 2.09 d − 18 3.16 d 15 95
Lγ 4 −1.36 d − 19 −1.85 4−1 2.04 d − 18 3.08 d 15 97
Lβ 3 −2.42 d − 19 −3.30 3−1 1.94 d − 18 2.92 d 15 103
Lα 2 −5.45 d − 19 −7.41 2−1 1.63 d − 18 2.47 d 15 122
1 −2.18 d − 18 −13.61
Balmer series
H∞ ∞ 0 0 ∞−2 5.45 d − 19 8.23 d 14 364
Hδ 6 −6.06 d − 20 −0.82 6−2 4.84 d − 19 7.31 d 14 410
Hγ 5 −8.72 d − 20 −1.19 5−2 4.58 d − 19 6.91 d 14 434
Hβ 4 −1.36 d − 19 −1.85 4−2 4.09 d − 19 6.17 d 14 486
Hα 3 −2.42 d − 19 −3.30 3−2 3.03 d − 19 4.57 d 14 656
2 −5.45 d − 19 −7.41
Paschen series
P∞ ∞ 0 0 ∞−3 2.42 d − 19 3.66 d 14 820
Pγ 6 −6.06 d − 20 −0.82 6−3 1.82 d − 19 2.74 d 14 1094
Pβ 5 −8.72 d − 20 −1.19 5−3 1.55 d − 19 2.34 d 14 1282
Pα 4 −1.37 d − 19 −1.85 4−3 1.06 d − 18 1.60 d 14 1875
3 −2.42 d − 19 −3.30

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .

CONTINUUM
0
energy levels 10–1 8J

Hα Hβ
Pα
13.6eV

–1

Lα Lβ

–2

0 1 2 3 4 5 6 7

Figure 9.2. A Grotrian diagram for the Lyman (L), Balmer (H) and Paschen (P)
series of hydrogen lines. The vertical axis is measured in units of 10−18 J. The
energy difference in electron volts between the n = 1 and the n = ∞ is shown by
the two-headed dark grey arrow labelled 13.6 eV.
158 Unit of thermodynamic temperature (kelvin)

known as bound-bound transitions. Transitions from a higher to a lower energy state

occur with the emission of a photon; these are also bound-bound transitions. An
arrow beginning at a horizontal line below the zero energy level and ending at a point
above the zero energy level represents a bound-free transition. Free electrons (i.e.,
those not bound to an atomic, ionic or molecular nucleus) may undergo free-free
transition in the region labelled ‘CONTINUUM’ in the figure.
The Rydberg formula: The following simple relationship between the inverse
wavelength λ, in metres, of a spectrum line whose initial energy level is nj and
whose final energy level is nk (for nj > nk ) was discovered by Rydberg in the
nineteenth century:
1 1 1
= RH 2 − 2 m−1 (9.33)
λ nj nk
where RH , the Rydberg constant in SI units, is for the hydrogen atom (Cox, 2000):

RH = 1.096 775 831 × 107 m−1 (9.34)

The inverse wavelength version of Rydberg’s formula may be written in terms of

frequency ν, and the speed of light c, as:
1 1
ν = c RH 2 − 2 Hz (9.35)
nj nk

Example: use the Rydberg formula to determine the frequency and

wavelength of the radio-frequency hydrogen line formed by the energy level
nj = 50 to nk = 51 transition
Substitute for c, RH , nj and nk in Equation (9.35) to compute the value for the line
frequency ν, in hertz:
 1 1 
ν50−51 = 2.997 924 × 108 × 1.096 776 × 107 − Hz
502 512
= 5.107 160 × 1010 Hz ≡ 51.072 GHz (9.36)

Similarly, substituting for RH , nj and nk in Equation (9.33) to compute firstly the

value for λ−1 and then invert to yield λ:
1  1 1  −1
= 1.096 775 831 × 107 − m
λ 502 512
= 170.356 569 m−1 (9.37)

and

λ = 5.870 041 × 10−3 m ≡ 5.870 mm (9.38)

 9.5 Spectral classification as a temperature sequence 159

9.5.3 The Saha–Boltzmann equation

The Saha–Boltzmann equation is valid when the condition of local thermodynamic
equilibrium (LTE) prevails, i.e., assuming the behaviour of the matter component
of the gas under consideration is governed by Maxwell–Boltzmann statistics, even
if the electron component deviates from statistical equilibrium (see, e.g., Collins,
1989).
Under conditions of LTE, the relationship between the number density of atoms,
Ni , in the ith state of ionization (i.e., the atom has had i electrons removed) to those
atoms, Ni+1 , in the (i + 1)th ionization state, was shown by Saha (1921) to be:
3
Ni+1 2 ui+1 (T ) (2 π me k T ) 2 − χi
Ne = e kT (9.39)
Ni ui (T ) h3
Equation (9.39) is sometimes written with electron pressure, Pe , substituted for Ne ,
the electron density, resulting in:
3 5
Ni+1 2 ui+1 (T ) (2 π me ) 2 (k T ) 2 − χi
Pe = e kT (9.40)
Ni ui (T ) h3
where h is Planck’s constant, k is Boltzmann’s constant, me is the mass of the
electron, χi is the ionization potential and is the energy required to move an electron
from the ground state of an i-times ionized atom to the free (i.e., unbounded) state,
T is the temperature of the gas and ui (T ) is the partition function defined by:
 −j
ui (T ) = gj e k T (9.41)
j

where j is the excitation potential above the ground state of the j th energy level
of the atom or ion and gj is the statistical weight (gj = 2 J + 1) of the j th energy
level, J being the total angular quantum number of the energy level.

Example: calculate the ratio of singly ionized aluminium to neutral

aluminium in a solar-type atmosphere at a point where the electron
pressure is equal to 3 Pa
Using Equation (9.40) and substituting the following values from Cox (2000)
and Aller (1963) (with transformation to SI units from cgs units where
necessary) gives: T = 5700 K, χ0 = 9.590 265 × 10−19 J, 2uu01 = 0.34, Pe = 3 Pa,
me = 9.1093 × 10−31 kg, h = 6.626075 × 10−34 J . s, k = 1.380658 × 10−23 J . K−1
3 5
N1 2 u1 (T ) (2 π me ) 2 )(k T ) 2 − χ0
= e kT (9.42)
N0 u0 (T ) Pe h3
= 47.27 (9.43)
160 Unit of thermodynamic temperature (kelvin)

i.e.,
N1 = 47.27 N0 (9.44)
so just over 2 % of the aluminium atoms are neutral and the rest are singly ionized.

9.5.4 Equivalent width of a spectral line

The profile of a spectral line formed in the atmosphere of a star is dependent on
many different variables, such as the atmospheric pressure, the temperature, the
rotation velocity of the star, the effects of any electric or magnetic fields present
and so on. The general shape may be described as Gaussian, unless the line is
saturated. The measured area contained within the line profile may be set equal to
that of a rectangular area, whose height is equal to that of an absorption line that is
completely black at the centre and whose width is then assigned a value such that the
profile area and the rectangular area match (see Figure 9.3). This width is known
as the equivalent width, Wλ , of the spectral line and is defined in wavelength
space to be:
 λmax 
1 − Fλ 
Wλ = dλ = F0 . (λ2 − λ1 ) (9.45)
λmin F0
where F0 is the flux density of the continuum fitted to the observed stellar spectrum
and Fλ is the flux density at wavelength λ over the wavelength range of interest
(λmin → λmax ). Collins (1989) notes that for narrow spectral lines with (λmax −
λmin )  λ0 it is possible to write the equivalent width, Wν , in frequency space as:
Wλ ν0
Wν ≈ (9.46)
λ0

F0 continuum level

Fλ

0
λmin λ1 λ0 λ2 λmax λ

Figure 9.3. The area enclosed by the Gaussian-shaped spectral line profile between
λmin and λmax , centred on wavelength λ0 is equal to the rectangular area, F0 . (λ2 −
λ1 ), whose width, (λ2 − λ1 ), is defined to be the equivalent width of the actual
spectral line.
 9.5 Spectral classification as a temperature sequence 161

By assigning appropriate values to the constants and variables in Equations

(9.39), (9.40) and (9.41) it is possible to derive the ratio of spectral line strengths
for selected atoms and ionized atoms. This ratio is a function of the effective
temperature of the star. The relative strengths of spectrum lines in stars may be
directly measured from stellar spectrograms and thus used in conjunction with the
line-strength ratios determined from the Saha–Boltzmann equation to produce an
estimate of the star’s temperature. Since the ratios of line strengths are also used
to classify stellar spectra (see Chapter 12) and determine spectral class, it becomes
possible to equate spectral class directly with stellar temperature in kelvin (see
Figure 9.4). More detailed discussions on the use of the Saha–Boltzmann equation
may be found in, e.g., Lang (2006), Irwin (2007), Collins (1989), Cox (2000) and
Aller (1963).

9.5.5 Colour temperature

For unreddened or dereddened monochromatic flux density measurements of stars
obtained via multicolour photometry, a direct relationship exists between spectral
type and colour index and hence between temperature and colour index.

temperature (K)

3500 5000 6000 8000 10 000 30 000

equivalent width

CaII HI

TiO HeI
HeII
CaI
FeI FeII MgII
SiIII

M0 K0 G0 F0 A0 B0 O5
spectral type

Figure 9.4. A plot showing the variation of the strength (equivalent width) of
various atoms (e.g., FeI iron) and ions (e.g., FeII singly ionized iron) against both
spectral type and temperature. This figure was adapted, reversed and updated from
the original, which appeared in Struve et al. (1959). The black lines are for neutral
atoms, the dark grey lines are for ions and the wider light grey line is for the
molecule TiO. It can readily be seen that the ratio of equivalent widths of atoms
and/or ions to other atoms and/or ions is a function of spectral type and temperature.
162 Unit of thermodynamic temperature (kelvin)

The colour temperature of an astronomical body may be determined if the relative

energy distribution of the body is known. Consider the apparent monochromatic
flux densities f1 and f2 , at two wavelengths λ1 and λ2 , then the ratio of f1 and f2
is the same as that derived using Planck’s law, so:
hc
f1 λ52 e λ2 k T − 1
= (9.47)
f2 λ51 e λ1hkcT − 1

where c is the speed of light, h is Planck’s constant, k is Boltzmann’s constant and

T is the colour temperature. Rewriting the flux values as Pogson magnitudes, m1
and m2 , gives:
f 
1
m1 − m2 = −2.5 log + c1 (9.48)
f2
where constant c1 , depends on the different zero points of the magnitude scales
used. The Wien approximation may be used in the optical part of the spectrum to
allow the colour temperature, Tc , to be written as:
 
2.5 log(e) h c λ11 − λ12
Tc =   5  (9.49)
k (B − V) − c1 + 2.5 log λλ21

Example: determine the colour temperature of an A0V star with colour

index (B − V) = 0.0
Cox (2000) gives the monochromatic flux densities in cgs units in the B and V
bands, in wavelength space, of such a star as:
−1
B = 6.40 × 10−9 erg . s−1 . cm−2 . Å
−1
V = 3.75 × 10−9 erg . s−1 . cm−2 . Å (9.50)
which in SI units are equivalent to:
B = 6.40 × 10−2 W . m−2 . m−1
V = 3.75 × 10−2 W . m−2 . m−1 (9.51)
with h being Planck’s constant, c the speed of light, k Boltzmann’s con-
stant and remembering that 1 W = 107 erg . s−1 , 1 m2 = 104 cm2 and 1 m = 1010 Å.
Substituting the appropriate values into Equations (9.48) and (9.49) gives:
 3.75 × 10−2 
(B − V) = 0 = −2.5 log + c1 (9.52)
6.40 × 10−2
rearrange for c1 :
c1 = −0.58 (9.53)
 9.5 Spectral classification as a temperature sequence 163

and for Tc :
Tc = 12 340 K (9.54)
Note that this value is considerably higher than the value for the effective tem-
perature of 9790 K given by Cox (2000). Selecting other pairs of photometric bands
would produce different values for the colour temperature. By way of illustration,
consider the red and near-infrared bands R and I.

Example: determine the colour temperature of an A0V star with colour

index (R − I) = 0.0
The computation is as for the previous example, substituting the appropriate
values for R and I wavelengths in metres and monochromatic flux densities in
W . m−2 . m−1 into Equations (9.48) and (9.49).

λR = 7.1 × 10−7 m fR = 1.75 × 10−2 W . m−2 . m−1

λI = 9.7 × 10−7 m fI = 8.40 × 10−3 W . m−2 . m−1
(R − I) = 0.0 (9.55)

The colour temperature so derived is:

Tc = 7398 K (9.56)

The colour temperature does not represent the physical temperature within the
photosphere of a star, but rather may be used to give an indication of the slope of
the energy distribution emitted by a star.

9.5.6 A lookup table to convert (B − V) colour index to effective temperature

A lookup table is defined in computer science to be a data structure such as a
numerical or alphanumerical array that may be used to replace a runtime compu-
tation. From a listing such as that given in Cox (2000) of effective temperature
against spectral type and various colour indices, it is possible to use linear interpo-
lation to estimate a value for the temperature of a star. The listed colour indices are
dereddened so that any observational colour index must first be corrected for the
effects of interstellar obscuration due to dust and gas. From the lookup table, extract
the values and colour index and corresponding effective temperature immediately
above (B − V)i and below (B − V)i+1 the observed star’s dereddened colour index
(B − V) . The interpolated effective temperature of the star, Teff , , is then calculated
from:
 Teff ,i+1 − Teff ,i 
Teff , = (B − V) + κ (9.57)
(B − V)i+1 − (B − V)i
164 Unit of thermodynamic temperature (kelvin)

where  
Teff ,i+1 − Teff ,i
κ = Teff ,i + (B − V)i (9.58)
(B − V)i+1 − (B − V)i

Example: determine the effective temperature of the Sun from a lookup

table (Cox, 2000)
Substitute the following data into Equations (9.57) and (9.58):
Teff ,i = 5790 K Teff ,i+1 = 5560 K
(B − V)i = 0.68 (B − V)i+1 = 0.63
(B − V) = +0.650
to give:
 5560 − 5790 
κ = 5790 + (0.63) = 8688 K
0.68 − 0.63
 5560 − 5790 
Teff , = (0.650) + 8688 = 5698 K
0.68 − 0.63
which may be compared with the value of 5777 K for the effective temperature of
the Sun given in Cox (2000). A plot of effective temperature against (B − V) colour
index is given in Figure 9.5. The location of the Sun in the plot is shown with an
arrow.

O5
40 000

30 000 B0
Teff K

20 000

A0
10 000 F0
G0
K0
M0
M5

0
Sun

–0.5 0.0 0.5 1.0 1.5

(B – V)

Figure 9.5. A colour index (B − V), versus (Teff ) for main sequence stars from
type O5 to M5 (Cox, 2000).
 9.6 Model stellar atmospheres 165

So using the spectrophotometric or lookup table methods outlined above it is

possible to replace spectral type and colour index by an approximate effective
temperature in SI unit kelvin, but the use of colour temperature rather than colour
index would generally be a poor substitute and at worst even misleading.

9.6 Model stellar atmospheres

Currently, the most realistic way of determining stellar photospheric temperatures
is to compare either an observational spectrogram or a set of discrete photometric
measurements with a computer-generated grid of model stellar atmospheres.
A model stellar atmosphere may be defined as a mathematical construct or
description of a real stellar atmosphere, with specified input temperature, pressure,
density and chemical composition, that obeys a set of fundamental physical laws.
The simplest and most common models assume that there are no magnetic or
electric fields present, no bulk movements in the atmosphere (e.g., convection or
rotation), the ideal gas laws are obeyed and the atmosphere, (generally considered to
be plane parallel or spherically symmetrical) is in local thermodynamic equilibrium
and hydrostatic equilibrium.
The data input to the models include the effective temperature, the gravitational
acceleration at the surface and the chemical composition of the atmosphere, nor-
mally cited relative to that of the solar atmosphere. The model which best fits the
observed emergent stellar flux against wavelength or frequency values is taken to
give a reasonable indication of the most likely actual values of effective tempera-
ture, surface gravity and chemical composition. If the resolution of the grid spacing
is too coarse then the use of a subset of models to allow a refined interpolated group
of values should be used.

9.6.1 Kurucz model stellar atmospheres

A major grid of model stellar atmospheres was published by Kurucz (1979) for
stars of spectral type O, B, A, F and G with effective temperatures from 5500 K
to 50 000 K and chemical abundances 0.01 solar, 0.1 solar and solar values. An
update and extension of this work in 1993 by Kurucz59 was made to include stars
of later spectral type and hence lower surface temperature (down to 3000 K). The
wavelength range runs from 9 nm to 160 μm with a grid step from 1 nm at the short-
wavelength end to 40 nm at the long. The revised chemical abundance relative to
the Sun (log Z) varies from log Z = +1.0 (i.e., 10 times the solar abundance) to
log Z = −5.0 (0.000 01 times solar abundance). Surface gravity (log g) varies in
the range log g = 0.0 (0.5) 5.0.

59 See http://www.stsci.edu/hst/observatory/cdbs/k93models.html
166 Unit of thermodynamic temperature (kelvin)

The output of the Kurucz models is in the format (λ, fλ ), where λ is in nm and
Fλ is in erg . s−1 . cm−2 . Hz−1 . sr−1 .

9.6.2 Conversion of Kurucz model output from cgs to SI units

The output of wavelengths is either in nm or Å, so the conversion to SI units in m is
achieved by dividing by 109 or 1010 . The flux output unit conversion is a little more
complex, the first stage being to convert the component parts of the composite unit
from cgs to SI form.

erg . s−1 . cm−2 . Hz−1 . sr −1 = (10−7 ) J . s−1 . (10−2 )−2 m−2 . Hz−1 . sr −1
= 10−3 W . m−2 . Hz−1 . sr −1
= 1023 Jy . sr −1 (9.59)

The total emergent monochromatic flux from the star is 4 π times the unit stera-
dian value above. If the radius of the emitting star is ρ metres and its distance is
R , then the conversion factor, conv, between Kurucz units and janskys becomes:
 ρ 2
× 1023

conv = 4 π (9.60)
R
or if fJ is the monochromatic flux density in janskys and fK is the monochromatic
flux density in Kurucz units, then:
 ρ 2
. 1023 × fK

fJ = 4 π (9.61)
R

Determination of values for a star’s radius and its distance

As an observational determination of the radius of a star is at present not generally
available for the majority of stars, those that do have such measures have been used
to calibrate a relationship between spectral type and radius. For main sequence stars,
the values of radius vary from 1.029 × 1010 m for spectral type O5 to 1.325 × 108 m
for spectral type M5 according to Straižys (1992). In Table 9.4, the conversion
from solar radii to SI units was achieved using a value for the solar radius of
R = 9.955 08 × 108 m (Cox, 2000).
The determination of stellar distances is dealt with in Chapter 6.

Example: plot the comparison between the absolute HST spectrum of Vega
and a Kurucz model stellar atmosphere
By the absolute HST spectrum of Vega is meant the unreddened spectrum that
would be observed at a distance of 10 pc or 3.085 677 6 × 1017 m. The model stellar
 9.6 Model stellar atmospheres 167

Table 9.4. Stellar radii in solar radius and SI units against spectral class and
luminosity class

Spectral Radius Spectral Radius Spectral Radius

luminosity R luminosity R luminosity R
class m class m class m
O5V 14.791 O5III 17.783 O5Iab 22.909
1.029 d 10 1.237 d 10 1.593 d 10
B0V 7.244 B0III 10.965 B0Iab 25.119
5.039 d 9 7.626 d 9 1.747 d 10
A0V 2.291 A0III 3.631 A0Iab 63.096
1.593 d 9 2.525 d 9 4.388 d 10
F0V 1.413 F0III 2.570 F0Iab 114.82
9.824 d 8 1.788 d 9 7.986 d 10
G0V 1.072 G0III – G0Iab 245.47
7.453 d 8 – 1.707 d 11
K0V 0.977 K0III 10.00 K0Iab 389.05
6.797 d 8 6.955 d 9 2.706 d 11
M0V 0.603 M0III 43.65 M0Iab 524.81
4.191 d 8 3.036 d 10 3.650 d 11
M5V 0.191 M5III 109.6 M5Iab –
1.325 d 8 7.626 d 10 –

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .

atmosphere is also computed as though the star were at a distance of 10 pc. The
Kurucz model selected has an effective temperature of 9750 K, a surface gravity
of log g = 4.00 and solar chemical abundances (log Z = 0.00). The radius of an
A0V star is, from Table 9.4, 1.593 × 109 m. Using the conversion between Kurucz
units and janskys given in Equation (9.61) and inserting the values for the stel-
lar radius and its distance, plus a value for the monochromatic flux density of
4.93 × 106 erg . s−1 . cm−2 . Hz−1 . sr−1 from the Kurucz model at a wavelength of
3 μm (or a frequency of 1014 Hz) gives:
 1.593 × 109 2
fJ = 4 π × 1023 × 4.93 × 106
3.0856776 × 1017
= 165.03 Jy (9.62)
Figure 9.6 shows segments of the observed and theoretical spectra from 1014 Hz
to approximately 2 × 1015 Hz.

9.6.3 Mean monochromatic flux densities from Kurucz models

The Kurucz model stellar atmospheres are essentially step functions, so computing
the mean monochromatic flux density output, fν , in a given frequency band is
168 Unit of thermodynamic temperature (kelvin)

104
8
7
6
5
4
monochromatic flux density (Jy)

3
a
Veg
2

103
8 Hα
7
6

el
5
4 od
c zm
3 ru
Ku
2

102
2 3 4 5 6 7 8 2
1014 1015
frequency (Hz)

Figure 9.6. Comparison between the HST spectrum of the A0V star Vega and a
Kurucz model stellar atmosphere, with parameters T = 9 750 K, log g = 4.00 and
log Z = 0.00. The effective temperature was deliberately set too low to allow for
easier comparison between the model-generated spectrum lines and the actual HST
spectrum. Note that the monochromatic flux densities of Vega and the Kurucz
model have been adjusted to appear as though they were both at the standard
distance of 10 pc.

simply a matter of summing the product of the flux elements φν and the elemental
frequency bandwidths δν and dividing by the sum of the frequency bandwidth
elements, i.e.,

φ ν δν
fν =  (9.63)
δν
Since, in general, the maximum and minimum frequency values lie between tab-
ulated model frequencies, end corrections using simple linear interpolation were
made. Monochromatic flux densities were calculated for various photometric bands
from the satellite ultraviolet (the ANS and TD1 satellites), through the optical spec-
trum (Vilnius and Johnson photometry), to the infrared (USNO-B I band, 2MASS
and IRAS photometry). The number of model data points available per band var-
ied from 1 for the IRAS 60μm and 100μm bands to a maximum of 95 for the
 9.6 Model stellar atmospheres 169

Table 9.5. Wavelength frequency and mean monochromatic flux density data for
Kurucz model (Teff = 9750 K, log g = 4.00, log Z = 0.00) photometric bands

Band λ λ ν ν Fν
ID nm nm THz THz Jy nK
ANS1 154 5 1940 63 206.4 7
ANS2 154 14 1935 187 215.9 18
ANS3 179 14 1666 138 327.9 17
ANS4 220 20 1363 124 390.0 22
ANS5 249 15 1203 72 397.7 18
ANS6 329 10 910 28 623.4 7
TD1 156 33 1916 409 216.1 38
TD2 196 33 1526 258 367.6 36
TD3 236 33 1268 178 382.4 36
TD4 274 31 1094 124 466.2 32
U 345 40 869 101 650.3 23
P 374 26 802 56 968.4 15
X 405 22 740 40 1922.5 13
Y 466 26 643 36 2005.0 15
Z 516 21 581 24 1809.1 12
V 544 26 551 26 1718.5 14
S 656 20 457 14 1322.9 12
U 367 66 817 148 955.6 36
B 436 94 688 150 1980.5 48
V 545 88 550 89 1728.5 46
I 808 185 371 86 1132.5 95
J 1235 162 243 32 725.9 34
H 1662 251 180 28 459.3 34
KS 2159 262 139 17 302.7 28
L 3600 1200 83 29 126.9 71
M 4800 800 63 11 71.2 42
12 μm 12 000 7000 25 16 15.1 42
25 μm 25 000 11 150 12 6 3.0 3
60 μm 60 000 32 500 5 3 0.5 1
100 μm 100 000 31 500 3 1 0.2 1

USNO-B I band. Table 9.5 lists the band identification, mean wavelength λ, band-
width λ, mean band frequency ν, frequency bandwidth ν, the mean absolute
monochromatic flux density, Fν , emitted by the model star atmosphere at a distance
of 10 pc, and number nK of Kurucz model data points used in computing the mean
monochromatic flux density. Wavelength units are nm and frequency units THz
(1012 Hz).
Figure 9.7 illustrates the output for the USNO-B I band of the Kurucz model
stellar atmosphere with Teff = 9750 K, log g = 4.00 and log Z = 0.00. The model
170 Unit of thermodynamic temperature (kelvin)

1300

1200
monochromatic flux density (Jy)

1100

1000

900

800

32 34 36 38 40 42
13
frequency x 10 (Hz)

Figure 9.7. USNO-B I band mean monochromatic flux density and frequency
bandwidth against Kurucz model stellar atmosphere output. The νmin and νmax
values for each monochromatic flux density sample computed are shown, which
creates the step-function-like appearance. The construction of the I band used
95 sample fluxes from the Kurucz model for T = 9750 K, log g = 4.00 and
log Z = 0.00. The light-grey line shows the mean monochromatic flux measured in
the bandwidth shown by the horizontal light-grey line. The black square is located
at the mean frequency and mean monochromatic flux density of the band.

star is taken to be at a distance of 10 pc (3.085 × 1017 m). The single light-grey

line plotted shows the frequency bandwidth and the mean value (the black square)
of the monochromatic flux density in janskys. The prominent line formation at the
low-frequency end of the band is due to the Paschen series of hydrogen.
By way of an example, a plot is shown in Figure 9.8 comparing the observed
photometry of the star HD73952, a member of the young open star cluster
IC2391 (Dodd, 2007), with two Kurucz model stellar atmospheres, with the input
parameters log g = 4.00, log Z = 0.00 and Teff = 9750 K and Teff = 11 750 K. The
respective monochromatic flux density outputs have been adjusted to be those
that would be observed were the star (and the models) at a distance of 10 pc and
dereddened. The points shown with error bars are those derived from observation,
the point itself being at the mean frequency and the mean monochromatic flux
 9.6 Model stellar atmospheres 171

104
8
7 HD73952
6
5
4 Hα
monochromatic flux density (Jy)

2
K
750
11
=
103 T
8
7
6
5
4

3
T = 9750 K
2

102
2 3 4 5 6 7 8 2
1014 1015
frequency (Hz)

Figure 9.8. A comparison between two computed model stellar atmospheres, with
log g = 4.00, log Z = 0.00 and temperatures 9750 K and 11 750 K, with actual pho-
tometric observations of the B8Vn star HD73952. The error bars attached to each
observational point represent the photometric bandwidth and an estimate of the
total error in the photometry. Circles are TD1 photometry, squares are Vilnius pho-
tometry, the inverted triangle is the USNO-B I band and the diamonds 2MASS
photometry. The monochromatic flux densities in janskys are adjusted to those that
would be observed were the models and the star at the standard 10 pc distance.

density. The error bar in the x direction being the frequency bandwidth and that
in the y direction being due to the uncertainty in the observational photometric
measurement.

9.6.4 Selection of best fitting model

Computing the weighted mean and weighted standard deviation of the differences
between the Kurucz model monochromatic flux density in a given photometric
band and the observational value was not used, given the wide variation in the
monochromatic flux density with frequency that could readily bias the mean towards
the difference value of the band with the greatest signal. Instead, the weighted
mean of the ratios of the observational to model monochromatic flux densities
were used. The weighting applied was the ratio of the measured average signal
strength to the standard deviation about that mean due to contributing variables,
172 Unit of thermodynamic temperature (kelvin)

such as photometric measurements, distance measurements, the determination of

the interstellar reddening value and the absolute calibration of the photometric
system.
If Ki is taken to be the Kurucz monochromatic flux density at the mean band
frequency νi and Oi to be the observed value with a measured standard deviation
about that value of σi , then the weighted mean ratio, R, between K and O for all
n measured photometric bands is:
n Oi2
i=1 σi Ki
R = n Oi
(9.64)
i=1 σi
and the weighted standard deviation, σR is:
 
 n Oi Oi 2
 i=1 σi ( Ki − R)
σ =   (9.65)
R n−1 n Oi
n i=1 σi

Example: estimating temperature using the best fitting model

stellar atmosphere
Using this technique on HD73952, a set of weighted χ 2 values against different
effective temperatures of Kurucz models with the same values of log g and log Z
may be determined. A second-order fit to the curve is then made and the minimum
value of weighted χ 2 parameter used to find the corresponding value of Teff .
In Figure 9.9 this may be seen to be approximately 10 900 K. Straižys (1992)
assigns that temperature to a star of type B8–9V, which would seem a good start in
determining a more precise temperature value using the model stellar atmospheres
in an iterative manner. (SIMBAD lists HD73952 to be spectral type B8Vn.)

9.7 Summary and recommendations

9.7.1 Summary
The SI unit of thermodynamic temperature, the kelvin, has been developed and
refined in parallel with other, non-absolute scales, such as those of Celsius and
Fahrenheit and one other absolute scale due to Rankine. Conversion formulae
between these systems are given, with examples. Sample temperatures of vari-
ous astronomical locations from the surface of the Earth to the cosmic background
radiation due to the Big Bang are given in kelvin.
The many different types of temperature, such as thermodynamic, kinetic, effec-
tive, Planck, blackbody, antenna, brightness and colour are defined and worked
examples of temperatures derived using appropriate observational techniques
(radiometry, photometry, spectrophotometry) given. The possibility of using a
temperature sequence in place of spectral classification is discussed in some detail.
 9.7 Summary and recommendations 173

weighted chi squared parameter 400

300

200

100

10 000 10 500 11 000 11 500 12 000

Teff (K)

Figure 9.9. A plot of a weighted χ 2 parameter against effective temperatures from

Kurucz model stellar atmospheres. The black points are actual values, the grey lines
merely link up the points and the solid black line is a second-order least squares
fit between the two variables. The minimum value of the weighted χ 2 parameter
corresponds to an effective temperature of 10 919 K.

The final section illustrates how model stellar atmospheres in conjunction with
coarse (broadband and intermediate-band photometry) and standard spectropho-
tometry may be used to calculate stellar temperatures.

9.7.2 Recommendations
The basic recommendation to most astronomers would be to carry on with what
they are doing now, as most already use the SI unit, the kelvin, as a measure of
temperature.
Some popular astronomical publications often cite temperatures in degrees Cel-
sius or Fahrenheit and should be encouraged to show the equivalent temperature
in kelvin in parenthesis. This would lead to an increasing familiarity with kelvin
amongst the interested laity.
Whilst it is possible to replace spectral class and dereddened colour indices by
a measure of temperature, there would seem to be no really good reason for doing
so, though it may be argued that for ease of comparison between observations and
theory, similar variables should be used.
 10
Unit of electric current (ampere)

10.1 SI definition of the ampere

The ampere is that constant current which, if maintained in two straight parallel
conductors of infinite length, of negligible cross section, and placed 1 metre apart
in vacuum, would produce between these conductors a force equal to 2 × 10−7
newton per metre of length.
The dimension of electric current is [I ], its unit is the ampere and its symbol
is A.

10.1.1 Possible future definition of the ampere

Under discussion at the present time is the redefining of the unit of electric current
in the following way:

The ampere is a unit of electric current such that the elementary charge is exactly
1.602 176 53 × 10−19 C, where 1 C (coulomb) = 1 A . s (ampere second).

10.1.2 Definition of electricity

Funk & Wagnalls New Standard Dictionary of the English Language (1946) gives
a general definition of electricity as:

Electricity: A material agency which, when in motion, exhibits magnetic, chemical and
thermal effects, and which, whether in motion or at rest, is of such a nature that when it is
present in two or more localities within certain limits of association, a mutual interaction
of force between such localities is observed.

And as a physics-related definition of electricity:

That branch of science that treats of this agency and the phenomena caused by it.

174
 10.2 SI and non-SI electrical and magnetic unit relationships 175

10.1.3 Definition of magnetism

Funk et al. (1946) give, as a general definition of magnetism:
Magnetism: that quality or agency by virtue of which certain bodies are productive of
magnetic force or susceptible to its action.

And as a physics-related defintion of magnetism:

The science that treats the laws and conditions of magnetic force.

10.2 SI and non-SI electrical and magnetic unit relationships

There are five major systems of electrical and magnetic units in use at present,
namely: the electrostatic system (esu), the electromagnetic system (emu), the cgs
Gaussian, the practical system and the SI system. Of these, the first three are based
on the cgs system for all mechanical quantities, but differ in that the esu system
is additionally based on the law of force between electric charges, whilst the emu
system is based on the corresponding law of force between magnetic poles with the
unrationalized cgs Gaussian being a hybrid using both esu and emu units.Adifficulty
arises in everyday use of either the esu or emu system in that particular units are
considered to be either too small or too large, so a further system, the practical
system, was developed. The practical system is based on the metre, kilogram,
second and ampere, similar to, but not identical with, the SI. The SI naturally uses
the metre, kilogram and second as the fundamental mechanical units, in conjunction
with the ampere as a fourth unit to define electrical and magnetic quantities.
The IAU recommends the use of the SI units only and particularly deprecates
(Wilkins, 1989) the use of the gauss or the gamma for magnetic flux densities
and the oersted for magnetic field strength. Burke & Graham-Smith (2002) state
in a footnote on page 107 of their radio astronomy textbook: ‘The usual unit in
astrophysics (for magnetic fields): 1 gauss = 10−4 tesla.’The rest of this section will
therefore consider mainly the relationship between the unrationalized cgs Gaussian
and the SI unit systems.

SI and unrationalized cgs Gaussian base units

Electric and magnetic derived units in the SI system are constructed from the four
base units: the kilogram, the metre, the second and the ampere, with dimensions
[M ], [L], [T ] and [I ].
The unrationalized cgs Gaussian system uses three base units: the gram, the
centimetre and the second only, and derives all electric and magnetic units from
appropriate combinations of them based on physical laws relating electromagnetism
to mechanics. Hence, the dimensional form of electric and magnetic units in the
cgs Gaussian system is [M ], [L] and [T ].
176 Unit of electric current (ampere)

So the SI system for electric and magnetic units may be thought of as

four dimensional and that of the unrationalized cgs Gaussian system as three
dimensional.

Example: derive the dimension of the unit of charge in both the

unrationalized cgs Gaussian system and that of the SI
In the unrationalized cgs Gaussian system, Coulomb’s law may be written as:
q1 q2
fG = CG (10.1)
r2
where fG is the force in dynes between two electric charges q1 and q2 , measured in
esu statcoulombs or franklins, r cm apart and CG , is a constant of proportionality.
CG is fixed, in the unrationalized case, by defining the force between two unit
charges (1 Fr) separated by 1 cm to be 1 dyne, in which case CG = 1. So Coulomb’s
law may be rewritten as:
q1 q2
fG = 2 (10.2)
r
by rearranging Equation (10.2) and taking q1 = q2 = q:

q 2 = fG r 2 (10.3)

so
1
q = fG2 r (10.4)
In dimensional terms, force and length may be expressed as:

dim[fG ] = [M] . [L] . [T ]−2 (10.5)

dim[r] = [L] (10.6)

so the dimension of the unrationalized cgs Gaussian unit of charge dim[q] is:
1
dim[q] = ([L] . [M] . [T ]−2 ) 2 [L]
1 3
= [M] 2 . [L] 2 . [T ]−1 (10.7)

In the SI case, charge Q, in coulombs, is defined to be the product of time in seconds

and the SI base unit, the ampere, so that:

dim[Q] = [I ] . [T ] (10.8)

In Table 10.1, a selection of named electric and magnetic derived units with
dimensions and combinations of base units are given for both SI and unrationalized
cgs Gaussian units.
The last two entries in Table 10.1 may be defined in the following manner.
 10.2 SI and non-SI electrical and magnetic unit relationships 177

Table 10.1. Names, symbols and dimensions of various electric and magnetic
units in SI and unrationalized cgs Gaussian form

Unit of Name Symbol Dimension

Electric current SI ampere A [I ]
3 1
cgsG esu statamp statA [L] 2 . [M] 2 . [T ]−2
Electric charge SI coulomb C =A. s [I ] . [T ]
3 1
cgsG franklin Fr [L] 2 . [M] 2 . [T ]−1
Magnetic field ampere/metre A . m−1 [I ] . [L]−1
strength SI
1 1
cgsG oersted Oe [L] 2 . [M] 2 . [T ]−2
Electric potential volt V [L]2 . [M] . [T ]−3 . [I ]−1
difference SI
1 1
cgsG esu statvolt statV [L] 2 . [M] 2 . [T ]−1
Capacitance SI farad F [L]−2 . [M]−1 . [T ]4 . [I ]2
cgsG (esu) cm cm [L]
Electric resistance SI ohm  [L]2 . [M] . [T ]−3 . [I ]−2
cgsG esu statohm stat [L]−1 . [T ]
Magnetic flux SI weber Wb [L]2 . [M] . [T ]−2 . [I ]−1
1 1
cgsG emu maxwell Mx [L] 2 . [M] 2
Magnetic flux tesla T [M] . [T ]−2 . [I ]−1
density SI
3 1
cgsG emu gauss G [L]− 2 . [M] 2
Inductance SI henry H [L]2 . [M] . [T ]−2 . [I ]−2
cgsG esu stathenry statH [L]−1 . [T ]2
Electric field volt/metre V . m−1 [L] . [M] . [T ]−3 . [I ]−1
strength SI
1 1
cgsG esu statvolt/cm V . cm−1 [L]− 2 . [M] 2 . [T ]−1
Electric flux coulomb/sq metre C . m−2 [L]−2 . [T ] . [I ]
density SI
3 1
cgsG esu statcoulomb/sq cm statC . cm−2 [L] 2 . [M] 2 . [T ]−1
Permittivity SI farad/metre 0 [L]−3 . [M]−1 . [T ]4 . [I ]2
of free space
Magnetic permeability henry/metre μ0 [L] . [M] . [T ]−2 . [I ]−2
SI of free space

The permittivity of free space, 0 , forms part of the constant of proportionality

CSI , in the SI statement of Coulomb’s law, i.e.:
q1 q2 1 q1 q2
fSI = CSI 2 = (10.9)
r 4 π 0 r 2
1
CSI = (10.10)
4 π 0
178 Unit of electric current (ampere)

where fSI is the force in joules between two charges q1 and q2 , measured in
coulombs, separated by a distance of r metres. The appearance of the factor 4 π
is due to SI units being rationalized. In general, for equations set out in rational-
ized units, 4 π appears in those involving spherical symmetry, 2 π in those with
cylindrical symmetry and none in those with plane symmetry.
The magnetic permeability of free space, μ0 , forms part of the constant of pro-
portionality ( 4μπ0 ) in the equation relating the magnetic force fM in joules, between
two current elements i1 and i2 in amperes, in conductors 1 and 2, of length dl1 and
dl2 metres, separated by distance r1,2 metres, where r1,2 is measured along a line
which makes an angle θ in radians to the direction of the flow of the current in
conductor 1.
μ0 −2
dfM = r i1 dl1 i2 dl2 sin θ (10.11)
4 π 1,2
Note that the constants 0 and μ0 are related in SI units by the expression:
1
0 μ0 = (10.12)
c2
where c is the velocity of light (Bleaney & Bleaney, 1962).

10.2.1 A comparison between some electric and magnetic mathematical

relationships in the SI and cgs Gaussian system
The various laws and definitions in the theory of electricity and magnetism take
different forms depending on the units in which the variables and constants in
the mathematical equations are expressed. Symbols used in Table 10.2 have the
following meanings, where normal text symbols refer to scalar quantities and bold-
face symbols to vector quantities:

q : electric charge J : electric current density

i : electric current E : electric intensity
 : permittivity B : magnetic induction
0 : permittivity of free space F : force
μ : magnetic permeability D : electric displacement
μ0 : permeability of free space H : magnetic field intensity
c : velocity of light in vacuo V : velocity of charge element
ρ : charge density P : polarization
σ : conductivity M : magnetization

For a more complete treatment of the above material see, e.g., Coulson & Boyd
(1979), Bleaney & Bleaney (1962) and Menzel (1960).
 10.2 SI and non-SI electrical and magnetic unit relationships 179

Table 10.2. Some electrical and magnetic mathematical definitions in SI and

unrationalized cgs Gaussian form

Name SI form cgs Gaussian form

Force field dF = E + V ∧ B dF = E + V∧ B
dq dq c
Electric field D = 0 E D=E
0 μ0 = c−2 0 μ0 = 1 
Polarization P = D − 0 E P = 41π D − 0 E
 
Magnetization M = μB − H M = 41π μB − H
0 0
Ampère’s law ∇ ∧ H = J + ∂∂D
t ∇ ∧ H = 4 πc J + 1c ∂∂D
t
Maxwell’s div D = ρ div D = 4 πρ
equations
div B = 0 div B = 0
curl E = − ∂B curl E = − 1c ∂B
∂t  ∂t 
curl H = σ E + ∂∂D
t curl H = 1c 4 π σ E + ∂∂D
   2 t 
Wave equation ∇ 2 (E , H) = μ μ0  0 ∂ 2 (E , H) μ
∇ 2 (E , H) = 2 ∂ 2 (E , H)
2
∂t c ∂t

10.2.2 Converting unrationalized cgs Gaussian units to SI units

Example 1: determine the relationship between the SI unit of charge, the
coulomb, and the unrationalized cgs Gaussian unit of charge, the statcoulomb
First, how many statcoulombs per charge are required to produce a force fG of 1 N
when the equal-valued charges qG are separated, rG = 1 m, apart?
The cgs Gaussian version of Coulomb’s law states that:

qG2
fG = (10.13)
rG2

substitute rG = 100 cm and fG = 105 dynes ( = 1 newton) and rearrange Equation

(10.13) to give:

qG2 = rG2 fG
= 1002 × 105 = 109
9
qG = 10 2 statcoulombs (10.14)

Second, how many coulombs per charge, qSI , are required to produce a force fSI
of 1 N when the charges are rSI = 1 m apart?
180 Unit of electric current (ampere)

The SI version of Coulomb’s law states that:

2
1 qSI
fSI = 2
(10.15)
4 π 0 rSI

Substitute rSI = 1 m and fSI = 1 N and rearrange Equation (10.15) to give:

2 1
qSI = r 2 fSI
4 π 0 SI
1
= (10.16)
4 π 0
Now 0 is a quantity defined by:
1
0 = (10.17)
μ0 c2
where c, the speed of light in vacuo, is a defined constant, as is μ0 , the permeability
of free space, hence 0 also has a fixed value. Substituting for μ0 ( = 4 π × 10−7 )
gives:
1
0 =
4 π × 10−7 c2
1
4 π 0 = −7 2
10 c
1 1
√ = (107 c−2 ) 2 (10.18)
(4 π 0 )

Comparing the values for the identical charges in the cgs Gaussian and SI units
gives:
7 9
10 2 c−1 (coulombs) ≡ 10 2 statcoulombs
9 7
10 2 − 2
1 C = −1 (statcoulombs)
c
= 10 c
= 2.997 924 58 × 109 statcoulombs (10.19)

by inverting Equation (10.19):

1
1 statcoulomb = = 3.335 640 95 × 10−10 C (10.20)
10 c
 10.2 SI and non-SI electrical and magnetic unit relationships 181

Example 2: determine the relationship between the SI unit of magnetic flux

density, the tesla (T), and the cgs Gaussian unit, the gauss (G)
First some definitions:

Magnetic flux, , is usually defined in terms of magnetic flux lines as the group
or number of such lines emitted outwards from the north pole of a magnetic body.
Its unit is the weber and its symbol Wb.

Magnetic flux density, B, is defined to be the amount of magnetic flux per unit
area perpendicular to the direction of the magnetic flux. Its unit is the tesla and its
symbol T. These quantities are related by the simple expression:


B= (10.21)
A

In SI units, when A is in m2 and  is in Wb, then B is in T (teslas). In

cgs Gaussian units, if A is in cm2 and  is in Mx (maxwells), then B is calcu-
lated in G (gauss). The geometry of the variable in Equation (10.21) is shown in
Figure 10.1.

MAGNETIC
FLUX

Φ
θ

Figure 10.1. Area A of solid-line rectangle is inclined by θ rad to the perpendicular

to the direction of the magnetic flux and presents an area A cos θ to the magnetic
flux and hence the magnetic flux density B, through A cos θ , is B =  / A cos θ .
The dotted axes (dark grey) run with and at right angles to the direction of the
magnetic flux.
182 Unit of electric current (ampere)

Next check the dimensional consistency between gauss and tesla, the dimensions
for each unit are given in Table 10.1 as:
1 3
dim[gauss] = [M] 2 . [L]− 2 (10.22)
dim[tesla] = [M] . [T ]−2 . [I ]−1 (10.23)
Also from Table 10.1, the cgs Gaussian dimension for electric current IG is given as:
1 3
dim[I ]G = [M] 2 . [L] 2 . [T ]−2 (10.24)
substitute for [I ] in Equation (10.23) to give:
1 3
dim[tesla] = [M] . [T ]−2 ([M] 2 . [L] 2 . [T ]−2 )−1
1 3
= [M] 2 . [L]− 2
= dim[gauss] (10.25)
so the gauss and the tesla are dimensionally consistent.
Mass in cgs Gaussian units is measured in grams (g) and length in centimetres
(cm); in SI units, mass is measured in kilograms (kg) and length in metres (m).
For a magnetic flux of 1 weber (Wb) passing perpendicularly through an area of
1 square metre (m2 ), the magnetic flux density is BSI = 1 T (tesla).
For a similar magnetic flux of 1 Wb (= 108 Mx (maxwells) in cgs Gaussian units)
passing through a 1 cm2 (= 10−4 m2 ) area, the magnetic flux density is given by:
BG = 10−4 T = 1 G (10.26)
and it follows that:
1 Wb = 1 T . m2
= 1 . (104 G) . (104 cm2 )
= 108 G . cm2
= 108 Mx (10.27)
A cgs Gaussian unit of magnetic flux density that was in common use in
astrophysics was the gamma γ , a submultiple of the gaussian given by:
1 γ = 10−5 G = 10−9 T (10.28)
A listing of transformations and conversions for other electric and magnetic units
from cgs Gaussian to SI is given in Table 10.3.
An example of an online magnetic unit converter is given at: http://www.
smpspowersupply.com/magnetic-unit-conversion.html.
 10.3 Magnetic fields in astronomy 183

Table 10.3. Names, symbols and dimensions of various electric and magnetic
units in SI and unrationalized cgs Gaussian form

Transformation cgs Gaussian SI unit and

Quantity Symbol from cgsG to SI unit conversion factor
√ 1
Charge Q (4 π 0 ) Q 1 statcoulomb 10 c C
√ 105 C . m−3
Charge density ρ (4 π 0 ) ρ 1 statcoulomb . cm−3 c
√ 1
Electric current I (4 π 0 ) I 1 statamp 10 c A
√ 103 A . m−2
Current density j (4 π 0 ) j 1 statamp . cm−2 c
Potential φ √ φ 1 statvolt 300 V
(4 π 0 )
Electric field E √ E 1 statvolt . cm−1 3 × 104 V . m−1
(4 π 0 )
Magnetic flux  1 Mx 10−8 Wb
√ B
Magnetic flux B 1G 10−4 T
(4 π μ0 )
density
√ A
Vector potential A 1 G . cm−1 102 c Wb . m−1
(4 π μ0 )
√ 103
Polarization P (4 π 0 ) P 1 statvolt . cm−1 C . m−2
c
√ 0 103
Electric D D 1 statvolt . cm−1 C . m−2
displacement 4π 4π c
0 107
Permittivity  1 statfarad . cm−1 F . m−1
 4 π c2
√ 4π
Magnetization M M 1G 4 π × 10−4 A . m−1
μ0
√ 1 103
Magnetic field H H 1 Oe A . m−1
4 π μ0 4π
μ0
Permeability μ 1 G . Oe−1 4 π × 10−7 H . m−1
μ
4 π 0 107
Electrical σ 1 s−1 mho . m−1
conductivity σ c2

10.3 Magnetic fields in astronomy

The discovery of magnetic fields throughout space associated with stars, planets, the
interplanetary and interstellar media has led to the realization of the importance of
incorporating such fields and their effects in the modelling of the universe in general,
and individual celestial bodies and their formation and evolution in particular.
As Table 10.4 shows, the size range of magnetic fields found in the cosmos is
enormous: from 10−10 T in interstellar space to 2.1 × 1011 T for a particular type of
highly evolved star known as a magnetar. The bulk of the magnetic flux densities
184 Unit of electric current (ampere)

Table 10.4. Magnetic flux densities (B) of various celestial objects in cgs
Gaussian and SI units

BG BSI
Object Notes G T
Solar System
Interplanetary space 1 d −6 − 1 d −5 1 d −10 − 1 d −9
Mercury equatorial value 3.4 d −3 3.4 d −7
Venus <4 d −6 <4 d −10
Earth 0.31 3.1 d −5
Mars <5 d −6 <5 d −10
Jupiter 4.3 4.3 d −4
Saturn 0.22 2.2 d −5
Uranus 0.23 2.3 d −5
Neptune 1.4 1.4 d −4
Sun
Sunspot penumbra 0.8 d 3 − 2.0 d 3 0.08 − 0.2
umbra 2d3−4d3 0.2 − 0.4
Plage 1.4 d 3 − 1.7 d 3 0.14 − 0.17
Prominences horizontal field 2 − 40 2 d −4 − 4 d −3
Corona 1 d −5 − 1 d 2 1 d −9 − 1 d −2
Interstellar space
general field 1 d −6 − 1 d −5 1 d −10 − 1 d −9
Stars
Vega A0V star 0.6 6 d −5
FK Com cool giant 60 − 272 6 d −3 − 2.7 d −2
Be 10 − 100 1 d −3 − 1 d −2
WR 1500 0.15
Flare ≥1000 ≥0.1
T Tau 1d3 − 1d4 0.1 − 1
WD 1d5 − 1d9 10 − 1 d 5
Neutron stars 1 d 12 − 1 d 13 1d8 − 1d9
Pulsar surface field 1 d 12 1d8
SGR 1806 -20 magnetar 2.1 d 15 2.1 d 11
Gaseous nebulae
Planetary 1 d −4 − 1 d −3 1 d −8 − 1 d −7
SNR 1 d −5 − 1 d −2 1 d −9 − 1 d −6
Galaxy
z = 0.692 8.4 d −5 8.4 d −9

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .

published are in the cgs Gaussian unit, the gauss, though a few are presented in the
SI unit, the tesla. Both units are given in Table 10.4.
Table 10.4 was compiled, in part, from the following sources: Cox (2000), Land-
street (2001), Gnedin (1997), Reddish (1978), Bemporad & Mancusco (2010),
 10.3 Magnetic fields in astronomy 185

Table 10.5. Solar total magnetic flux in unrationalized cgs

Gaussian and SI units

cgs Gaussian SI
Total magnetic flux at/in Mx Wb
Solar minimum 1.5 − 2.0 d 23 1.5 − 2.0 d 15
Solar maximum 1.0 − 1.2 d 24 1.0 − 1.2 d 16
Small active region 3 d 20 3 d 12
Large active region ≥1 d 22 ≥1 d 14

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .

Table 10.6. Planetary magnetic dipole fields in

unrationalized cgs Gaussian and SI units

cgs Gaussian SI
Planet G . cm−3 T . m−3
Mercury 2 − 6 d 22 2 − 6 d 12
Venus <1 d 21 <1 d 11
Earth 7.84 d 25 7.84 d 15
Mars <1 d 22 <1 d 12
Jupiter 1.55 d 30 1.55 d 20
Saturn 4.6 d 28 4.6 d 18
Uranus 3.9 d 27 3.9 d 17
Neptune 2.2 d 27 2.2 d 17

In this table the shorthand notation m d n has been used where m d n ≡ m × 10n .

Lignières et al. (2009), Korhonen et al. (2009) and Wolfe et al. (2008). The surface
magnetic flux density of the magnetar was extracted from the McGill SGR/AXP
Online-Catalog.60
Other than magnetic flux density, Cox (2000) also lists examples of total magnetic
flux for the Sun and active solar regions (see Table 10.5) and planetary magnetic
dipole fields (see Table 10.6).
The relationship between cgs Gaussian units and SI units in Table 10.5 is simply:
1 Wb = 108 Mx.
In Table 10.6, the relationship between the cgs Gaussian unit and the SI unit is:

1 T . m3 = 10−10 G . cm3 (10.29)

this follows from the relations 1 G = 10−4 T and 1 cm = 10−2 m.

60 See http://www.physics.mcgill.ca/∼pulsar/magnetar/main.html
186 Unit of electric current (ampere)

10.3.1 Measurement of astronomical magnetic fields

Until recently, the majority of astronomical magnetic fields were detected and
measured using the Zeeman effect in the observed spectrum of the celestial object.
For bodies in the Solar System, including the Sun itself, magnetometers carried
by various space probes have been able to measure directly their magnetic fields.
At radio frequencies, techniques have been developed that depend on the way in
which electrons behave in the presence of a static magnetic field to measure the
strength of the field (see Burke & Graham-Smith, 2002).

Zeeman effect
The effect of a magnetic field of flux density B on the spectrum of an atom with
singlet lines (S = 0 where S is the spin value) is to split such lines into 2 J + 1
(where J is the total angular momentum) equally spaced lines either side of the
original singlet line. In cgs Gaussian units, the energies EM , in ergs, of the levels
are given by:
ehB M
EM = E0 ± (10.30)
4 π me c
where E0 is the energy level of the zero field singlet, e is the charge of the electron
in statcoulombs, where:
1 C = 10 c stat C (10.31)

me is the rest mass of the electron in grams, h is Planck’s constant in erg . s, c is the
speed of light in cm . s−1 , B is the magnetic flux density in gauss, and M takes an
integer value of 0 or ± 1 (see Lang 2006).
Since:
EM = h νM (10.32)

Equation (10.31) may be rewritten in terms of frequency ν, as (Aller, 1963):

eB M
ν M = ν0 ± (10.33)
4 π me c

where νM is the frequency of the singlet line split by the magnetic field and ν0 is
the frequency of the zero field singlet. Now, Equation (10.34) may be rewritten in
SI units as:
eB M
νM = ν0 ± (10.34)
4 π me
 10.3 Magnetic fields in astronomy 187

where ν is in hertz, e is the charge of an electron in coulombs, me is the rest mass

of an electron in kilograms and B is the magnetic flux density in teslas.

Example: show that the dimension of the shift due to the Zeeman splitting
of a singlet spectral line in both cgs Gaussian and SI units is the same
(a) cgs Gaussian case:
eG BG
νG = (10.35)
4 π mG cG
Use Table 10.1 to give the dimensions for each of the units:
dim[eG ] . dim[BG ]
dim[ νG ] =
dim[mG ] . dim[cG ]
3 1 3 1
([L] 2 . [M] 2 ] . [T ]−1 ) . ([L]− 2 . [M] 2 . [L] . [T ]−1 )
=
[L] . [M] . [T ]−1
= [T ]−1 (10.36)

The unit which has a dimension of [T ]−1 is the hertz, the unit of frequency.
(b) SI case:
eSI BSI
νSI = (10.37)
4 π mSI
Again, using Table 10.1:
dim[eSI ] . dim[BSI ]
dim[ νSI ] =
dim[mSI ]
([I ] . [T ]) . ([M] . [T ]−2 . [I ]−1 )
=
[M]
= [T ]−1 (10.38)

so the SI unit has a dimension of [T ]−1 , showing dimensional consistency between

the two systems of units.

Example: determine the numerical value for the coefficient bSI in

Equation (10.39)
eSI
bSI = (10.39)
4 π mSI
Substitute SI values for eSI and mSI in Equation (10.40):

1.602 189 072 × 10−19

bSI =
4 π 9.109 389 7 × 10−31
= 1.399 634 438 × 1010 (10.40)
188 Unit of electric current (ampere)

So the frequencies of the singlet line and the first pair of lines either side due to the
presence of a magnetic field of flux density B T are given by:

νSI = ν0 ± bSI B (10.41)

Example: at the rest frequency of the 21-cm HI line, what would be the
approximate amount of Zeeman splitting resulting from a background
magnetic flux density of 1 nT, similar to that found in the Milky Way Galaxy?
(a) SI case:
ν21 = 1.420 405 751 768 × 109 Hz (Cox, 2000) is the radio frequency of the HI line,
which corresponds approximately to a wavelength of 21.106 114 cm. The frequency
separation, in a magnetic field of flux density of 10−9 T, is 2 ν21 , and may be
written as:
eSI BSI
2 νSI = 2
4 π mSI
(1.602 × 10−19 ) (10−9 )
2
4 π (9.109 × 10−31 )
 28 Hz (10.42)

(b) cgs Gaussian case:

In cgs Gaussian units, the magnetic flux density, BG is equal to 10−5 G:
eG B G
2 νG = 2
4 π mG c
(4.805 × 10−10 ) (10−5 )
2
4 π (9.109 × 10−28 ) (2.9979 × 1010 )
 28 Hz (10.43)

The Zeeman effect in wavelength units

The relationship between the frequency ν, and wavelength λ, of a spectral feature
is given by:
c
ν= (10.44)
λ
where c is the speed of light. Differentiate ν with respect to λ then:
dν c
=− 2 (10.45)
dλ λ
Replace the differentials (d ν, d λ) by small differences ( ν, λ):
c
ν = ∓ λ (10.46)
λ2
 10.3 Magnetic fields in astronomy 189

or
λ2
λ = ∓ ν (10.47)
c
Now, in SI units, the frequency shift νSI , in hertz, is given by Equation (10.38),
substitute λSI for νSI thus:
λ20
λSI = ∓ νSI
c
λ2 eSI
=∓ 0 BSI (10.48)
4 π mSI c
where λ0 is the wavelength in metres of the spectral feature in the absence of a
magnetic field and λ0 ∓ λSI are the central wavelengths, in metres, of the first
Zeeman components formed in the presence of a magnetic field of flux density BSI
in teslas, with eSI being the charge of the electron in coulombs and mSI the mass
of the electron in kilograms.
If cgs Gaussian units are used with the charge of the electron eG measured in esu
charge units (statcoulombs), the electron mass mG in grams, the wavelengths λ in
centimetres, and the speed of light in cm . s−1 , then by substituting for νG from
Equation (10.36), Equation (10.48) gives:
λ20 eG
λG = ∓ BG (10.49)
4 π mG c 2
where BG is the magnetic flux density in gauss.
In optical astronomy, the variables that are measured are λ0 and λ, and that
which is to be determined is B. Equations (10.49) and (10.50) may then be
rewritten as:
4 π c mSI λSI
BSI = (10.50)
λ20 eSI
and
4 π c2 mG λG
BG = (10.51)
λ20 eG
Example: determine the global surface magnetic flux density of the magnetic
A-type star HD94660 in both teslas and gauss
(a) SI case:
Approximate measurements were made of a small portion of the spectrum of
HD94660 between 6 144 Å and 6 154 Å (as presented by J. D. Landstreet at the
Leverhulme lecture on stellar magnetism.). The Zeeman doublet of interest in this
narrow spectral region is due to singly ionized iron (FeII), with a laboratory-
measured central wavelength of 6 149.238 Å (Moore, 1959). The separation of
190 Unit of electric current (ampere)

the minima of the Zeeman components was 2 λ = 0.32 Å . In SI units, λ0 =

6.1492 × 10−7 m and λ = 1.60 × 10−11 m; substitute these values into Equation
(10.49) with the SI values for c, m and e:

4 π . (9.109 × 10−31 ) . (2.998 × 108 ) . (1.6 × 10−11 )

BSI =
(6.1492 × 10−7 )2 . (1.602 × 10−19 )
= 0.91 T (10.52)

(b) cgs Gaussian case:

In the cgs Gaussian case, wavelength is measured in cm, the speed of light in
cm . s−1 , electron mass in g and electron charge in esu or statC, so Equation (10.50)
becomes:

4 π . (9.109 × 10−28 ) . (2.998 × 1010 )2 . (1.6 × 10−9 )

BG =
(6.1492 × 10−5 )2 . (4.805 × 10−10 )
= 9100 G (10.53)

Since 1 T = 104 G, the global magnetic field values in each set of units are equal
to one another and may be compared with the value of − 2200 G (= − 0.22 T) for
the mean longitudinal magnetic field obtained by Bagnulo et al. (2002) using the
FORS1 spectropolarimeter at the ESO VLT.

10.3.2 Magnetic fields of Solar System objects and interplanetary space

from spaceborne magnetometers
The exploration of the magnetic fields of the Solar System is somewhat different
from that of the rest of the universe in that it has recently become possible to
send space probes to sample the fields directly rather than by using spectroscopy
or spectropolarimetry. The instrument that is used to carry out measurements to
determine the intensity and orientation of magnetic flux lines is known as a fluxgate
magnetometer.61
The Mercury orbiting space probe MESSENGER carries such a magnetometer,
which was able to measure the direction and magnitude of the planet’s magnetic
field during the January and October 2008 flybys (see Alexeev et al. 2010 and
Anderson et al. 2010). Both cited papers use SI units for the equatorial surface
field (given in nT) and the latter, Anderson et al. (2010), also presents the values
for dipole and multipole fields in units of nT . RM 3 , where R is the radius of
M
Mercury.

61 See, e.g., http://whatis.techtarget.com/definition/

 10.3 Magnetic fields in astronomy 191

10.3.3 Faraday rotation

The plane of polarization of an electromagnetic wave is rotated in the presence of
a magnetic field, with a component along the direction of propagation of the wave.
This is known as the Faraday effect and the rotation of the polarization plane as
Faraday rotation.
Rotation measure, Rm , may be defined as an integral along the line of sight from
the celestial object being observed to the observer at a distance of r pc, of the
product of the electron density ne , and the line-of-sight component of the general
magnetic field B|| (the integral numerator in Equation 10.54).
Dispersion measure, Dm , is another integral along the line of sight, this time of
the electron density ne alone, from the celestial body of interest to the observer,
again over a distance of r pc (the integral denominator in Equation 10.54).
Since the magnitude of the Faraday rotation is proportional to the product of the
electron density and the magnetic field component along the line of sight, the ratio
of the rotation measure to the dispersion measure yields the flux density of this
component of the magnetic field.
Burke & Graham-Smith (2002) and Weisberg et al. (2004) define the magnitude
of the mean line-of-sight component of the magnetic field, <B|| >, in cgs Gaussian
units, along that line from the source celestial object (e.g., a pulsar), to the observer,
r pc distant on Earth, to be:
 s=r
ne B . ds
< B|| > = s=0s=r (10.54)
ne ds
s=0
R 
= 1.232 × 10−6
m
G (10.55)
Dm
where Rm is measured in rad . m−2 and Dm is measured in pc . cm−3 . Note that
whilst the dispersion measure unit may also be thought of as cm−2 , the common
expression used by astronomers is pc . cm−3 . In SI units the dispersion measure
has units of m−2 (≡ m . m−3 ) or, when IAU approved units are used, the dispersion
measure is measured in pc . m−3 .
In cgs Gaussian units, the dimension of the quotient Rm / Dm in Equation (10.55)
is given by:
dim[Rm ] [L] . [L]−1 . [L]−2
=
dim[Dm ] [L] . [L]−3
=1 (10.56)
192 Unit of electric current (ampere)
3 1
and since dim[B|| ] = [L]− 2 . [M] 2 , then the dimension of the coefficient in Equation
3 1
(10.55) must also be [L]− 2 . [M] 2 to balance the dimensional equation.
In SI units, the dimensions of the quotient Rm /Dm is still unity and that of B|| is:

dim[B|| ] = [M] . [T ]−2 . [I ]−1

[M] . [T ]−2
= 3 1
[L] 2 . [M] 2 . [T ]−2
3 1
= [L]− 2 . [M] 2 (10.57)

The unit (rad . m−2 ) of the rotation measure is not dependent on which of the
cgs Gaussian, IAU/SI or SI units is used.
The unit of the dispersion measure in cgs Gaussian units is the pc . cm−3 , the
appropriate unit in IAU/SI units would be pc . m−3 and that in pure SI units is
m . m−3 . The conversions from cgs Gaussian to IAU/SI and SI units and gauss to
teslas are:

1 pc . m−3 = 106 pc . cm−3 (10.58)

106
1 m . m−3 = = 3.240 × 1011 pc . cm−3 (10.59)
3.086 × 1016
1 T = 104 G (10.60)

Applying these conversion factors to the cgs Gaussian dispersion measures alters
the proportionality coefficient, PG , in the following ways:

PG = 1.232 × 10−6 (10.61)

 106 
PIAU = 1.232 × 10−6 = 1.232 × 10−4 (10.62)
104
 106 
PSI = 1.232 × 10−6 4 16
= 3.992 × 10−21 (10.63)
10 × 3.086 × 10
where PIAU is the proportionality coefficient for the IAU/SI unit and PSI is that for
the pure SI unit.

Example: evaluate the line-of-sight component of the galactic magnetic field

in the direction of the pulsar J2022+2854 located near the Cygnus/Vulpecula
border
Weisberg et al. (2004) carried out a comprehensive series of observations of Faraday
rotation measures of pulsars using the Arecibo radio telescope at a frequency of
 10.3 Magnetic fields in astronomy 193

430 MHz. Included in their programme was the pulsar J2022+2854, at a distance of
around 3 kpc (Verbiest et al., 2010). The observed values for the rotational measure
and dispersion measure were: Rm = −73.7 rad . m−2 and Dm = 24.623 pc . cm−3 .
Given these measurements, the mean line-of-sight magnetic flux densities, < B|| >,
were calculated in each of the systems of units.
(a) cgs Gaussian units:
R 
m
< B|| >G = PG (10.64)
Dm
where
Rm = −73.7 rad . m−2 (10.65)

Dm = 24.623 pc . cm−3 (10.66)

so
 −73.7 
−6
< B|| >G = 1.232 × 10 = −3.688 × 10−6 G = −3.688 μ G (10.67)
24.623
(b) IAU/SI units:

Rm = −73.7 rad . m−2 (10.68)

Dm = 2.4623 × 107 pc . m−3 (10.69)

so
 −73.7 
< B|| >IAU = 1.232 × 10−4
2.4623 × 107
= −3.688 × 10−10 T = −368.8 p T = −3.688 μG (10.70)

(c) SI units:

Rm = −73.7 rad . m−2 (10.71)

2.4623 × 107
Dm = = 7.979 × 10−10 m . m−3 (10.72)
3.086 × 1016
so
 −73.7 
< B|| >SI = 3.992 × 10−21
7.979 × 10−10
= −3.687 × 10−10 T = −368.7 p T = −3.687 μG (10.73)

So, to calculate the galactic field magnetic flux density in teslas from Faraday
rotation measures, input Rm in units of rad . m−2 , Dm in units of m . m−3 and use
PSI as the value for the coefficient of proportionality.
194 Unit of electric current (ampere)

10.4 Electric fields in astronomy

The presence of electric fields, potentials and currents has been detected in inter-
planetary space, on and about the Sun, in planetary atmospheres (e.g., the Earth,
Jupiter and Saturn) and in space near planetary bodies.

10.4.1 The Stark effect on stellar spectra

The Stark effect due to an electric field is analogous to the Zeeman effect due to a
magnetic field. The electric field causes the splitting of spectral lines in a manner
that depends on the strength of that field. To be observable in stellar spectral features
would require a uniform electric field over the entire stellar atmosphere – a phe-
nomenon that is neither expected nor observed. However, theoretical calculations
of Stark broadening parameters of likely interest to astrophysicists have recently
been made for use with spectra obtained with the Far Ultraviolet Spectroscopic
Explorer satellite (FUSE) by Alonso-Medina et al. (2009).

10.4.2 Measured electric fields, potentials and currents in, around,

near and on astronomical bodies
Some examples of measurements made, and of units used, in astronomy of electric
fields, potential, power and currents follow.

Solar electric fields and currents

Spectra and magnetograms of the NOAA active region 6233 were obtained at the
Mees Solar Observatory by de la Beaujardière et al. (1993) on 28 and 29 August
1990. On the later date they found values of the maximum Jmax , and minimum
Jmin , vertical current density to be + 26 mA . m−2 and −20 mA . m−2 and the total
positive current I+ , and total negative current I− , over the entire active region to
be +8.8 × 1012 A and −8.9 × 1012 A, respectively. Whilst the authors used SI units
for these electrical measurements, they plotted their image-plane magnetograms of
the longitudinal magnetic field in cgs Gaussian units (G).

Possible near-tail electric current by Mercury

From the Mercury Messenger spacecraft flybys, Anderson et al. (2010) have
inferred the presence of a near-tail electric current density of 100 nA . m−2 to
account for low field intensities recorded by the spacecraft near the equator of
Mercury. Again SI units are used for electric current density.
 10.5 Summary and recommendations 195

Jupiter’s lightning
Lightning on Jupiter was detected by the Cassini spacecraft whilst enroute to Saturn
in early 2001 and the observations analyzed by Dyudina et al. (2004). The most
powerful lightning storm recorded was measured to emit 800 MW of power in the
Hα line, corresponding to some 40 GW of broadband optical power.

Saturn’s aurorae
A study of the electric currents in the polar ionosphere of Saturn was carried out by
Cowley et al. (2004). They reported potentials in kV units, electron energy fluxes in
mW . m−2 , upward field-aligned current intensities in mA . m−1 and particle number
densities in cm−3 (cgs Gaussian units); in SI, the unit would be particles . m−3 .

10.5 Summary and recommendations

10.5.1 Summary
It is arguable that the relationships between the SI electrical and magnetic units
and those based on the older cgs system are more complex than those of the
other fundamental units. There are five major systems of units currently in use,
designed variously by physicists, chemists, electrical engineers and metrologists.
Since astronomers have, in the past, tended to favour the cgs Gaussian system of
units, this chapter has concentrated mainly on the relationship between that system
and the SI.
Examples are given of the names, symbols and dimensions of a range of derived
electric and magnetic units in SI and unrationalized cgs Gaussian form. Of impor-
tance is the difference between the defining mathematical equations for the two
unit systems, examples of which are set out in tabular form.
Worked examples of how to convert unrationalized cgs Gaussian units to SI units
are given and a table of coefficients to enable ready transformations between the
systems. Examples of celestial magnetic fields and units used to express them and
methods of measuring them (e.g., Zeeman splitting, Faraday rotation) are described.
The chapter finishes with a brief outline of electric fields in astronomy and
examples of the measurements made and units used.

10.5.2 Recommendations
Units of electricity and magnetism are manifold, and great care needs to be taken
in ascertaining to which of the five major systems the unit belongs (esu, emu,
cgs Gaussian, practical or SI). The conversion coefficients are not just powers of
ten, but often involve the physical and mathematical constants c, μ0 , 0 and π .
196 Unit of electric current (ampere)

Given that the IAU strongly recommends using SI units in all cases and that the
most common electric or magnetic unit used by astronomers is that for magnetic flux
density, such an outcome should be easy to accomplish. Perusal of relevant astro-
nomical literature to date indicates that SI units are becoming the unit of choice in
this area, so perhaps all that is needed is to encourage those still using cgs Gaussian
for research or teaching to change to SI. Until such a desirable situation is achieved,
a very strong recommendation is made to all astronomers when publishing results
to specify precisely which system or systems of units they are using.
 11
Unit of amount of substance (mole)

11.1 SI definition of the mole

1. The mole is the amount of substance of a system that contains as many elementary
entities as there are atoms in 0.012 kilogram of carbon 12 (12 C); its symbol is mol.
2. When the mole is used, the elementary entities must be specified and may
be atoms, molecules, ions, electrons, other particles, or specified groups of such
particles.
The dimension of amount of substance is [N ].

11.1.1 Possible future definition of the mole

Presently under discussion is redefining the unit of amount of substance in the
following way:

The mole is an amount of substance such that the Avogadro constant is exactly
6.022 141 5 × 1023 mol−1 (per mole).

11.2 Avogadro’s constant and atomic masses

In keeping with the proposed new definition of the mole, Avogadro’s constant may
be defined as the number NA of elementary entities per mole of substance which has
the (current) value 6.022 141 79 × 1023 mol−1 (Mohr et al., 2007). So the number
of atoms in 0.012 kg of 12 C is 6.022 141 79 × 1023 .
Note that the dimension of Avogadro’s constant is [N]−1 and its symbol mol−1 .

11.2.1 Atomic and molar masses

In SI units, the atomic mass unit (amu) is defined to be exactly 1/12 the mass of
one atom of the 12 C isotope.

197
198 Unit of amount of substance (mole)

In SI units, the molar mass of a substance is defined to be the mass in kg of

1 mol of the elementary entities (e.g., atoms or molecules) composing the substance.
Hence
0.012 kg 12 C = 1 mol (12 C atoms) (11.1)
It follows that the mass m(12 C) of one atom of 12 C in kg is:
0.012
m(12 C) =
NA
= 1.992 646 54 × 10−26 kg (11.2)

and that the mass of one atomic mass unit is given by:
1.992 646 54 × 10−26
1 amu =
12
= 1.660 538 78 × 10−27 kg (11.3)

In cgs units (still in common use in modern university textbooks, e.g., Chang
2005) the mass of one atom of 12 C is 1.992 646 54 × 10−23 g and that of 1 amu is
1.660 538 78 × 10−24 g. The conversion factor between cgs units and SI units is
0.001 (1 g = 0.001 kg; 1 kg = 1000 g).

11.2.2 Average atomic mass

The published atomic mass of an atom in general refers to a mean value determined
for a sample of that element composed of all its naturally occurring isotopes. For
example, hydrogen has three isotopes, 1 H or hydrogen with a nucleus consisting
of a single proton, 2 H or deuterium with a nucleus containing a single proton and
a single neutron, and 3 H or tritium with a nucleus containing one proton and two
neutrons.
Hence, the average atomic mass of hydrogen is the sum of the atomic masses
of hydrogen, deuterium and tritium, each multiplied by the mole fraction of that
isotope which occurs in the sample of hydrogen gas whose average atomic mass is
to be determined.
The mole fraction is a dimensionless quantity Xi which, in the case of a sample
of isotopes of a given element, expresses the ratio of the number of moles of one
isotope ni to the number of moles of all the isotopes present, nT (see, e.g., Chang
2005).
ni
Xi = (11.4)
nT
An example of an astronomical use of mole fractions is given by Flasar et al.
(2005), who used early Cassini space probe infrared observations of the atmosphere
 11.2 Avogadro’s constant and atomic masses 199

of Saturn’s largest moon, Titan, to determine the stratospheric mole fractions of the
molecules CH4 (methane) and CO (carbon monoxide) to be 1.6 ± 0.5 × 10−2 and
4.5 ± 1.5 × 10−5 , respectively.

Example: determine the average atomic mass of carbon

In the case of carbon there are three naturally occurring isotopes: 12 C, 13 C and 14 C.
Carbon 14 is a radioactive isotope with a half life of 5700 y, a value that has proved
to be extremely useful in dating human artifacts. 12 C has a mass of 12 amu and
is the major constituent of a typical terrestrial sample of carbon with a percentage
mole fraction of 98.93 %, 13 C has a mass of 13.003 35 amu and a percentage mole
fraction of 1.07 % and 14 C has a mass of 14.003 24 amu and a percentage mole
fraction of <10−10 %. The average atomic mass, m(C) of a terrestrial sample of
carbon is thus:
98.93 1.07 10−10
m(C) = × 12 + × 13.003 35 + × 14.003 24
100 100 100
= 12.010 74 amu (11.5)

A comprehensive listing of the average atomic masses for each of the elements is
given by Wieser & Berglund (2009) and a periodic table based on the data by G. P.
Moss is available on the web.62
The average atomic mass m of an element with n isotopes, where the mole
fraction of the ith isotope of mass mi is xi , is given by:

n
m= xi mi (11.6)
i=1

11.2.3 Molecular mass

The molecular mass of a molecule is defined to be the sum of the individual average
atomic masses of the atoms which make up the molecule.

Example: calculate the molecular mass of methane (CH4 ) and carbon

monoxide (CO) in atomic mass units
CH4

molecular mass of CH4 = (atomic mass of C) + 4 (atomic mass of H)

= 12.011 + 4 (1.0079)
= 16.042 6 amu (11.7)

62 See http://www.chem.qmul.ac.uk/iupac/AtWt/table.html
200 Unit of amount of substance (mole)

CO
molecular mass of CO = (atomic mass of C) + (atomic mass of O)
= 12.011 + 15.999
= 28.010 amu (11.8)
Note that the mass of 1 mole of CH4 is 0.016 043 kg and that of CO is
0.028 010 kg, since the molar mass in kg of a molecule is numerically equal to
0.001 times its molecular mass (there is a direct equivalence if cgs units are used,
i.e., 1 mole of CH4 has a mass of 16.043 g and 1 mole of CO has a mass of 28.010 g).
For further examples of the use of the mole, see Mills et al. (1993).

11.2.4 Mass spectrometer

An important instrument in determining the chemical composition of astronomical
objects such as meteorites, lunar and martian rock samples and the solar wind is
the mass spectrometer. In the first two cases, the sample to be examined may be
analyzed on Earth whilst, generally, the second pair require a miniaturized mass
spectrometer flown onboard a spacecraft.
The mass spectrometer is designed to determine the ionic, atomic and molecular
masses present in a gaseous sample of the object of interest. The gaseous sample is
bombarded by high-energy electrons to produce positive ions that may readily be
accelerated between pairs of oppositely charged plates, following which, the ions
are injected into a circular path by a magnet. The radius of the circular path is a
function of the ratio of e/m, with a larger value of this ratio producing a tighter
radius curve. So the mass of the ion, and hence that of the atom or molecule from
which it is produced, is also a function of the curve radius. Thus, it is possible to
determine the ratio of the quantities of different isotopes present in a sample and
the atomic constituents of a molecule.
In the case of the direct measurement of the solar wind, the detected particles
are already ionized. Weygand et al. (1999) determined the argon and neon isotopic
ratios of the solar wind from measurements made with the CELIAS isochronous
time-of-flight mass spectrometer (Hovestadt et al., 1995) flown onboard the SOHO
spacecraft. The isotopic ratios they measured for the solar wind were:
36 Ar 20 Ne
38 Ar
= 5.8 ± 1.1 and 22 Ne
= 14.7 ± 3.0

The SOHO63 spacecraft also records the speed of the solar wind and its particle
density at a given time. For example, at 23 h 44 m on 12 October 2010 the solar

63 See http://sohowww.nascom.nasa.gov/
 11.2 Avogadro’s constant and atomic masses 201

wind speed was 363 km . s−1 and its particle density in cgs units, 3.89 cm−3 , which
is equivalent to 3.89 × 106 m−3 in SI units since 1 cm−3 ≡ 106 m−3 .
An example of a ground-based determination of isotopic ratios is the examination
of five silicon carbide grains by Amari et al. (1992) obtained from the Merchison
carbonaceous meteorite. The sample produced values for the 26Al / 27Al isotopic
ratio that varied from 0.20 ± 0.01 to 0.61 ± 0.04.

11.2.5 Some further constants involving the mole

Faraday constant
If Q is the quantity of charge measured in coulombs flowing through an electrolytic
solution and n the number of moles of monovalent ions released at an electrode in
that solution, then the Faraday constant F is defined by:

Q
F= (11.9)
n

The dimension of the Faraday constant is [I ] . [T ] . [N ]−1 and its unit C . mol−1 .
Since the charge carried by a single monovalent ion is equivalent to the electric
charge e of an electron, then Equation (11.9) may be rewritten as:

F = NA e (11.10)

where NA is Avogadro’s constant.

Using the values for e and NA given by Mohr et al. (2007), the Faraday constant
is equal to:

F = 6.022 141 79 × 1023 × 1.602 176 487 × 10−19

= 9.648 533 98 × 104 C . mol−1 (11.11)

Universal or molar gas constant

(1) Boyle’s law: at constant n (number of moles of molecules present) and T
(temperature of the gas in kelvin), the volume V of the gas in m3 is inversely
proportional to the pressure of the gas in pascals (V ∝ P −1 ).
(2) Charles’ law: at constant n and P , the volume of the gas is directly proportional
to the temperature in kelvin (V ∝ T ).
(3) Avogadro’s law: at constant P and T , the volume of the gas is directly
proportional to the number of moles of molecules present (V ∝ n).
202 Unit of amount of substance (mole)

These three gas laws may be combined to create a universal, ideal or perfect gas
law in which:
VP
R= (11.12)
nT
where R is the universal or molar gas constant that has dimension
[L]2 . [M ] . [T ] . [N ]−1 . [
]−1 and unit J . mol−1 . K−1 , and is equal to (Mohr et al.,
2007):
R = 8.314 472 J . mol−1 . K −1 (11.13)
In cgs units, R = 8.314 472 × 107 erg . mol−1 . K−1 (see, e.g., page 94 et seq.
in Aller (1963) for a complete cgs treatment of the ideal gas laws as relevant to
astrophysics).

Molar volume of an ideal gas

The molar volume Vm of an ideal gas is given by:
RT
Vm = (11.14)
P
The dimension of the molar volume of an ideal gas is [L]3 . [N ]−1 and its unit is
m3 . mol−1 .
If T = 273.15 K and P = 100 kPa, then Vm = 22.710 981 × 10−3 m3 . mol−1 . For
the standard atmosphere, T = 273.15 K and P = 101.325 kPa, and the molar volume
is Vm = 22.413 996 × 103 m3 . mol−1 . These values of Vm are listed in Mohr et al.
(2007).

Molar masses of selected and subatomic particles

(1) Electron molar mass = 5.485 799 0943 × 10−7 kg . mol−1
(2) Proton molar mass = 1.007 276 466 77 × 10−3 kg . mol−1
(3) Neutron molar mass = 1.008 664 915 97 × 10−3 kg . mol−1
The dimension of each of the above subatomic molar masses is [M ] . [N ]−1 and
its unit is kg . mol−1 . These molar masses are given in CODATA recommended
values of the fundamental physical constants (Mohr et al., 2007).

11.3 Astrochemistry and cosmochemistry

Both astrochemistry and cosmochemistry are specialized branches of astronomy.
According to Cowley (1995), cosmochemistry deals with the chemical processes
responsible for the observed abundances of ions, atoms and molecules in the uni-
verse and the nuclear processes that cause such ions, atoms and molecules to
form.
 11.3 Astrochemistry and cosmochemistry 203

The short definition of astrochemistry in The Oxford Encyclopedic English Dic-

tionary (Hawkins & Allen, 1991) and the rather longer one in the Oxford Dictionary
of Astronomy (Ridpath, 2007) both define astrochemistry as a subset of cosmo-
chemistry that deals primarily with interstellar chemistry (i.e., low temperature and
pressure ambient physical conditions).

11.3.1 Interstellar chemistry

More than 100 interstellar and circumstellar molecules have been discovered
(Herbst, 2001) using radio and infrared astronomy over the past 50 years, from
the simplest H2 hydrogen molecule to those comprising 10 or more atoms (e.g.,
H C11 N ; C H3 C O C H3 ). The majority of these molecules are formed in interstel-
lar clouds of low density and at low temperatures. In astrophysics, the chemical
reactions which are of most importance are of the bimolecular type (Lang,
2006), e.g.,
A+ B C → AB + C (11.15)
where A and C are atoms and AB and BC are molecules. The rates of change with
time of the number densities, NA etc., are related in the following way:
dNA d dNAB dNC
− =− dt = = = k nA NBC (11.16)
dt N BC dt dt
Viala (1986) published a steady-state model of the chemical composition of
interstellar clouds in which the rates of gas-phase reactions k are written in the
general form:
β
k = (a T + b) T −α e− T cm3 . s−1 (11.17)
where the coefficients a, b, α and β for 1074 reactions are given. The units of a
and b are cm3 . s−1 , β is in kelvins and α is dimensionless (a ratio).

Example: derive the equation of gas-phase reactions for the chemical

reaction C + H2 O → C O + H2
From Table A in Viala (1986), the values of the appropriate coefficients for the
chemical reaction are:

a = 0 cm3 . s−1 b = 2.10 × 10−14 cm3 . s−1

α = −0.50 β =0 (11.18)

Substitute these values into Equation (11.17):

k = 2.10 × 10−14 T 0.5 cm3 . s−1 (11.19)

204 Unit of amount of substance (mole)

In SI, the coefficients would be in units of m3 . s−1 (where 1 m3 . s−1 ≡

106 cm3 . s−1 )
so that for the C + H2 O → C O + H2 reaction:

a = 0 m3 . s−1 b = 2.10 × 10−8 m3 . s−1

α = −0.50 β =0 (11.20)
k = 2.10 × 10−8 T 0.5 m3 . s−1 (11.21)

Table A in Viala (1986) also contains an entry, dE, for variation in enthalpy
(defined by Brimblecombe et al. 1998 as the heat energy associated with a chem-
ical change) during the chemical reaction, with a negative value indicative of an
exothermic reaction and positive for an endothermic reaction. The unit used is the
kcal . mol−1 .

Example: convert the value of the variation in enthalpy dE from cgs units to
SI units for the chemical reaction C + H2 O → C O + H2
From Viala (1986), the value of dE for this reaction given in Table A is
−140.076 kcal . mol−1 and there is an exact relationship between calories and
joules, such that 1 cal ≡ 4.184 J and 1 kcal ≡ 4184 J, hence:

−140.076 kcal . mol−1 = −140.076 × 4184 J . mol−1

= −5.860 78 × 105 J . mol−1 (11.22)

11.4 Summary and recommendations

11.4.1 Summary
The mole, the unit of amount of substance, is almost certainly the least used of all the
SI base units by astronomers as it sits more comfortably in the realm of the chemical
rather than the physical sciences. The fundamental relationships between the mole
and Avogadro’s constant is described. This will prove to be of great importance
should the SI definition of the mole be directly linked to Avogadro’s constant.
Atomic, molar and molecular masses are defined, as is the atomic mass unit
and the mole fraction. Examples are given of deriving the average atomic mass
of carbon isotopes and the molecular masses of methane and carbon monoxide. A
brief description of the mass spectrometer is followed by examples of observations
made with such an instrument in the laboratory and from space.
Compound units involving the mole are described and examples given, including
transforming cgs to SI units. In the final section of the chapter, astrochemistry,
cosmochemistry and interstellar chemistry are defined and a worked example given
of deriving the equation of gas-phase reactions for the case of C + H2 O.
 11.4 Summary and recommendations 205

11.4.2 Recommendations
Wilkins (1989) makes no specific reference to the mole or its usage, though the SI
version is implied from the overriding recommendation from the IAU that all SI
units are acceptable.
The limited appearances of the mole in modern astronomical literature means
that it has to be assumed that a definition of the unit in a particular source is based on
12 C rather than the earlier usage of 16 O or 1 H. When compound units are used, there

is a tendency for some parts of the unit to come from the cgs range (e.g., cm3 . mol−1
rather than m3 . mol−1 ) and, as is not uncommon in astronomical papers, textbooks
or reference works, sometimes a mixture of both cgs and SI units. This practice
should be avoided as confusion is the likely outcome.
 12
Astronomical taxonomy

12.1 Definition of taxonomy

Taxonomy is defined as the science of classification and is derived from the ancient
Greek word, τ αξ ις , meaning arrangement, order, regularity (Liddell & Scott,
1996).

12.2 Classification in astronomy

Funk & Wagnalls New Standard Dictionary of the English Language (1946) defines
classification as:

The act or process of arranging by classes; a grouping into classes; the putting together of
like objects or facts under a common designation; a process based on similarities of nature,
attributes, or relations. Classification may proceed by the gathering together of similar things
into a class, or by the unfolding of general groups into narrower or more specific divisions.

To classify, therefore, is to arrange in a class or classes on the basis of observed

resemblances and differences. For this to proceed, two pieces of information are
needed: an identity (the name of that which is to be classified) and an attribute
(does the identified object have or could it have the necessary information for it to
be classified as having the attribute), e.g., does Sirius (the identity – the name of
the star) have a spectral type (the attribute or classification)? The answer is yes and
the spectral type of Sirius is A1. It should be noted that the group of identifying
names or definitions of a class of objects may also constitute a classification, e.g.,
the recent IAU definitions of types of bodies in the Solar System in which the new
class, dwarf planet, contains the objects Pluto, Ceres and Eris.
For most of its history, astronomy was a purely observational science, relying
entirely on receiving from the Universe electromagnetic radiation (light, infrared,
ultraviolet, X-rays, γ -rays, radio waves), particular material (solar wind, cosmic
rays) and larger assemblies of matter (dust, meteorites). This led to astronomy
developing initially as a science of classification. Most forms of astronomical

206
 12.3 Classification of stellar objects 207

Table 12.1. A simple classification of the

baryonic matter in the Universe

Group Subgroup
The Universe Clusters of galaxies
Clusters of galaxies Galaxies
Intergalactic gas
Intergalactic dust
Galaxies Stars
Star clusters
Interstellar gas
Interstellar dust
Stars Double stars
Exoplanets
The Solar System
The Solar System The Sun
Solar wind
The planets
Dwarf planets
Satellites
Small Solar System bodies

taxonomy at the present time are quantitative, i.e., they depend on the measurement
or measurements of physical attributes (position, size, distance, motion, emergent
flux etc.) made using objective instrumentation. Earlier forms of taxonomy relied
on a qualitative assessment of the observational attribute, e.g., in estimating the
brightness of a star by visual comparison with others nearby of known brightness,
or the assignment of a spectral class from the examination of a spectrogram, or deter-
mining the morphological type of a galaxy from a photographic or digital image.
Whilst such subjective methods are still in use, they are far less common than
they once were. Examples of objective measurements and classifications, where
relevant, are given in the chapters on individual SI units. This chapter contains
some examples of subjective classifications that originally depended solely on a
visual estimation of some parameter or ratio, e.g., the comparison of spectral line
strengths, the apparent shape of a galaxy or the magnitude of a star. Such classi-
fications are generally dimensionless. Table 12.1 sets out one possible scheme for
subdividing the Universe into classifiable groupings and subgroupings.

12.3 Classification of stellar objects

Stellar objects include stars, double stars, star clusters and planetary or gaseous
nebulae.
208 Astronomical taxonomy

A star itself may be defined as a spherical or ellipsoidal gaseous object that is

both massive enough, and hot enough near its centre, to sustain energy-generating
nuclear reactions. If the mass of the body is less than about 80 times the mass of
the planet Jupiter, then the object will have insufficient mass to be able to sustain
internal nuclear reactions and it would not be classified as a star.
Stars may be classified in many different ways, such as by location, brightness,
colour, chemical composition, mass, age etc.

12.3.1 Stellar identity

The earliest form of identification of individual stars was to assign them names and
to name prominent groupings of stars.

Individually named objects

All the bright stars and a good many fainter, naked-eye ones have individual names.
For example, Sirius, the apparently brightest star in the night sky. This is the Greek
name for the star, meaning sparkling or scorching (Allen, 1899). Sirius has other
names in other languages. Some stars are named for their position on the sky,
such as Polaris, the north polar star, or for their position within a constellation (a
grouping of stars forming a generally recognizable pattern), an example of which
is Betelgeuse, a bright orange–red star in the constellation of Orion, whose name
loosely translates from the Arabic as ‘The armpit of the giant’. Many of the proper
names of stars are associated with the mythology of various races and often with
their agricultural calendars or for navigational purposes.

Constellations
The entire sky is subdivided into 88 groups of stars of varying sizes, called constel-
lations, well-known examples of which are Orion, Ursa Major (The Great Bear),
Gemini (The Twins), Crux Australis (The Southern Cross) and Scorpio (The Scor-
pion). Some date back thousands of years and others, mainly constellations near
the south celestial pole, are more recent inventions that followed the exploration of
the southern hemisphere by European navigators. In the early seventeenth century,
just prior to the use of the telescope for astronomical purposes, Bayer produced
a star atlas (the Uranometria) in which the stars forming the constellations were
identified with a Greek letter, assigned approximately in order of decreasing star
brightness from α to ω. Thus, Sirius is also known as α Canis Majoris and Mirzam,
the second brightest star in the constellation, as β Canis Majoris. If all the Greek
letters are used, then the letters of the Roman alphabet or numbers assigned by
Flamsteed (the first Astronomer Royal, appointed in 1675) are used.
 12.3 Classification of stellar objects 209

Table 12.2. A selection of identifiers from different catalogues for

the star Sirius

Catalogue Identifier
Name Sirius, Dog Star
Uranometria α CMa, 9 CMa
Bonner Durchmusterung BD-16 1591
5th Fundamental Catalogue FK5 257
Henry Draper Catalogue HD 48915
HIPPARCOS Catalogue HIP 32349
Bright Star Catalogue HR 2491
Infrared Astronomical Satellite IRAS 06429-1639
ROSAT (X-ray) RX J0645.1-1642
Smithsonian Astrophysical Observatory SAO 151881
TD1 (Ultraviolet Photometry) TD1 8027

Faint star identities

The advent of the telescope permitted the observation of large numbers of much
fainter stars. The identity assigned to such stars is generally the name of the
astronomer who constructed the star catalogue, followed by a running number. The
introduction of photography resulted in another large increase in the number of
stars observed, eventually resulting in the need for automated measuring machines
to construct the catalogues. Stars in these publications are commonly identified by
the name of the observatory or institute that carried out the photographic measure-
ments or the name of the benefactor who provided the money for the observational
programme, followed by a number that is related to the position of the star on the
sky at a particular date.
By way of example, the astronomical database SIMBAD64 lists 54 different
identifiers for the star Sirius, from which the ten shown in Table 12.2 were selected.

12.3.2 Classification by brightness

The first attempt at assigning a value for the brightness of a star was due to Hip-
parchus, a second-century (BCE) Greek astronomer, who produced a catalogue of
1080 stars in which each star was labelled as being of the first (very bright), sec-
ond, third, fourth, fifth or sixth (very faint star, only just visible to the naked eye)
magnitude.
This system was extended to the 7th, 8th, 9th etc. magnitudes after the telescope
revealed objects that could not be seen with the naked eye. This rather loose ordinal

64 See http://simbad.u-strasbg.fr/simbad/
210 Astronomical taxonomy

method of classifying the brightness of a star was given a little more mathematical
rigour by Pogson in 1850, when he set a factor of exactly 100 as the difference in
brightness between a first and a sixth magnitude
√ star. So a one magnitude difference
5
in brightness corresponds to a factor of 100 or approximately 2.512. It should be
noted that the human eye has a logarithmic response to stimulation by light.
The last major catalogue that relied on the human eye to estimate the brightness
of stars was the Bonner Durchmusterung or BD. This contained all the stars from
the north celestial pole to a declination of −2◦ observed by Argelander with a
72 mm Fraunhofer telescope (King, 1955). The catalogue lists some 320 000 stars
with positions to 0s .1 in right ascension and 0 .1 in declination. Argelander visually
estimated magnitudes to ±0.1 to a limiting magnitude of 9.5 (van Biesbroeck,
1963). Subsequent photographically determined measures of Argelander’s stars in
the BD catalogue are remarkably similar to his eye estimates down to magnitude 9.0.
Visual estimates of stellar magnitudes are still made by a group of dedicated
amateur astronomers whose particular interest is stars that vary in brightness over a
long period of time (∼100 d) or that vary erratically or unpredictably. An example
of one of the techniques used is the fractional method (Sidgwick, 1955), where
two non-varying comparison stars near to the variable star are selected, one of
which is slightly brighter than the variable and the other slightly fainter (ideally the
magnitude difference between the comparison stars should not exceed 0.4). The
magnitude of the variable star obviously lies between the two comparison stars.
The brightness interval between the comparison stars is expressed as an integer
number of parts (normally less than 10) and an estimate made of the fraction of that
number that best represents the brightness of the variable. If the comparison stars
are of magnitudes m1 and m2 , where m2 > m1 , and the number of steps between
them is n, then if the variable is estimated to be k steps brighter than the fainter
comparison star, its magnitude mvar , will be:
k (m2 − m1 )
mvar = m2 − (12.1)
n

12.3.3 Spectral classification of stars

The earliest serious attempt at classifying stellar spectra was carried out by
Fr. Angelo Secchi. He began a spectroscopic study of the stars in 1862 (Hearn-
shaw, 1986), which eventually led to a classification scheme consisting of five
classes. This was superseded by that of the Harvard College Observatory in the
1890s. Under the direction of E. C. Pickering, a survey of stellar spectra recorded
photographically with an 11-inch telescope equipped with an objective prism were
analyzed initially by Antonia Maury and later by Annie Jump Cannon. Further
observations of stars in the southern hemisphere were obtained using the 13-inch
 12.3 Classification of stellar objects 211

Table 12.3. Spectral features associated with different spectral classes

Spectral type Effective temperature (K) Selected spectral features

O ≥42000−34000 HeII, HeI, SiIV
B 30000−11400 HeI, SiIV, SiIII, SiII, MgII
A 9800−8100 HI, FeII, MgII, SiII, CaII, MnI
F 7300−6200 CaII, HI, FeI, CrI, CaI, CN
G 5900−5300 CaII, FeI, HI, CH, NaI
K 5200−4400 CrI, TiI, CaI, FeI, CH, CN, TiO
M 3900−3100 TiO, VO
L 2000−1300 NaI, KI, RbI, CsI, LiI, MgH, CaH, CrH, FeH
MgH, CaH, CrH, FeH
T 1300−700 CH4 , H2 O, NaI, KI
(Y) 600−

Boyden telescope in South Africa to complete the coverage of the entire sky. Clas-
sification was at first limited to magnitude 6. Since these photographs, plus others
taken with a 10-inch telescope in Peru for the survey, had a much fainter limit-
ing magnitude, the catalogue was able to be extended. Altogether, Miss Cannon
classified nearly 400 000 stars by eye over a period of 45 years. The results were
published in the Henry Draper catalogue plus its extensions. This gargantuan work
formed the basis of the modern scheme of classifying stellar spectra.
The HD (Henry Draper) classifications are essentially one dimensional, with
spectral class assigned by the relative strengths of the spectral lines of ions and
atoms and the bands of molecules. The letters O, B, A, F, G, K, M, L, T, (Y) form
the spectral sequence, with each class being subdivided by a number from 0 to 9,
e.g., A1 (spectral type of Sirius), G2 (spectral type of the Sun) and M2 (spectral
type of Betelgeuse). In physical terms, the spectral sequence from O to (Y) is one of
decreasing temperature. The defining spectral features (Keenan, 1963; Kirkpatrick,
2005) and effective temperatures (Cox, 2000; Burningham et al., 2008; Leggett
et al., 2009) of each class are given in Table 12.3. Class (Y) is not yet well defined
due to a lack of suitable candidate objects.

Luminosity classification
To convert the MK classification from a one- to a two-dimensional scheme, pairs of
spectral lines were selected for standard stars of known spectral type and absolute
luminosity. By comparing the line ratios exhibited by stars of unknown luminosity
class with those of the standard stars, an estimate of the luminosity class for the
candidate star may be made. Luminosity is a function of the surface area and hence
the radius of the star. If the large and small radius stars of the same spectral type have
212 Astronomical taxonomy

Table 12.4. Line pairs used to classify luminosity for different

spectral class ranges (Cox, 2000)

Spectral type range Useful line pairs

O9 → B3 SiIV, HeI (411.6–412.2 nm) / HeI (414.4 nm)
B0 → B3 NII (399.5 nm) / HeII (400.9 nm)
B1 → A5 HI Balmer line wings
A3 → F0 MgII (441.6 nm) / MgII (448.1 nm)
F0 → F8 CaI (417.2 nm) / CaI (422.6 nm)
F2 → K5 FeI (404.5–406.3 nm) / SrII (407.7 nm)
CaI (422.6 nm) / SrII (407.7 nm)
G5 → M Discontinuity near 421.5 nm
K3 → M CaI strength increasing 421.5 nm → 426.0 nm

similar masses, then the larger star must have a lower gas density and pressure, and
surface gravity, than the smaller star. These differences show up in the appearance
of certain spectral lines; those of a very large star are narrow and sharply defined,
whilst those of a smaller radius star show the effects of pressure broadening and
appear wider and less clearly defined.
Table 12.4 lists the spectral line pairs that are of particular value in assigning
luminosity classes to stars of spectral type O to M. The fainter types, L, T and (Y),
are for the most part not truly stars and are currently termed L dwarfs, T dwarfs and
(Y) dwarfs.
There are eight main luminosity classes: class O (hypergiants), class I (super-
giants), class II (bright giants), class III (giants), class IV (subgiants), class V (main
sequence or dwarfs), class VI (subdwarfs), and class VII (white dwarfs). Some
examples of the two-dimensional spectral type / luminosity class system are Betel-
geuse (M2Iab),  CMa (B2II), Arcturus (K2III), α Cru (B0.5IV), the Sun (G2V)
and α Cen A (G2V).

12.3.4 Classification of star clusters

A star cluster may be defined as a group of stars that, forming a physical system,
appear at approximately the same distance from the observer, in approximately the
same direction, sharing similar transverse and radial motions.
Star clusters may be located by their appearance as an enhanced stellar density
region on the sky. There are two major types of star cluster:

1. The open or galactic clusters, which typically are located in or near to the
galactic plane of the Milky Way Galaxy, with member stars that generally are
not centrally concentrated and have less than 1 000 members.
 12.3 Classification of stellar objects 213

2. The globular clusters, which are strongly spherical in shape, have upwards of
1 000 members (exceeding 100 000 stars for very large examples) and are located
in a roughly spherical distribution about the centre of the Milky Way Galaxy.
Schemes for classifying clusters generally depend on the number of members,
location relative to the galactic plane, ages, colours and spectral types of the
members.

The Trumpler classification of galactic star clusters

Trumpler (1930) devised a classification scheme for galactic star clusters that uses
three parameters: degree of concentration of cluster stars, range in brightness of
cluster members and number of cluster members. The descriptors within each
parameter are:

1. Degree of concentration
I: Detached clusters with a strong central concentration
II: Detached clusters with little central concentration
III: Detached clusters with no noticeable concentration
IV: Clusters not well detached but with strong field concentrations.
2. Range of brightness
1: Most of the cluster stars are approximately the same apparent brightness
2: A medium brightness range between the stars in the cluster
3: The cluster is composed of a mixture of bright and faint stars
3. Number of stars in the cluster
p: Poorly populated clusters with less than 50 member stars
m: Medium rich clusters with between 50 and 100 member stars
r: Rich clusters with over 100 member stars.
In addition, the letter ‘n’ may be added at the end of the classification to signify
the presence of any form of nebulosity.

In the first instance, the degree of concentration and the range of brightness may
be assessed by eye, as may the counting of those stars qualitatively presumed to
be cluster members. A recent example of a Trumpler classification for the southern
hemisphere open cluster IC2391 of II3r is given by Dodd (2004).

12.3.5 Morphological classification of galaxies

Even a casual glance at the illustrations of galaxies in most modern general astro-
nomical textbooks (e.g., Kutner, 2003) reveals that galaxies do not all look the
same. A scheme for classifying galaxies according to their visual appearance was
devised by Hubble in the early 1920s (Hubble, 1926). A photographic selection of
214 Astronomical taxonomy

galaxies belonging to the different classes is set out in The Hubble Atlas of Galax-
ies by Sandage (1961) and online examples may be found on the Hubble Space
Telescope website.65

The Hubble classification of galaxies

The majority of known galaxies may be placed on a diagram, which is generally
referred to as the tuning fork diagram (see Figure 12.1). On the left-hand side of
the diagram are found elliptical galaxies, with their degree of ellipticity, E:
b
E = 10 . (1 − ) (12.2)
a
where a is the length of the semi-major and b that of the semi-minor axis, increasing
from E0 to E7, the most elliptical type so far found. In the centre of the tuning fork
are found spherically shaped galaxies with prominent bulges but no spiral arms
(types S0 and SB0, where the S is for spiral and the B for barred). The upper-right
portion of the diagram contains a sequence of spiral galaxies (Sa, Sb, Sc) with
decreasing central bulge size relative to the entire galaxy and increasing openness
of the spiral-arm structure. A parallel sequence in the lower right of the tuning fork
is occupied by the barred spirals (SBa, SBb, SBc) in which the spiral arms begin,
not from the galactic central bulge itself, but from the ends of a bar that runs through
the galactic centre.
Some galaxies do not fit this simple classification scheme and the classes Irreg-
ular, Lenticular, Ring, Seyfert and Peculiar were added later. Table 12.5 sets out
some of the more obvious morphological and astrophysical properties of the main
types of galaxies.

Sa Sb Sc

bulge decreasing >>>

E0 E4 E7 S0

increasing
ellipticity >>> SB0
increasing
openness >>>

SBa SBb SBc

Figure 12.1. Hubble classification of galaxies.

65 http://hubblesite.org/newscenter/archive/images/galaxy
 12.4 Classification of Solar System objects 215

Table 12.5. Morphological and astrophysical properties of elliptical and spiral

galaxies

Hubble Nuclear Central General Young Gas and Integrated

type bulge bar shape stars dust colour
E no no ellipsoidal no small red
amounts
S yes no flattened yes yes blue
SB yes yes flattened yes yes blue

12.4 Classification of Solar System objects

In 2006, at the 26th General Assembly of the International Astronomical Union
(IAU), resolutions were discussed on the definitions of the terms planet, dwarf
planets and small Solar System bodies. The IAU press release was as follows:
Contemporary observations are changing our understanding of planetary systems, and it is
important that our nomenclature for objects reflect our current understanding. This applies
in particular, to the designation planets. The word planet originally described wanderers
that were known only as moving lights in the sky. Recent discoveries lead us to create a
new definition, which we can make using currently available scientific information.

12.4.1 Definition of a planet in the Solar System

A planet is a celestial body that (a) is in orbit around the Sun, (b) has sufficient mass
for its self-gravity to overcome rigid-body forces so that it assumes a hydrostatic
equilibrium (nearly round) shape, (c) has cleared the neighbourhood around its
orbit, and (d) is not a satellite.
The eight objects defined now as planets are, in increasing distance from the
Sun: Mercury, Venus, Earth, Mars, Jupiter, Saturn, Uranus, and Neptune.

12.4.2 Definition of a dwarf planet in the Solar System

A dwarf planet is a celestial body that (a) is in orbit around the Sun, (b) has
sufficient mass for its self-gravity to overcome rigid-body forces so that it assumes a
hydrostatic equilibrium (nearly round) shape, (c) has not cleared the neighbourhood
around its orbit, and (d) is not a satellite.
The dwarf planet Pluto is recognized as an important prototype of a new class
of trans-Neptunian objects.
Examples of dwarf planets are: Pluto, Ceres and Eris.
216 Astronomical taxonomy

12.4.3 Definition of small Solar System bodies

All other bodies, except satellites, orbiting the Sun shall be referred to collectively
as small Solar System bodies.
These currently include most of the Solar System asteroids, most of the trans-
Neptunian objects (TNOs), comets, and other small bodies.
Examples of such small Solar System bodies are: Vesta, Pallas, Eros (all aster-
oids); Comet Halley, Comet McNaught (both comets); and meteors (whilst still in
orbit about the Sun).

12.4.4 Definition of a satellite

The IAU did not produce a formal definition for a satellite. Funk et al. (1946) define
such a body as:
A smaller body attending and revolving round a larger one; commonly a secondary planet
round a primary one.

The Earth’s Moon is a satellite, as are: Phobos and Deimos of the planet Mars;
Io, Ganeymede, Europa and Callisto of the planet Jupiter; Iapetus, Mimas and Titan
of the planet Saturn; Umbriel, Titania and Oberon of the planet Uranus; Triton and
Nereid of the planet Neptune; and Charon of the dwarf planet Pluto.

12.5 Astronomical databases and virtual observatories

Until recently, information about celestial bodies was generally published as a
printed volume or volumes (e.g., The Astronomical Almanac (annual), the Bright
Star Catalogue, the HD catalogue of stellar spectra). The advent of computer sys-
tems has led to the production of machine readable catalogues, which may be held
online or made available as CD or DVD discs.

12.5.1 Astronomical databases

Both general-purpose databases and those with a particular astronomical research
area in mind have been set up online. The following is a small selection of some of
the data that are available, with their web address.
SIMBAD astronomical database is a compendium of information about stars,
including their positions, distances, proper motions, radial velocities, pho-
tometry, spectroscopy and bibliography.
http://simbad.u-strasbg.fr/simbad/
 12.5 Astronomical databases and virtual observatories 217

Table 12.6. Members of the International Virtual Observatory Alliance

Virtual observatory web address

International Virtual Observatory Alliance http://www.ivoa.net/
Armenian Virtual Observatory http://www.aras.am/Arvo/arvo.htm
Virtual Observatory United Kingdom http://www.astrogrid.org/
Australian Virtual Observatory http://www.aus-vo.org/
Chinese Virtual Observatory http://www.china-vo.org/
Canadian Virtual Observatory http://services.cadc-ccda.hia-iha.
nrc-cnrc.gc.ca/cvo/
European Virtual Observatory http://www.euro-vo.org/
German Astrophysical Virtual Observatory http://www.g-vo.org/
Hungarian Virtual Observatory http://hvo.elte.hu/en/
Japanese Virtual Observatory http://jvo.nao.ac.jp/
Korean Virtual Observatory http://kvo.kasi.re.kr/
National Virtual Observatory United States http://www.us-vo.org
Observatoire Virtuel France http://www.france-vo.org/
Russian Virtual Observatory http://www.inasan.rssi.ru/eng/rvo/
Spanish Virtual Observatory http://svo.laeff.inta.es/
Italian Virtual Observatory http://vobs.astro.it/
Virtual Observatory India http://vo.iucaa.ernet.in/

The European Southern Observatory (ESO) archive includes images and

data, the digitized sky surveys, the HIPPARCOS and TYCHO catalogues,
records of images obtained with the various ESO telescopes and ESO
publications.
http://archive.eso.org/
The image and catalogue archive of the United States Naval Observatory (USNO)
includes the astrometric USNO catalogues.
http://www.nofs.navy.mil/data/fchpix/
The Washington Double Star Catalogue is also housed at the USNO.
http://ad.usno.navy.mil/wds/
For astronomers with a particular interest in star clusters, the WEBDA site of
the University of Geneva will prove very useful.
http://obswww.unige.ch/webda/

12.5.2 Virtual observatories

The virtual observatory (VO) combines astronomical databases at different cen-
tres with specially written software to create a scientific research environment
that facilitates work on a variety of astronomical, astrophysical and cosmological
research programmes.
218 Astronomical taxonomy

There are a number of virtual observatories throughout the world. Their web
addresses are given above in Table 12.6, along with that of the International
Virtual Observatory Alliance (IVOA), whose primary task is to develop agreed
interoperability standards with the member VOs.

12.6 Summary and recommendations

12.6.1 Summary
The classification of celestial bodies according to particular attributes is often a
useful first stage in grouping together objects that may later be shown to have sim-
ilar physical properties. In the examples given in this chapter, this expectation is
realized with the morphological classification of galaxies, but not with the group-
ing of stars according to their apparent observed brightness, since stars do not all
have the same intrinsic brightness and do not all lie at the same distance from the
observer.
The development of astronomical databases and virtual observatories allows
high-quality data to be examined by a far wider group of research astronomers
than was previously the case and should lead to a greater understanding of the
significance of the data.

12.6.2 Recommendations
When celestial bodies are classified by measurement (e.g., positions, distances,
motions, flux outputs, temperatures etc.) and published in an online database or
virtual observatory, then such measurements should be given in SI units. If other
units are used, then they should be accompanied by SI units.
 References

Abdo, A. A., Ackermann, M., Ajello, M. et al. (2009). Discovery of pulsations from the
pulsar J0205+6449 in SNR 3C 58 with the FERMI γ -ray space telescope ApJL, 699,
L102–L107.
Abdo, A. A., Ackermann, M., Ajello, M. et al. (2010). The first catalog of active galactic
nuclei detected by the FERMI Large Area Telescope. ApJ, 715, 429–457.
Abramowitz, M. & Stegun, I. A. eds (1972). Handbook of Mathematical Functions. New
York: Dover Publications Inc., p. 67.
Albacete-Colombo, J. F., Damiani, F., Micela, G., Sciortino, S. & Harnden, F. R., Jr. (2008).
An X-ray survey of low-mass stars in Trumpler 16 with CHANDRA. A&A, 490,
1055–1070.
Alexeev, I. I., Belenkaya, E. S., Slavin, J. A. et al. (2010). Mercury’s magnetospheric
magnetic field after the first two MESSENGER flybys. Icarus, 209, 23–39.
Alder, K. (2004). The Measure of All Things. London: Abacus.
Allen, C. W. (1951). Dex. Observatory, 71, 157.
Allen, R. H. (1899). Star Names and their Meanings. New York: G. E. Stechert & Co.
Aller, L. H. (1963). Astrophysics: The Atmospheres of the Sun and Stars, 2nd edn. New
York: The Roland Press Company.
Alonso-Medina, A., Colón, C., Montero, J. L. & Nation, L. (2009). Stark broadening of
PbIV spectral lines of astrophysical interest. MNRAS, 401, 1080–1090.
Amari, S., Hoppe, P., Zinner, E. & Lewis, R. S. (1992). Interstellar SiC with unusual isotopic
compositions: Grains from a supernova? ApJ, 394, L43–L46.
Anderson, B. J., Acuña, M. H., Korth, H. et al. (2010). The magnetic field of Mercury. Space
Sci. Rev., 152, 307–339.
Astronomical Almanac for 1995, The (1994). London: HMSO.
Atkin, K. (2007). Size matters. A&G, 48, 4.7.
Atwood, W. B. Abdo, A. A., Ackermann, M. et al. (2009). The Large Area Telescope on the
FERMI γ -ray space telescope mission. ApJ, 697, 1071–1102.
Bagnulo, S., Szeifert, T., Wade, G. A., Landstreet, J. D. & Mathys, G. (2002). Mea-
suring magnetic fields of early type stars with FORS1 at the VLT. A&A, 389,
191–201.
Barlow, C. W. C. & Bryan, G. H. (1956). Elementary Mathematical Astronomy. London:
University Tutorial Press Ltd., p. 125.
Battat, J. B. R., Murphy, T. W., Jr., Adelberger, E. G. et al. (2009). The Apache Point Lunar-
ranging Operation (APOLLO): Two years of millimeter-precision measurements of
the Earth–Moon range. PASP, 121, 29–40.

219
220 References

de la Beaujardière, J.-F., Canfield, R. C. & Leka, K. D. (1993). The morphology of flare

phenomena, magnetic fields and electric currents in active regions. III NOAA active
region 6233 (1990 August). ApJ, 411, 378–382.
Bedding, T. R., Butler, R. P., Carrier, F. et al. (2006). Solar-like oscillations in the metal-
poor subgiant ν Indi: constraining the mass and age using asteroseismology. ApJ, 647,
558–563.
Beech, M. (2008). The reluctant parsec and the overlooked light-year. Observatory, 128,
489–494.
Bemporad, A. & Mancusco, S. (2010). First complete determination of plasma physical
parameters across a coronal mass ejection driven shock. ApJ, 720, 130–143.
Bessel, F. W. (1838). On the parallax of 61 Cygni. MNRAS, 4, 152–161.
Bessell, M. S. (1992). In The Astronomy and Astrophysics Encyclopedia, ed. S. P. Maran.
Cambridge: Cambridge University Press, p. 403.
Bessell, M. S. (2001). In Encyclopedia of Astronomy & Astrophysics, Vol. 2, ed. P. Murdin,
Bristol: IoP, pp. 1638–1644.
BIPM (2006). The International System of Units, 8th edn. France: CIPM.
Blanco, V. M. & McCuskey, S. W. (1961). Basic Physics of the Solar System. Reading, MA:
Addison-Wesley Publishing Company, Inc.
Blauuw, A., Gum, C. S., Pawsey, J. L. & Westerhout, G. (1960). The new IAU system of
galactic coordinates (1958 revision). MNRAS, 121, 123–131.
Bleaney, B. I. & Bleaney, B. (1962). Electricity and Magnetism. Oxford: Clarendon Press.
Bohlin, R. C. & Gilliland, R. L. (2004). Hubble Space Telescope absolute spectrophotometry
from the far ultraviolet to the infrared. AJ, 127, 3508–3515.
Bok, B. J. (1937). The Distribution of the Stars in Space. Chicago, IL: The University of
Chicago Press.
Bradley, P. A. (1993). Methods of asteroseismology for white dwarf stars. Baltic Astr., 2,
545–558.
Brimblecombe, S., Gallannaugh, D. & Thompson, C., eds. (1998). The Hutchinson
Encyclopedia of Science. Oxford: Helicon
Brown, M. E. & Schaller, E. L. (2007). The mass of the dwarf planet Eris. Science, 316,
1585.
Browning, D. R., ed. (1969). Spectroscopy. London: McGraw-Hill.
Burke, B. F. & Graham-Smith, F. (2002). An Introduction to Radio Astronomy. 2nd edn.
Cambridge: Cambridge University Press.
Burningham, B., Pinfield, D. J., Leggett, S. K. et al. (2008). Exploring the substellar
temperature region down to ∼ 500 K. MNRAS, 391, 320–333.
Cairns, W. (2007). About the Size of It. London: MacMillan.
Capitaine, N. & Guinot, B. (2008). The astronomical units. arXiv:0812.2970v1 [astro-ph].
Cardarelli, F. (2003). Encyclopedia of Scientific Units, Weights and Measures. London:
Springer.
Chambers, G. F. (1889). Astronomy. I: The Sun, Planets and Comets. Oxford: Clarendon
Press.
Chang, R. (2005). Chemistry, 8th edn. New York: McGraw-Hill.
Cohen, M., Walker, R. G., Barlow, M. J. & Deacon, J. R. (1992). Spectral irradiance cali-
bration in the infrared. I. Ground-based and IRAS broadband calibrations. AJ, 104,
1650–1657.
Cohen, M., Wheaton, W. A. & Megeath, S. T. (2003). Spectral irradiance calibration in the
infrared XIV: The absolute calibration of 2MASS. AJ, 126, 1090–1096.
Collins, II, G. W. (1989). The Fundamentals of Stellar Astrophysics. New York: W. H.
Freeman and Company.
 References 221

Coulson, C. A. & Boyd, T. J. M. (1979). Electricity. London: Longman.

Cowley, C. R. (1995). An Introduction to Cosmochemistry. Cambridge: Cambridge Univer-
sity Press.
Cowley, S. W. H., Bunce, E. J. & Prangé, R. (2004). Saturn’s polar ionospheric flows and
their relation to the main auroral oval. Ann. Geophys., 22, 1379–1394.
Cox, A. N. (2000). Allen’s Astrophysical Quantities, 4th edn, ed. A. N. Cox. New York:
Springer.
Culhane, J. L. & Sanford, P. W. (1981). X-ray Astronomy. London: Faber & Faber.
Dodd, R. J. (2004). Data mining in the young open cluster IC2391. MNRAS, 355,
959–973.
Dodd, R. J. (2007). Unified absolute spectrophotometry for star clusters, ed. C. Sterken.
ASP Conference Series, 364, pp. 237–254.
Duffard, R., Ortiz, J. L., Santos Sanz, P. et al. (2008). A study of the photometric variations
on the dwarf planet (136199) Eris. A & A, 479, 877–881.
Dyson, F. W. (1913). Report of the Royal Astronomical Society meeting on 14 March 1913.
Observatory, 36, 160.
Dyudina, U. A., del Genio, A. D., Ingersoll, A. P. et al. (2004). Lightning on Jupiter observed
in the Hα line by the Cassini imaging science subsystem. Icarus, 172, 24–36.
Emerich, C., Lamarre, J. M., Gispert, R. et al. (1988). Temperature of the nucleus of comet
Halley. ESA Proceedings of International Symposium on the Diversity and Similarity
of Comets, pp. 703–706.
Evans, D. S. (1954). Teach Yourself Astronomy. London: English Universities Press Ltd.
Evensen, K. M., Wells, J. S., Petersen, F. R. et al. (1972). Speed of light from direct
frequency and wavelength measurements of the methane-stabilized laser. PRL, 29,
1346–1349.
Ferdman, R. D., Stairs, I. H., Kramer, M. et al. (2010). The precise mass measurement of
the intermediate-mass binary pulsar PSR J1802–2124. ApJ, 711, 764–771.
Ferriere, K. (2001). The interstellar environment of our galaxy. Rev. Mod. Phys., 73, 1031–
1066.
Fixler, J. B., Foster, G. T., McGuirk, J. M. & Kasevich, M. A. (2007). Atom interferometer
measurement of the Newtonian constant of gravity. Science, 315, 74–77.
Flasar, F. M., Achterberg, R. K., Conrath, B. J. et al. (2005). Titan’s atmospheric tempera-
tures, winds and composition. Science, 308, 975–978.
Fouqué, P., Chevallier, L., Cohen, M. et al. (2000). An absolute calibration of DENIS.
A&AS, 141, 313–317.
Funk I. K., Thomas C., Vizetelly F. M. & Funk C. E. eds. (1946). Funk & Wagnalls New
Standard Dictionary of the English Language, Vols I and II. London: The Waverley
Book Company Ltd.
Gezari, D. Y., Schmitz, M., Pitts, P. S. & Mead, J. M. (1993). Catalog of Infrared
Observations, 3rd edn. Greenbelt, MD: NASA reference publication, p. 1294.
Girard, G. (1994). The third periodic verification of national prototypes of the kilogram
(1988–1992). Metrologia, 31, 317–336.
Gnedin, Y. N. (1997). Chromospheres, activity and magnetic fields. In Fundamental Stellar
Properties: The Interaction Between Observation and Theory, ed. T. R. Bedding et al.
Dordrecht: Kluwer, pp. 245–252.
Gould, A., Udalski, A., Monard, B. et al. (2009). The extreme microlensing event
OGLE-2007-BLG-224: Terrestrial parallax of a thick-disk brown dwarf. ApJL, 698,
147–151.
Hansen, C. J., Kawaler, S. D. & Trimble, V. (2004). Stellar Interiors: Physical Principles,
Structure and Evolution. New York: Springer-Verlag.
222 References

Harris, III, D. L., Strand, K. Aa. & Worley, C. E. (1963). Empirical data on stellar masses,
luminosities and radii. In Basic Astronomical Data, ed. K. Aa. Strand. Chicago IL:
University of Chicago Press, pp. 273–292.
Hawkins, J. M. & Allen, R. eds. (1991). The Oxford Encyclopedic English Dictionary.
Oxford: Clarendon Press.
Hearnshaw, J. B. (1986). The Analysis of Starlight. Cambridge: Cambridge University
Press.
Hearnshaw, J. B. (1996). The Measurement of Starlight: Two Centuries of Astronomical
Photometry. Cambridge: Cambridge University Press.
Henriksen, M. J. & Tittley, E. R. (2002). CHANDRA observations of the A3266 galaxy
cluster merger. ApJ, 577, 701–709.
Herbst, E. (2001). In Encyclopedia of Astronomy & Astrophysics, Vol. 2, ed. P. Murdin,
Bristol: IoP. pp. 1258–1266.
Hohle, M. M. Eisenbeiss, T., Mugrauer, M. et al. (2009). Photometric study of the OB
star clusters NGC1502 and NGC2169 and mass estimation of their members at the
University Observatory Jena. AN, 330, 511.
Holland, W. S., Greaves, J. S., Zuckerman, B. et al. (1998). Submillimeter images of dusty
debris around nearby stars. Nature, 392, 788–791.
Hollis, J. M., Chin, G. & Brown, R. L. (1985). An attempt to detect mass loss from α Lyrae
with the VLA. ApJ, 294, 646–648.
Hovestadt, D., Hilchenbach, M., Bürgi, A. et al. (1995). CELIAS: Charge, Element and
Isotope Analysis System for SOHO. Sol. Phys., 162, 441–481.
Huang, T.-Y., Han, C.-H., Yi, Z.-H., & Xu, B.-X. (1995). What is the astronomical unit of
length? A&A, 298, 629–633.
Hubble, E. P. (1926). Extragalactic nebulae. ApJ, 64, 321–369.
Hubble, E. P. (1929). A relation between distance and radial velocity among extra-galactic
nebulae. PNAS, 15, 168–173.
Irwin, J. A. (2007). Astrophysics: Decoding the Cosmos. Chichester: John Wiley & Sons
Ltd.
Józsa, G. I. G., Garrett, M. A., Oosterloo, T. A. et al. (2009). Revealing Hanny’s Voorwerp:
radio observations of IC2497. A&A, 500, L33–L36.
Karovska, M. & Sasselov, D. (2001). In Encyclopedia of Astronomy & Astrophysics, vol. 4,
ed. P. Murdin. Bristol: IoP, pp. 3068.
Kaye, G. W. C. & Laby, T. H. (1959). Tables of Physical and Chemical Constants and some
Mathematical Functions, 12th edn. London: Longmans, Green and Co.
Keenan, P. C. (1963). Classification of stellar spectra. In Basic Astronomical Data, ed. K. Aa.
Strand. Chicago: University of Chicago Press, pp. 78–122.
King, H. C. (1955). The History of the Telescope. London: Charles Griffin & Co. Ltd.
Kirkpatrick, J. D. (2005). New spectral types L and T. ARA&A., 43, 195–246.
Klioner, S. A. (2007). Relativistic scaling of astronomical quantities and the system of
astronomical units. A&A, 478, 951–958.
Korhonen, H., Hubrig, S., Berdyugina, S. V. et al. (2009). First measurement of the magnetic
field on FK Com and its relation to the contemporaneous star-spot locations. MNRAS,
395, 282–289.
Kurucz, R. L. (1979). Model atmospheres for G, F, A, B and O stars. ApJS, 40, 1–340.
Kutner, M. L. (2003). Astronomy: A Physical Perspective, 2nd edn. Cambridge: Cambridge
University Press.
Landstree, J. D. (2001). In Encyclopedia of Astronomy & Astrophysics, vol. 2, ed. P. Murdin.
Bristol: IoP, pp. 1508–1514.
Lang, K. R. (2006). Astrophysical Formulae. Vols I and II: Berlin: Springer-Verlag.
 References 223

Leggett, S. K., Cushing, M. C., Saumon, D. et al. (2009). The physical properties of four
600 K T dwarfs. ApJ, 695, 1517–1526.
Leschiutta, S. (2001). In Encyclopedia of Astronomy & Astrophysics, vol. 4, ed. P. Murdin.
Bristol: IoP, pp. 3313–3315.
Liddell, H. G. & Scott, R. (1996). A Greek – English Lexicon. Oxford: Clarendon Press.
Lignières, F., Petit, P., Böhm, T. & Aurière, M. (2009). First evidence of a magnetic field
on Vega. A&A, 500, L41–L44.
Longair, M. S. (1989). Royal Observatory Edinburgh: Research and Facilities Handbook.
Edinburgh: ROE, p. 79.
Love, S. G. & Brownlee, D. E. (1993). A direct measurement of the terrestrial accretion rate
of cosmic dust. Science, 262, 550–553.
Lovell, B. & Clegg, J. A. (1952). Radio Astronomy. London: Chapman & Hall Ltd.
Lucas, P. W., Tinney, C. G., Burningham, B. et al. (2010). Discovery of a very cool brown
dwarf amonst the ten nearest stars to the Solar System. arXiv:1004.0317v1[astro-
ph.SR].
Lyne, A. G., Pritchard, R. S. & Graham-Smith, F. (1993). 23 years of Crab pulsar rotational
history. MNRAS, 265, 1003–1012.
McCarthy, D. D. & Petit, G. (2003). IERS Technical Note, 32. IERS Conventions (2003).
Frankfurt am Main: Verlag des Bundesamtes für Kartographie und Geodäsie.
Manchester, R. N., Hobbs, G. B., Teoh, A. & Hobbs, M. (2005). The Australia Telescope
National Facility pulsar catalogue. AJ, 129, 1993–2006,
Mather, J. C., Cheng, E. S., Cottingham, D. A. et al. (1994). Measurement of the cosmic
microwave background spectrum by the COBE FIRAS instrument. ApJ, 420, 439–444.
Mayes, V. (1994). Unit prefixes for use in astronomy. QJRAS, 35, 569–572.
Menzel, D. H. (1960). Fundamental Formulas of Physics. Vols 1 and 2. New York: Dover
Publications Inc.
Mills, I., Cvitaš, T., Homann, K., Kallay, N. & Kuchitsu, K. (1993). Quantities, Units and
Symbols in Physical Chemistry, 2nd edn. International Union of Pure and Applied
Chemistry. Oxford: Blackwell Science.
Mills, I. M., Mohr, P. J., Quinn, T. J., Taylor, B. N. & Williams, E. R. (2005). Redefinition
of the kilogram: a decision whose time has come. Metrologia, 42, 71–80.
Miura, T., Arakida, H., Kasai, M. & Kuramata, S. (2009). Secular increase of the astronom-
ical unit: a possible explanation in terms of the total angular-momentum conservation
law. PASJ, 61, 1247–1250.
Moffatt, J. (1950). A New Translation of the Bible. London: Hodder and Stoughton Ltd.
Mohr, P. J., Taylor, B. N. & Newell, D. B. (2007). The fundamental physical constants. Phys.
Today, 60(7), 52–55.
Monet, D. G., Levine, S. E., Canzian, B. et al. (2003). The USNO-B Catalog. AJ, 125,
984–993.
Moore, C. E. (1959). A Multiplet Table of Astrophysical Interest. Washington, DC: US
Department of Commerce.
Moore, P. A. (2001). In Encyclopedia of Astronomy & Astrophysics, vol. 1, ed. P. Murdin.
Bristol: IoP, p. 464.
Muhleman, D. O., Holdridge, D. B. & Block, N. (1962). The astronomical unit determined
by radar reflections from Venus. AJ, 67, 191–203.
Murray, C. A. (1989). The transformation of coordinates between the systems of B1950.0
and J2000.0, and the principal galactic axes referred to J2000.0. A&A, 218, 325–329.
Noerdlinger, P. D. (2008). Solar mass loss, the astronomical unit, and the scale of the Solar
System. 2008arXiv0801.3807N, 1–31.
Pannekoek, A. (1961). A History of Astronomy. London: George Allen & Unwin Ltd.
224 References

Patilla, P. (2000). Measuring Up Size. London: Belitha Press Ltd.

Pease, D. O., Drake, J. J. & Kashyap, V. L. (2006). The darkest bright star: CHANDRA
X-ray observations of Vega. ApJ, 636, 426–431.
Penzias, A. A. & Wilson, R. W. (1965). A measurement of excess antenna temperature at
4080 Mc/s. ApJ, 142, 419–421.
Perryman, M. A. C., Lindegren, L., Kovalevsky, J. et al. (1997). The Hipparcos Catalogue.
A&A, 323, L49–L52.
Pitjeva, E. V. (2005). High-precision ephemerides of planets – EPM and determination of
some astronomical constants. Solar System Res., 39, 176–186.
Pitjeva, E. V. & Standish, E. M. (2009). Proposals for the masses of the three largest asteroids,
the Earth – Moon mass ratio and the astronomical unit, Celest. Mech. Dyn. Astr., 103,
365–372.
Reddish, V. C. (1978). Stellar Formation. Oxford: Pergamon Press.
Ridpath, I. (2007). Oxford Dictionary of Astronomy, 2nd edn. Oxford: Oxford University
Press.
Rufener, F. & Nicolet, B. (1988). A new determination of the Geneva photometric passbands
and their absolute calibration. A&A, 206, 357–374.
Rutherford, E. (1929). Origin of actinium and age of the Earth. Nature, 123, 313–314.
Sackman, I.-J., Boothroyd, A. I. & Kraemer, K. E. (1993). Our Sun. III: Present and future.
ApJ, 418, 457–468.
Saha, M. N. (1921). On the physical theory of stellar spectra. Proc. Roy. Soc. London, A99,
135–138.
Sandage, A. (1961). The Hubble Atlas of Galaxies. Publications of the Carnegie Institution
of Washington No. 618, Washington.
Schramm, D. N. (1990). The Age of the Universe: Concordance. In Astrophysical Ages
and Dating Methods, ed. E. Vangioni–Flam, et al. Gif-sur-Yvette, France: Editions
Frontières, pp. 365–383.
Shane, C. D. & Wirtanen, C. A. (1954). The distribution of the extragalactic nebulae. AJ,
59, 285–306.
Shklovskii, I. S. (1970). Possible causes of the scalar increase in pulsar periods. Sov. Astr.,
13, 562–565.
Sidgwick, J. B. (1955). Observational Astronomy for Amateurs. London: Faber & Faber
Ltd.
Spencer-Jones, H. (1956). General Astronomy. London: Edward Arnold (Publishers) Ltd.
Stencel, R. E. Creech-Eakman, M., Hart, A. et al. (2008). Interferometric studies of the
extreme binary  Aurigae: pre-eclipse observations. ApJL, 689, L137–L140.
Sterken, C. & Manfroid, J. (1992). Astronomical Photometry: A Guide. Dordrecht: Kluwer
Academic Publishers.
Straižys, V. (1992). Multicolor Stellar Photometry. Tucson, AZ: Pachart Publishing House.
Struve, O., Lynds, B. & Pillans, H. (1959). Elementary Astronomy. New York: Oxford
University Press.
Sumi, T., Udalski, A., Szymański, M. et al. (2004). The Optical Gravitational Lensing
Experiment: catalogue of stellar proper motions in the OGLE II Galactic bulge fields.
MNRAS, 348, 1439–1450.
Taylor, J. H., Manchester, R. N. & Lyne, A. G. (1993). Catalog of 558 pulsars. ApJS, 88,
529–568.
Thompson, G. I., Nandy, K., Jamar, C. et al. (1978). Catalogue of Stellar Ultraviolet Fluxes.
London: Science Research Council.
Thorsett, S. E. (2001). In Encyclopedia of Astronomy & Astrophysics, vol. 3, ed. P. Murdin.
Bristol: IoP, pp. 2177–2183.
 References 225

Trimble, V. (2010). A review of An Introduction to the Theory of Stellar Structure and

Evolution, 2nd edn. Observatory, 130, 185–186.
Trumpler, R. J. (1930). Lick Observatory Bulletin, No. 420.
Urry, C. M. (1988). X-ray timing of active galactic nuclei. Lecture Notes in Physics, 307,
257–274.
van Biesbroeck, G. (1963). Star catalogues and charts. In Basic Astronomical Data, ed. K.
Aa. Strand. Chicago IL: University of Chicago Press, pp. 471–480.
van Duinen, R. J., Aalders, J. W. G., Wesselius, P. R. et al. (1975). The ultraviolet experiment
onboard the Astronomical Netherlands Satellite – ANS. A&A, 39, 159–163.
van Leeuwen, F. (2007). Hipparcos, the New Reduction of the Raw Data. London: Springer.
Verbiest, J. P. W., Lorimer, D. R. & McLaughlin, M. A. (2010). Lutz – Kelker bias in pulsar
parallax measurements. MNRAS, 405, 564–572.
Viala, Y. P. (1986). Chemical equilibrium from diffuse to dense interstellar clouds. I: Galactic
molecular clouds. A&AS, 64, 391–437.
Wall, J. V. & Jenkins, C. R. (2003). Practical Statistics for Astronomers. Cambridge:
Cambridge University Press.
Weaver, T. A., Zimmerman, G. B. & Woosley, S. E. (1978). Pre-supernova evolution of
massive stars. ApJ, 225, 1021–1029.
Weisberg, J. M., Cordes, J. M., Kuan, B. et al. (2004). Arecibo 430 MHz pulsar polarimetry:
Faraday rotation measures and morphological classifications. ApJS, 150, 317–341.
Wesselius, P. R., van Duinen, R. J., Aalders, J. W. G. & Kester, D. (1980). Ultraviolet colours
of main-sequence stars. A&A, 85, 221–232.
Wesselius, P. R., van Duinen, R. J., de Jong, A. R. W. et al. (1982). ANS ultraviolet
photometry, catalogue of point sources. A&AS, 49, 427–474.
Weygand, J. M., Ipavich, F. M., Wurz, P. Paquette, J. A. & Bochsler, P. (1999). Plasma
dynamics and diagnostics in the solar transition region and corona: Determination of
the argon isotope ratio of the solar wind using SOHO / CELIAS / MTOF. Proceedings
of 8th SOHO Workshop, ESA SP-446, pp. 701–705.
White, V., ed (2008). British Astronomical Association, London: BAA. Handbook for 2008.
Wieser, M. E. & Berglund, (2009). Atomic weights of the elements 2007 (IUPAC technical
report). Pure Appl. Chem., 81, 2131–2156.
Wilkins, G. A. (1989). IAU Style Manual, Comm. 5. In IAU Transactions, XXB.
Winget, D. E., Nather, R. E., Clemens, J. C. et al. (1994). Whole Earth Telescope observa-
tions of the DBV white dwarf GD358. ApJ, 430, 839–849.
Wolfe, A. M., Jorgenson, R. A., Robishaw, T., Heiles, C. & Prochaska, J. A. (2008). An
84 μG magnetic field in a galaxy at redshift z = 0.692. Nature, 455, 638–640.
Wollard, E. W. & Clemence, G. M. (1966). Spherical Astronomy. New York and London:
Academic Press.
Wright, E. L. (2006). A cosmology calculator for the World Wide Web. PASP, 118,
1711–1715.
Yang, Y.-G. (2009). BVR observations and period variation of the neglected contact binary
V343 Orionis. PASP, 121, 699–707.
Zacharias, N., Urban, S. E., Zacharius, M. I. et al. (2000). The first US Naval Observatory
CCD astrograph catalog. AJ, 120, 2131–2147.
 Index

2MASS (Two Micron All Sky Survey) photometry, barycentric coordinate time (TCB), 56
132 barycentric dynamic time (TDB), 56
baryonic matter, 207
absolute elliptical orbit, 102 Bayer, 208
absolute magnitude, 140 bequerel, 66
absolute zero, 148 binary star, 102
acceleration, 88 BIPM (Bureav International des Poids et Mesures), 3,
accuracy definition, 5 10
accurate definition, 5 blackbody radiation, 151
Airy transit circle, 37 Boltzmann constant, 146, 152
altazimuth coordinate system, 38 Bonner Durchmusterung, 210
altitude, 38 bound-bound transitions, 158
ampere, 174 bound-free transition, 158
angstrom unit, 27 Boyle’s law, 201
angular acceleration, 60 brief history of units, 7
angular momentum, 111 brightness, 133
angular quantum number, 159 brightness temperature, 154
angular velocity, 60
ANS photometry, 132 candela, 113
antenna temperature, 154 Cannon, 210
apparent magnitude, 140 Cavendish, 94
area, 89 celestial equator, 32, 39
Arecibo radio telescope, 192 celestial latitude, 40
Argelander, 210 celestial longitude, 40
asteroseismology, 97, 104 celestial sphere, 36
astrochemistry, 202 CELIAS, 200
astronomical classification, 206 celsius to kelvin temperature, 148
astronomical taxonomy, 206 centre of a sphere, 36
astronomical unit, 22, 75 cerenkov radiation, 136
atomic mass unit, 197 cesium fountain atomic clocks, 68
atomic time, 55 CGPM (Conférence Général des Poids et Mesures),
attribute, 206 3, 9
average atomic mass, 198 cgs Gaussian units, 175
Avogadro’s constant, 197 cgs system, 14
Avogadro’s law, 201 CHANDRA, 52
axis of a great circle, 36 characteristic age, 68
azimuth, 38 Charles’ law, 201
colour index, 140
BAAS (British Association for the Advancement of colour temperature, 161
Science), 10 column mass density, 107
barycentre, 102 Condamine, 9

226
 Index 227

constellation, 208 ESO archive, 217

continuum, 158 esu electrostatic units, 10, 175
conversion factors to radians, 30 excitation potential, 159
converting proper motions, 63 exoplanet, 106
coordinated universal time (UTC), 55 exothermic reaction, 204
cosmochemistry, 202
cosmological red shift, 88 fahrenheit to kelvin temperature, 148
coulomb, 155 Faraday constant, 201
Coulomb’s law, 176, 177 Faraday rotation, 191
Crab pulsar, 68 First Point of Aries, 35, 56
Crux, 49 Flamsteed, 208
Curie, 66 flattening, 72
flux density, 130
d, 20 force, 108
daughter nucleus, 66 franklin, 176
declination, 40 free electrons, 158
Delambre, 9, 72 free-free transitions, 158
DENIS photometry, 132 fundamental epoch, 42
density, 106
galactic centre, 42
deuterium, 198
galactic coordinate system, 42
dex, 20
galactic equator, 42
dimension base units, 24
galactic latitude, 42
dimension definition, 5
galactic longitude, 42
dimension derived units, 24
galactic longitude of the ascending node, 42
dimensional analysis, 3
galaxy M31, 50
dimensional equations, 25
galaxy number densities, 51
dimensionless units, 25
gamma, 182
dispersion measure, 191 gamma-rays, 135
dwarf planet, 206, 215 gas-phase reactions, 203
dynamical time, 55 Gauss, 10
Gaussian distribution, 7
e-magnitudes, 139 Gaussian gravitational constant, 94
Earth, eccentricity, 73 general precession, 41
Earth, mass, 99 general relativity solar mass units, 99
Earth, mean radius, 73 Geneva photometry, 132
Earth, rotation, 55 geocentric coordinate time (TCG), 56
Earth, size, 72 geocentric gravitational constant, 96
eclipsing binaries, 102 geocentric radius, 73
ecliptic, 39, 40 geodetic equator, 73
ecliptic coordinate system, 40 geodetic latitude, 73
effective temperature, 149 geographic poles, 36
electricity definition, 174 geographical latitude, 37
electromagnetic radiation, speed, 84 geographical longitude, 37
electron, 154 geoid, 73
electron molar mass, 202 gigaparsec, 82
electron pressure, 159 Giorgi, 10
electron transitions, 156 gravitational constant, 76, 94
electron volt, 134, 155 gravitational microlensing, 78, 97
emu electromagnetic units, 10, 175 gravitational red shift, 88
endothermic reaction, 204 great circle, 36
energy, 108 Greenwich mean time, 56
energy states, 154 Grotrian diagram, 156
entropy, 147
Epsilon Aurigae, 50 half-life, 67, 199
equatorial coordinate system, 39 HD catalog, 211
equinoctial points, 40 heliocentric gravitational constant, 96
equivalent width, 160 helioseismology, 97
error definition, 5 hertz, 60
228 Index

high-frequency bands, 134 magnetar, 183

Hipparchus, 125, 209 magnetic flux, 181
HIPPARCOS catalogue, 79 magnetic flux density, 181
HIPPARCOS / TYCHO satellite, 63 magnetic permeability of free space, 178
horizon, 38 magnetism definition, 175
Hubble classification of galaxies, 214 magnetometer, 190
Hubble Space Telescope, 214 magnitudes, 23, 137
Huygens, 9 magnitudes, advantages, 138
hydrogen atom energy levels, 156 magnitudes, disadvantages, 138
hydrostatic equilibrium, 165 Mappa Mundi, x
mass loss, 105
IAU unit recommendations, 20 mass spectrometer, 200
ideal gas laws, 165 Maury, 210
identity, 206 Maxwell, 10
International Astronomical Union (IAU), 3, 42 maxwell, 181
International atomic time (TAI), 55 Mayes’ unit prefixes, 17
International Celestial Reference Frame (ICRF), 64 mean, 6
International Electrotechnical Commission, 10 mean coordinates, 42
International Prototype Kilogram (IPK), 92 mean orbital angular velocity, 61
International Space Station, 100 measurement definition, 4
International Virtual Observatory Alliance (IVOA), megaparsec, 82
218 Merchison carbonaceous meteorite, 201
interstellar absorption, 140 Mercury’s electric tail, 194
interstellar chemistry, 203 metre, 72
interstellar clouds, 203 milliarcsecond, 22
ionization potential, 159 milliare, 8
IRAS, 132 millimeter bands, 134
ISO date time standard, 58 mistake definition, 6
isophote, 50 model stellar atmospheres, 165
Modified Julian Date (MJD), 57
jansky, 27, 32, 114, 133 molar mass, 198
Julian date, 56 molar volume of an ideal gas, 202
julian day number, 56 mole, 197
Julian year, 22, 58 mole fraction, 198
Jupiter’s lightning, 195 molecular mass, 199
Mollweide’s projection, 35
kayser, 136
moment of inertia, 109
Kepler’s third law, 75, 102
month, 58
kilogram, 92
Moon, distance, 74
kiloparsec, 81
Moon, horizontal parallax, 75
kinetic energy, 109
kinetic temperature, 147
Kurucz model stellar atmospheres, 165 nanoradian, 22
neutron molar mass, 202
leap second, 57 Newton, 76
light year, 23, 82 Newton’s second law, 103
linear velocity, 83 Newtonian gravitational constant, 96
local hour angle, 43 north celestial pole, 32, 39
local sidereal time, 43 nucleocosmochronology, 66
local thermodynamic equilibrium, 159, 165 nutation, 41, 57
longitude of the ascending node, 45
look-back time, 88 oblate spheroid, 72
lookup table, 163 oblateness, 89
low-frequency bands, 129 obliquity of the ecliptic, 40
luminous intensity, 115 optical retro-reflectors, 74
lunisolar precession, 41
Palomar testbed interferometer, 50
Méchain, 9, 72 parent nucleus, 66
Magna Carta, 8 parsec, 23, 80
 Index 229

partition function, 159 sidereal year, 40

perfect gas law, 202 SIMBAD, 216
periodic table, 199 small circle, 36
permittivity of free space, 177 SOHO, 200
photometry, 113 solar constant, 115
Picard, 9 solar electric fields, 194
Pickering, 210 solar mass, 23, 97
Planck temperature, 151 solar parallax, 77
Planck’s constant, 93, 152 Solar System masses, 100
Planck’s law, 151 solar wind particle density, 200
planet, 215 solar wind speed, 200
planetary ephemerides, 96 south celestial pole, 32, 39
planetary precession, 41 Southern Sky Survey, 65
plate scale, 47 space velocity, 86
Pogson, 125, 210 spectral classification, 154, 210
Pogson magnitudes, 162 spectral irradiance, 124
poles of a great circle, 36 spectroscopic binary, 102
potential energy, 108 sphere, 36
power, 109 spherical astronomy, 36
practical system electrical units, 10, 175 spherical triangle, 36
precise definition, 5 square arcminute, 48
pressure, 109 square arcsecond, 48
prime meridian, 37, 38 square degree, 48
probable error, 7 standard atmosphere, 202
proper-motion catalogues, 63 standard definition, 4
proper motions, 61 standard deviation, 7
proton molar mass, 202 standard error, 7
pulsar, 68, 192 standard gravitational parameter, 96
star cluster, 212
quantity definitions, 4 Stark effect, 194
statcoulombs, 176
radial velocity, 85 Stefan-Boltzmann constant, 149
radian definition, 30 Stefan–Boltzmann constant unit conversion, 149
radioactivity, 66 Stefan–Boltzmann law, 153
radiometry, 113 stellar identity, 208
radius of a sphere, 36 stellar luminosity classification, 211
rankine to kelvin temperature, 148 stellar magnitudes, 209
rationalized units, 178 stellar magnitudes to candelas, 115
reasons for standardizing units, 2 stellar masses, 102
recession velocity, 87 stellar radius, 166
red shift, 87 stellar space motions, 85
revolution period, 60 steradian, 48
right ascension, 39 Strömgren photometry, 132
rotation measure, 191 submillimeter bands, 134
rotation period, 60 Sumi catalogue, 65
rotational angular velocity, 61 surface gravity, 88, 165
Rydberg constant, 158 Système International d’Unités, 12
Rydberg formula, 158
TD1 photometry, 132
Saha–Boltzmann equation, 159 temperature scales, 147
satellite, 216 terrestrial coordinate system, 36
Saturn’s aurorae, 195 terrestrial dynamic time (TDT), 56
Secchi, 210 terrestrial equator, 37
second definition, 54 terrestrial meridians, 37
SI base units, 12 tesla, 181
SI derived units, 12 thermodynamic temperature, 146
SI unit prefixes, 17 Thomson, 10
sidereal time, 56 time conversions, 58
230 Index

time definition, 54 virtual observatories, 217

time scales, 54 volume, 89
Tower of Babel, x
transverse velocity, 86 wavenumbers, 23, 136
trigonometric parallax, 79 WEBDA, 217
tritium, 198 Wein’s displacement constant, 153
Trumpler, 213 Wein’s displacement law, 153
tuning fork diagram, 214
X-rays, 134
unit definition, 4
Universal Time (UT), 55
USNO catalogues, 217 year, 58

Vernal Equinox, 39 Zeeman effect, 186

vertical circle, 38 zenith, 38
Vilnius photometry, 132 zenith distance, 38