Chapter 2. Spectral Signatures.............................................................................................

1
2.1 Introduction ................................................................................................................ 1
2.2 Visible radiation and plant characteristics ................................................................. 2
2.3 Near-infrared radiation and plant characteristics ....................................................... 4
2.4 Middle-infrared radiation and plant characteristics ................................................... 5
2.5 Signatures at the canopy level.................................................................................... 5
2.6 Reflectance modelling of a vegetation canopy .......................................................... 7
2.6.1 Introduction ........................................................................................................ 7
2.6.2 Indices for estimating LAI ................................................................................. 9
2.6.3 The Weighted Difference Vegetation Index (WDVI)........................................ 9
2.7 References ................................................................................................................ 12
Appendix to chapter 2: Some notes on aerial photography ..................................................... 13

i
Chapter 2. Spectral Signatures

2.1 Introduction

By sight humans can use visible differences in the reflection of the sunlight to recognize
vegetation. Part of the reflected radiation observed with remote sensing (RS) techniques
coincides with the part visible to the human eye. It is therefore sensible to take advantage of
those differences in vegetation observable by everybody in the field. A general characteristic
of vegetation is its green colour during most of the growing season. The pigment chlorophyll
causes the green colour. Different forms of this pigment exist, but green-yellow chlorophyll
occurs in all photo-synthesizing plants and plays a dominant role in the absorption of solar
energy. A single crop may have various shades of green. The causes may vary: growth stage,
water supply, manuring, diseases, etc. The different green tones are an important
characteristic in the visual judgement of crops.

Differences in reflected light, not only in the green, may also originate from differences in the
surface of the foliage. Leaves may be covered with wax or have a hairy coat. These
differences in light reflection can help in the identification of plant species. Another botanical
characteristic is the structure of vegetation, which causes differences of reflection. Examples
are the positions of the leaves (phyllotaxis) and their distribution. In relation with the angle of
incident radiation this may greatly influence the reflection. Also, the growth stage or
phenological stage of the plants has an influence on the reflection. This may be caused by
differences of structure (shapes of flowers, ears, etc.) as well as by differences in colour.

The amount of reflected radiation should be considered relative to the amount of incoming
radiation in order to get a quantitative measure. This quantitative measure is called the
reflectance and can be expressed as a factor (reflectance factor) or as a percentage (reflectance
percentage). This reflectance (ρ) can be determined by dividing the reflected radiation (Lr) by
the incoming radiation (Li) for a certain wavelength (λ) or wavelength interval (∆λ):

ρ (λ ) = Lr (λ ) / Li (λ ) (2.1)

This definition is often applied with field radiometers. In general, with airborne or spaceborne
sensors the incoming radiation is not measured. The reflectance can then be determined by
comparing the reflected radiation from an unknown object (Lobj) with that from a reference
(Lref) of which we know the reflectance properties (ρref). Again this approach can be applied
for a certain wavelength or a wavelength interval:

ρ obj (λ ) = Lobj (λ ) / Lref (λ ) ∗ ρ ref (λ ) (2.2)

A third approach is to model the incoming radiation and the atmospheric effects. The model is
then used to translate the measured radiation by the sensor back to the reflectance of the
object at the Earth surface`(see chapter 4 on pre-processing).

In the 0.4-2.5 µm part of the electromagnetic spectrum, three regions may be distinguished
with respect to the interaction of the plant cover with solar EM energy: the visible, near-
infrared and middle-infrared part of the EM spectrum (Figure 2.1).

1
2.2 Visible radiation and plant characteristics

In the first range of wavelengths, the visible light (VIS) ranging from 0.4 to 0.7 µm, various
pigments, such as chlorophyll, xanthophyll (yellow), and carotene (orange), influence the
reflection. In most plant species two types of chlorophyll (a and b) determine the reflection,
mainly by absorption of blue and red light and to a lesser degree of green light (see Figure
2.1). The energy in these spectral bands is used for the displacement of electrons, which
initiates the synthesis of carbohydrates from atmospheric CO2 and absorbed groundwater.
Green-yellow chlorophyll a is present in all photo-synthesizing plants and plays a dominant
role. Higher plants and green algae contain blue-green chlorophyll b, although in small
quantities. Both chlorophylls absorb the visible light to a large extent, and have two
absorption peaks one in the blue (approx. 0.45 µm) and one in the red (approx. 0.65 µm)
region of the electromagnetic (EM) spectrum. As a result of this and also of the
hypersensitivity of the eye to green, vegetation reveals itself to the eye in various shades of
green. Subsequently, the peak of the reflectance in the visible light occurs at approx. 0.54 µm.

% VIS NIR MIR

100
transmittance
80
absorptance
(chlorophyll)

60
absorptance
40
(water)

20 reflectance

0
0.4 0.8 1.2 1.6 2.0 2.4 2.8
Wavelength (µm)

Figure 2.1: Average course of reflectance, absorptance and transmittance of a green healthy
plant leaf as a percentage of the irradiation (Knipling, 1970).

A growing plant lacking certain nutrients may produce less chlorophyll. The plant turns pale
in places, and a distinct increase in the blue and red reflectance is observed by spectrometer
measurements, bringing it to an equal level with the green reflectance. It is well known that
under certain conditions, e.g., the ageing of plants, the yellow and red pigments show up
clearly as the chlorophyll disintegrates. This is the case with deciduous trees, whose leaves
get various colours just before they fall.

For each mixture of tree species the limit of recognition of a certain species by the so-called
spectral signature is different. The spectral signature – usually reproduced graphically – is the
characteristic of the reflection of an object defined by the reflectance as a function of (plotted
against) the wavelength.

2
Also, at a certain instant there are differences of reflectance in the visible region of the EM
spectrum of different plant species growing under the same conditions. Generally these
differences occur in the green and blue reflectance (see Figure 2.2) produced by differences in
the pigment content and of the leaf surface, such as hair, wax, etc. Under field conditions one
has to take account of other factors influencing the spectral reflectance. Examples are the
conditions of nutrition and the water supply of a crop causing an immediate change in both
the signature of the leaves and the crop structure. Differences in the reflectance of visible light
are also produced by infections of the leaves' surface, e.g., by fungi spores. Well known is the
discolouring of the leaves by spores of blight, but bacterial and viral infections may also
influence leaf colour. Figure 2.3 shows an example of the reflectance curve of wheat leaves
with a heavy (1) and a light (2) infection of grain blight. The figure clearly shows that the
reflectance in the red at 0.65 µm is high for the heavily infected (1) leaves.

citrus
tomato
60
Reflectance (%)

sorghum
cotton

40

20

0
0.5 1.0 1.5 2.0 2.5
Wavelength (µm)
Figure 2.2: Spectral reflectance of representative leaves of four agricultural crops measured
with a Beckman DK-2A laboratory spectrometer (Shay et al., 1970).

100

80
Reflectance (%)

60

40
(1)
20 (2)

0
0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
Wavelength (µm)
Figure 2.3: Wheat leaves heavily (1) and lightly (2) infected by grain blight (Keegan et al.,
1956).

3
2.3 Near-infrared radiation and plant characteristics

The second range of wavelengths from 0.7 to 1.3 µm approx. (near-infrared radiation, NIR) is
mainly determined by the absence of absorption by pigments (see Figure 2.1). This means that
the energy passes through the leaf (the leaf is transparent) or that it is reflected. It is apparent
from the various reflectance curves (measurements on individual leaves!) that approximately
50% of the NIR energy is reflected by the leaf. However, Figure 2.2 shows that this
percentage varies widely for different plant species. During the development of a leaf the
reflectance in this range of wavelengths also changes.

Various experiments have tried to obtain an insight into which parts of the leaf are responsible
for NIR reflection. It has been established for this range of wavelengths that a leaf becomes
very transparent if the air channels between the cells of the leaf are filled with fluid. This gave
reason to a theory that reflection takes place in the leaf at the transition of air and cellulose
cell walls. The reflection of NIR appeared to increase at first when leaves shrivelled up. This
could be explained by ruptures between the cells, producing more transitions of air to cell
walls (see Figure 2.4).

There are also differences in reflectance between the leaves of groups of plant species, known
as monocotyledons and dicotyledons. In principle, leaf sections of the latter species (e.g.,
potatoes, sugar beet, Papilionaceae) show two different layers of cells between the outer
walls. The upper layer is the palisade parenchyma, which consists of tightly packed oblong
cells perpendicular to the epidermis. The lower layer, the sponge parenchyma, has a
completely different structure; the cells are rounder with many intercellular air cavities.
Hence, it appears that the reflection occurs mainly in the sponge parenchyma at the transition
of air and the cell wall. In contrast, monocotyledonous leaves (e.g., cereals and grasses) do
not contain the sponge parenchyma layer with many air cavities. Apart from the theory that
reflection takes place mainly in the air cavities of the sponge parenchyma, it has been
suggested that reflection is also determined by the number of cell walls parallel with the
epidermis of the leaf. These walls are certainly more numerous in leaves with additional
layers of palisade cells. Of course, the reflection also depends on the orientation of the leaf in
the vegetation. The preceding theories are based on measurements of individual leaves.
100
12 dried leaves
80
Reflectance (%)

60
1 dried leaf
12 fresh leaves
40

1 fresh leaf
20

0
0.5 1.0 1.5
2.0
Wavelength (µm)
Figure 2.4: Total reflectance of fresh and dried cotton leaves measured with a laboratory
spectrometer (Shay et al., 1970).

4
In summary, it can be stated that the uppermost leaves of vegetation reflect a considerable
part in the NIR region (up to 1.3 µm), and that the remainder is transmitted (see Figure 2.1)
hitting the underlying leaves. Measurements demonstrate an increasing reflectance in this NIR
region as more leaves occur in the layers. Dependent on the leaf angle distribution, the
reflectance increases up to five or six leaves underneath one another. This can also be
demonstrated in the laboratory by piling up leaves and measuring the reflectance with a
spectrometer (Figure 2.4).

2.4 Middle-infrared radiation and plant characteristics

In the third region of wavelengths ranging from 1.3 to 2.5 µm (called middle-infrared (MIR)
or shortwave infrared (SWIR)), a great deal of energy is absorbed by water in the cells (see
Figure 2.1). The figure shows, like Figure 2.2 and Figure 2.4, that the absorption peaks fall at
1.9 and 1.4 µm. It should be pointed out that weak absorption bands of water also occur at 1.1
and 0.96 µm. From Figure 2.4 it is evident that dried leaves, in contrast with fresh ones, show
a considerably higher reflectance in the range from 1.3 to 2.5 µm, and that in the curves the
dips caused by water absorption have vanished. Measurements on leaves with more gradual
differences of moisture content gave the reflectance curves shown in Figure 2.5. The
reflectance in the region of 0.7 to 1.3 µm as well as in the region of 1.3 to 2.5 µm increases
with a decreasing moisture content. However, in the latter region the shape of the curve also
changes.

100 < 40% moisture content

40-54% moisture content
54-66% moisture content
80
Reflectance (%)

> 66% moisture content

60

40

20

0
0.5 0.9 1.3 1.7 2.1 2.5
Wavelength (µm)

Figure 2.5: Influence of the moisture content of maize leaves on the spectral reflectance
(Hoffer & Johannsen, 1969).

2.5 Signatures at the canopy level

To obtain an impression of the spectral behaviour of various natural objects under different
conditions, the reflectance of some objects will first be discussed as a function of the
wavelength (spectral signature). Subsequently, some results from experimental agricultural
fields will be given, where the reflectance is expressed as a function of time (temporal
signature).

5
 50
barley
40
Reflectance (%)

dry soil
30

20

10 wet soil

0
0.4 0.5 0.6 0.7 0.8 0.9 1.0
Wavelength (µm)

Figure 2.6: Spectral reflectance of a green barley crop as compared to the reflectance of a wet
and dry soil (field spectroradiometer measurements at the Minderhoud Farm, 1982).

The soil (Figure 2.6) shows a constant, small increase of reflectance in the investigated part of
the spectrum (0.4-1.0 µm). The measurements were carried out with a spectroradiometer in
the field. The wet soil clearly has a lower reflectance than the dry soil. There is also an
obvious contrast between the reflectances of bare soil and vegetation in the visible (red and
green) and in the NIR regions of the spectrum. In the visible part, absorption of radiation is
caused by the chlorophyll in the leaves, which is particularly strong in the red and the blue
regions. As a result, a low reflectance occurs for green vegetation in this part of the spectrum.
Absorption of radiation by vegetation is very low in the NIR region. Therefore, the
reflectance of vegetation is larger than that of bare soil in the NIR region.

Figure 2.7 shows the progress of the reflectance of a barley crop during the growing season,
obtained by means of multispectral aerial photography. This is the recording of radiation on
black-and-white films in separate small spectral bands by a selection of films and filters. The
green and red reflectances generally showed a similar pattern. The reflectance in the visible
bands (green and red) decreased with increasing growth at the beginning of the growing
season (days 0-75). On the first recording date the soil was bare, resulting in almost equal
reflectance in all bands. During complete soil cover (days 75-100) the reflectance remained
almost constant in the visible bands. At the end of the growing season (from day 100) the
reflectance increased again as a result of the senescence (yellowing) of the crop. Generally,
the effects in the NIR band were opposite to those in the visible bands.

6
 60

50
Reflectance (%)

40 bare soil

30 NIR

20 red
green
10

0 50 100 150
days after sowing

Figure 2.7: Reflectances in the green, red and NIR regions as a function of time for a barley
crop under normal cultivation conditions (Minderhoud Farm, 1982).

2.6 Reflectance modelling of a vegetation canopy

2.6.1 Introduction

By considering a crop to be built up of horizontal (leaf-) layers, an impression of the crop

characteristics which might be estimated directly from reflectance measurements (Figure 2.8)
can be obtained.

The reflectance and the transmittance of a green leaf in the visible part of the EM spectrum
amounts to 10% or less each (see also Figure 2.1). This means the absorptance is at least 80%.
If 10% of the incident radiation is reflected by a top leaf, the contribution to the total
measured reflectance of a second leaf under the top leaf would be approximately 1% of the
reflectance of that top leaf. This implies that, in the visible region, the reflectance of only the
top layer would determine the total reflectance of a crop. The visual observation of plants
confirms this reasoning. Hence, the large contrast in reflectances of bare soil and vegetation in
the visible region will be ideal for estimating the degree of soil cover, the reflectance attaining
a minimum value (maximum absorption) at the instant that the soil is completely covered.

In the NIR part of the spectrum, the reflectance and transmittance of a green leaf amounts to
approximately 50% each (see Figure 2.1). A green leaf hardly absorbs any NIR radiation.
Under these conditions, leaves or crop layers under the top layer contribute significantly to
the total measured reflectance. In the simplest case of the reflectance and transmittance both
amounting to 50%, the contribution of a second leaf (or layer) would be about 15% of the
incident radiation, which is not negligible. The contributions of the successive leaves or layers
decrease exponentially. This multiple reflectance denotes the NIR reflectance to be
particularly suitable for estimating the so-called leaf area index (LAI) ("counting the number
of leaf-layers"). This LAI is defined as the total one-sided leaf area per unit ground area.

These considerations grossly simplify the reality, but they do explain the general concept.

7
 VISIBLE NIR
100 100
50
10 12.5
0.1 3.1
1 1
10 1 50 12.5
25 6.2
2 2
1
25

3 3

Figure 2.8: Schematic illustration of reflectance and transmittance of radiation through crop
layers (leaf canopy) in the visible and NIR regions, respectively.

When modelling the relationship of reflectance and crop characteristics, a number of

disturbing factors should be considered. One of these is the soil reflectance, strongly
determined by the moisture content of the soil. This is explained in Figure 2.9 by reflectance
measurements in the red and NIR regions of a barley crop (from a different growing season
than that in Figure 2.7). The influence of the soil moisture content on the reflectance was the
same in both bands. The reflectance was relatively low at a high soil moisture content (days
54 and 84 after sowing), and relatively high at a low moisture content (days 40 and 60 after
sowing) (see arrows in Figure 2.9).

50 wet

40
Reflectance (%)

30 wet
NIR
20
red

10 dry
dry

0 50 100 150

days after sowing

Figure 2.9: Reflectances in the red and NIR regions as a function of time (barley, Minderhoud
Farm, 1983). An upward arrow indicates dry soil; a downward arrow indicates wet soil.

8
The soil reflectance greatly influences the relationships between the reflectance and the
degree of soil cover, and between the reflectance and the LAI. With a low degree of soil cover
the soil reflectance contributes largely to the total reflectance measured in the various spectral
bands. The moisture content of a particular soil type will be the most important factor
determining the soil reflectance. It is important to know how this happens, and to what extent
this influence depends on the wavelength in the part of the spectrum concerned.

2.6.2 Indices for estimating LAI

The green, red and NIR reflectances could be employed as variables to estimate, for instance,
the LAI. Many investigations have been conducted to assess vegetation characteristics, such
as biomass and LAI, by means of combinations of reflectances in various spectral bands. Such
a combination of reflectance values, the vegetation index, also serves to correct for
undesirable influences of varying soil reflectance or atmospheric conditions on the results.
These kinds of disturbances are particularly unwanted in spatial and multitemporal analyses.
In addition, a vegetation index for estimating the LAI should still be sensitive to changes in
LAI when the soil is fully covered (LAI > 2 or 3). This means that the NIR reflectance should
play a central role in such an index.

The first investigations into vegetation indices concerned the NIR/red ratio by Rouse et al.
(1973, 1974). Rouse and his colleagues found this ratio to be suitable – when applied to
satellite data – for the estimation of crop characteristics owing to a partial correction for the
solar position and atmospheric influence. They also used the normalized difference vegetation
index (NDVI = [NIR-red]/[NIR+red]) for the same purpose. Many more vegetation indices
have been developed and can be found in the literature. The next section will illustrate one
particular example.

2.6.3 The Weighted Difference Vegetation Index (WDVI)

A simplified, semi-empirical reflectance model for estimating LAI of a green canopy

(vegetative stage) was introduced by Clevers (1988, 1989). In this model it is assumed that in
the multitemporal analysis the soil type is given and the soil moisture content is the only
varying property of the soil. For estimating LAI, a "corrected" (adjusted) NIR reflectance was
calculated by subtracting the contribution of the soil in line of sight (by the sensor) from the
measured reflectance of the composite canopy-soil scene. This corrected NIR reflectance was
ascertained as a weighted difference between the measured NIR and red reflectances (called
WDVI = weighted difference vegetation index), assuming that the ratio of NIR and red
reflectances of bare soil is constant, independent of soil moisture content (which assumption
is valid for many soil types). Subsequently, this WDVI was used for estimating LAI
according to the inverse of an exponential function.

The simplified reflectance model derived by Clevers (1988, 1989) consists of two steps.
Firstly, the WDVI is calculated as:

WDVI = rir − C ∗ rr (2.3)

with rir = total measured NIR reflectance

rr = total measured red reflectance

9
and

C = rs ,ir / rs ,r (2.4)

rs,ir = NIR reflectance of the soil

rs,r = red reflectance of the soil.

Secondly, the relation between WDVI and LAI is modelled as:

LAI = −1 / α ∗ Ln(1 − WDVI / WDVI ω ) (2.5)

with α describing the rate with which the function of Equation 2.5 runs to its asymptotic
value, and WDVIω as the asymptotic limiting value for the WDVI. Parameters α and WDVIω
have to be estimated empirically from a training set (Clevers, 1988). The combination of
Equation 2.3 and Equation 2.5 is the simplified, semi-empirical, reflectance model: CLAIR
model ("Clevers Leaf Area Index by Reflectance" model).

The main assumption is that C is a constant, meaning that the ratio of the NIR and red
reflectance of the soil is independent of the soil moisture content. The validity of this
assumption for many soil types is confirmed by results obtained by e.g. Condit (1970) and
Stoner et al. (1980). For many soil types, there is only a slight monotonic increase in
reflectance with increasing wavelength (e.g. Condit, 1970).

As an example, the WDVI as a function of time is shown in Figure 2.10 for the barley crop
used in Figure 2.9. The WDVI is determined as the difference between the NIR and red
reflectances in Figure 2.9. In case of the soil on the Minderhoud Farm, the value of 1 can be
used for C in Equation 2.4. This value is calculated by determining the soil reflectance at
arbitrary instants. The influence of the moisture content of the soil during the beginning of the
growing season is eliminated in this way (compare NIR and red reflectances in Figure 2.9).

Finally, Figure 2.11 shows an example of the regression of the LAI on the WDVI for barley
in the vegetative growth stage. The practical usefulness of the model described above is
treated extensively by Clevers (1986).

Developments in this field over the last decade will be dealt with in a subsequent remote
sensing course.

10
 WDVI (%) 50
wet
40

30

20 wet

10
dry

0 50 100 150
dry
days after sowing

Figure 2.10: WDVI as a function of time (barley, Minderhoud Farm, 1983). An upward arrow
indicates dry soil; a downward arrow indicates wet soil.

LAI
8

6
LAI = -1/α • ln(1 - WDVI / WDVI∞)

0 20 40 60
WDVI

Figure 2.11: Regression of LAI on WDVI of a barley crop (vegetative stage).

11
2.7 References

Clevers, J.G.P.W., 1986. Application of remote sensing to agricultural field trials. Ph.D.
Thesis, Agricultural University Wageningen Papers 86-4, 227 pp.
Clevers, J.G.P.W., 1988. The derivation of a simplified reflectance model for the estimation
of leaf area index. Remote Sens. Environ. 25: 53-69.
Clevers, J.G.P.W., 1989. The application of a weighted infrared-red vegetation index for
estimating Leaf Area Index by correcting for soil moisture. Remote Sens. Environ. 29:
25-37.
Condit, H.R., 1970. The spectral reflectance of American soils. Photogramm. Eng. Remote
Sens. 36: 955-966.
Hoffer, R.M. & C.J. Johannsen, 1969. Ecological potentials in spectral signature analysis. In:
Remote Sensing in Ecology. Ed. P.L. Johnson, Univ. of Georgia Press, Athens, Georgia,
pp. 1-16.
Keegan, H.J. et al., 1956. Spectrometric and colorimetric study of diseased cereal crops. US
Department of Commerce.
Knipling, E.B., 1970. Physical and physiological basis for the reflectance of visible and near-
infrared radiation from vegetation. Remote Sens. Environ. 1: 155-159.
Rouse, J.W. Jr., R.H. Haas, J.A. Schell & D.W. Deering, 1973. Monitoring vegetation
systems in the Great Plains with ERTS. Earth Res. Techn. Satellite-1 Symp., Goddard
Space Flight Center, Washington D.C., pp. 309-317.
Rouse, J.W. Jr., R.H. Haas, D.W. Deering, J.A. Schell & J.C. Harlan, 1974. Monitoring the
vernal advancement and retrogradation (green wave effect) of natural vegetation.
NASA/GSFC Type III Final Report, Greenbelt, Md., 371 pp.
Shay, J.R. et al., editors, 1970. Remote Sensing, with special reference to agriculture and
forestry. Washington, Nat. Acad. of Sciences, 424 pp.
Stoner, E.R., M.F. Baumgardner, L.L. Biehl & B.F. Robinson, 1980. Atlas of soil reflectance
properties. Agric. Exp. Station, Purdue Univ., W-Lafayette, Indiana, Res. Bull. 962, 75
pp.

12
Appendix to chapter 2: Some notes on aerial photography
Aerial photographic systems operate in the visible and the near-infrared (NIR) parts of the
EM spectrum. The region of the photographic wavelengths generally ranges from 0.3 to 0.9 µm.
Radiation (reflected by objects on the Earth's surface) touches a light-sensitive coating through
an optical system (camera). This coating (film) acts as the photographic receptor. The main types
of film employed in aerial photography (see figures A1 to A3) are:
- panchromatic (black-and-white) film (PAN), one emulsion layer;
- black-and-white infrared film (BWIR), one emulsion layer;
- normal colour film (TC: true colour), three emulsion layers;
- colour infrared film (FC: false colour), three emulsion layers.
Figures A1 to A3 characterize the spectral sensitivity of each type of film with respect to the
wavelength of the incident radiation (0.3-0.9 µm). Note the characteristic differences of the film
types.

A number of simultaneously taken recordings in different spectral bands is called multispectral

photography (MSP). Multispectral photography requires a camera with more than one lens, or
several cameras simultaneously observing the same terrain. Only black-and-white films are to be
employed. In order to make recordings in narrow regions of the EM spectrum (high spectral
resolution) one may combine photographic filters and films of specific sensitivities.

LOG (sensitivity)

2 BWIR - film

1
PAN - film
0
blue green red near-infrared

0.4 0.5 0.6 0.7 0.8 0.9

Wavelength (µm)

Figure A1: Spectral sensitivity of panchromatic (PAN, solid line) and black-and-white infrared
film (BWIR, dotted line) (Source Kodak Publ. M-29, 1982).

Although black-and-white panchromatic film has long been the standard film type for aerial
photography, many applications currently involve the use of colour film. The most important
advantage of using colours is that the human eye allows the distinction of many more colour
gradations than grey tones.
A change in the relative quantity of blue, green, or red light coming from the object changes the
colour we associate with it. In short, we perceive all colours by synthesizing relative amounts of
just three colours. Blue, green, and red are termed additive primaries. White light can be thought
of as the mixture of blue, green, and red light. Various combinations of the three additive
primaries can be used to produce other colours. When red and green light are mixed, yellow light
is produced. Mixture of blue and red light results in the production of magenta light (bluish-red).
Mixing blue and green results in cyan light (bluish-green).

13
 Blue-blocking filter Y
(yellow filter) M
LOG (sensitivity) C

2 base
Y
C backing
1 M
TC-film
0

blue green red

0.4 0.5 0.6 0.7
Wavelength (µm)

Figure A2: Spectral sensitivity of normal (true) colour film (TC). Y = layer sensitive to blue
(produces yellow dye), M = layer sensitive to green (produces magenta dye), C = layer sensitive
to red (produces cyan dye). (Source Kodak Publ. M-29, 1982).

LOG (sensitivity)
2 yellow filter W12
M
Y
FC - film
1
C
0

green red near-infrared

0.5 0.6 0.7 0.8 0.9
Wavelength (µm)

Figure A3: Spectral sensitivity of colour infrared film (FC). Y = layer sensitive to green
(produces yellow dye), M = layer sensitive to red (produces magenta dye), C = layer sensitive
to infrared (produces cyan dye). (Source Kodak Publ. M-29, 1982).

Like the eye, colour television and computer monitors operate on the principle of additive colour
mixing through use of blue, green, and red dots on the picture screen. When viewed at a distance,
the light from the closely spaced screen elements forms a continuous colour image.

Whereas colour television simulates different colours through additive mixture of blue, green,
and red lights, colour photography is based on the principle of subtractive colour mixing using
superimposed yellow, magenta, and cyan pigments (dyes). These three pigment colours are
termed the subtractive primaries, and each results from subtracting one of the additive primaries
from white light.
The subtractive mixture of yellow, magenta, and cyan dyes on a photograph is used to control the
additive mixture of blue, green, and red light reaching the eye of the observer. To accomplish

14
this, colour film is manufactured with three emulsion layers that are sensitive to blue, green, and
red light but contain yellow, magenta, and cyan pigments after processing. We perceive the
colours in the original scene through the subtractive process. Where a blue object was present in
the scene, the image appears blue. Green and red are produced in an analogous fashion. Other
colours are produced in accordance with the proportions of blue, green, and red present in the
original scene.

The assignment of a given pigment colour to a given spectral sensitivity range is a film
manufacturing parameter that can be varied arbitrarily. The colour of the dye developed in any
given emulsion layer need not bear any relationship to the colour of light to which the layer is
sensitive. Any desired portions of the photographic spectrum, including the NIR, can be recorded
on colour film with any colour assignment.

In contrast to “normal” colour film, colour infrared film (also called false colour film) is
manufactured to record green, red, and the photographic portion of the NIR (0.7 – 0.9 µm) scene
energy in its three emulsion layers. The dyes developed in each of these layers are again yellow,
magenta, and cyan. The result is a “false colour” film in which blue images result from objects
reflecting primarily green energy, green images result from objects reflecting primarily red
energy, and red images result from objects reflecting primarily in the NIR portion of the EM
spectrum. The basic structure and spectral sensitivity of colour infrared film are shown in figure
A3. Note that there is some overlap in the sensitivities of the layers.
Because of this colour shift the colours deviate strongly from the usual colour scale. We speak of
“pseudo colour” or “false colour”. Since green vegetation reflects particularly strong in the NIR
portion of the spectrum (see Figure 2.6), vegetation mostly shows up in red tones on a colour
infrared photograph.

15