Uniform distributions and solid angle dependence in random

pulsar geometry generation

Hayden Tornabene, Yuping Huang, and Joel Weisberg
April 6, 2016
We first derive an expression that yields a uniform distribution of magnetic axes on the
surface of a sphere, specifically for between 0 and . We note that is the angle between the
rotation axis and the magnetic axis as seen in Figure 1. The problem we consider only requires
one dimension (quantified by ), but we include the azimuthal (longitude) angle for the sake
of completeness. Two alternate techniques are given for generating the uniform magnetic axis
distribution. Next, in order to fully simulate the geometry, the observer is brought into the
picture by deriving the distribution of the angle between the spin pole and the observers line
of sight. Finally, we address a question from Simon Johnston considering the observers role.

Figure 1: Geometry of a pulsar where is the angle between the rotational and magnetic axes. Figure
reference: http://inspirehep.net/record/1095127/plots .

First Approach to Uniformly Distributed m - Integration and

CDFs

There are two separate ways to consider the problem of a uniform distribution of magnetic axes m, as
show in Figure 2. The first solution considers the integration of the cumulative distribution function. We
begin with a brief explanation of cumulative distribution function, or cdf. A cdf describes the probability

that some real-valued variable X with some probability distribution will be found to have a value less than
or equal to x. That is,
FX (x) = P (X x).

(1)

The probability that X lies within the interval (a,b] is given,

P (a < X b) = FX (b) FX (a),

(2)

where is the angle from the north rotational pole to the magnetic axis (or the polar angle) and is the
azimuthal angle in the xy-plane from the x-axis with 0 < 2, and is the solid angle. The probability
that a point lies in an infinitesimal cone of solid angle is given as,
P ()d,

(3)

where the normalization is,

Z Z
P ()d = 1.

(4)

The probability density function (pdf) for uniform density over a unit sphere is given as,
1
.
(5)
4
As and are chosen independently, we can write the pdf, f() as the product of the two probability density
functions that describe and . As solid angle is a function of and , for d = sin d d, the joint
probability density function for and is given,
f () =

sin
.
(6)
4
Thus the two marginal distributions, that is the probability density functions for each variable with the other
variable integrated out, are given as,
f (, ) =

Z
f () =

f (, )d =
0

f () =

f (, )d =
0

sin
;
2

(7)

1
2

(8)

As the cdf and the pdf are related by the following,

F 0 (x) = f (x),

(9)

and from the definition in Equation 2 and the fundamental theorem of calculus, for a uniform generated set
of and , the cdfs for F() and F() are defined as,

1 cos
sin
d =
;
2
2
0
0
Z
Z
1

F () =
f ()d =
d =
.
2
0
0 2
Z

F () =

f ()d =

(10)

(11)

Now that we have two cumulative distribution functions that describe and , we consider two random
variables, v and u, that are uniform over the interval [0,1]. These variables represent the random numbers
created by a uniform random number generator. In other words, we can define the random numbers in terms
of the cdf where, v = F() and u = F(). Solve for and and we get the following inverse cdfs:

v = F () =

1 cos
;
2

(12)

= F1 (v) = cos1 (2v 1).

(13)

= F1 (u) = 2u

(14)

And similarly for ,

An incorrectly distributed array of points around a sphere concentrates values at the poles as shown in
Figure 2 whereas a correct distribution is truly uniform with the consideration of the inverse cdfs, shown in
Figure 3.

Figure 2:
Incorrect two-dimensional ( and ) uniform distribution.
points are centrally concentrated when the incorrect distribution is used.
http://www.bogotobogo.com/Algorithms/uniform distribution sphere.php .

Note that the

Figure reference:

Second Approach to Uniformly Distributed m - Differentials

One may alternatively consider the problem in terms of differentials. This formulation comes directly
from Yuping Huang, and we include it here for the sake of completeness. We consider some vector m pointing
in space with a tilt with respect to the z-axis of an orthogonal (x, y, z) coordinate system. We can consider
m as the magnetic vector and z as the spin axis. Thus the angle is the angle formed between m and z
and is the longitudinal angle. For , the solid angle at m where, as in the first approach,
d = sin d d.

(15)

We again denote f () has the probability density function (pdf) of . As is to be uniformly distributed
around a sphere, then
f () d = C.

(16)

This equation says that if we take some infinitesimal patch of solid angle d, no matter where one picks
this patch, the probability the vector points in that direction is the same, namely the constant C, where
C = 1/(4).

Figure 3: Correct two-dimensional ( and ) uniform distribution. Figure reference: Same as Fig. 2.
As and are chosen independently, again, the pdf f () may be written as the product of pdfs describing
f () and f (). From Eq. 16, we then have,
f ()f () sin d d = C.

(17)

In order for the product of these two PDFs to be equal to some constant C, we notice that both f ()d and
f () sin d will also have to be constant. For A and B constant,
f ()d = A,

(18)

f () sin d = B.

(19)

and

We want to find a function f () (particularly a pdf) that satisfies the above constant equation. (Note that
the same can be done for .) One may propose that f () = B sin satisfies the equation, and indeed it
does. This means that the probability of being at a certain angle is proportional the the value of sine at
that angle. (This is not saying sin is uniformly distributed, as this would require an equation with pdf,
f (sin )d(sin ) = D). Instead, we manipulate Equation 19 and find,
f () d( cos ) = B.

(20)

So we have found that ( cos ) is uniformly distributed, given the following consideration:
If we consider a distribution f ( cos ) instead of f (), this pdf, along with would generate a uniform
distribution where again, [0,], can be completely determined by and -(cos ). Equation 15 becomes,
d = d( cos )d,

(21)

f ()f ( cos ) d( cos )d = C.

(22)

and thus Equation 17 becomes,

The separation of variables again gives,

f ( cos ) d( cos ) = B.

(23)

Hence, (-cos ) is uniformly distributed on [-1,1], and since it is an odd function, cos is similarly uniformly
distributed.
Notice that the above two approaches yield identical results. However, the first approach provides a
explicit function for (and ) that may be used in some randomized statistical routine aimed at probing
possible pulsar geometry (for v and u some random value of and respectively). The second approach
gives a more geometric argument, useful for conceptualizing the uniform distribution in terms of probability
density functions.

Additional issue - The role of the observer via the spin-pole to

line-of-sight angle

The above two sections dealt with alternate, equivalent techniques for uniformly distributing a set of
magnetic axes on the face of a sphere with a fixed spin axis; characterizing the location of each magnetic
axis with the two angles (, ). However, in order to fully simulate an observers view, the observer must
be brought into the analysis! Specifically, we wish to randomly align the spin axes of a set of spheres with
respect to us. We do so via , the angle between the north spin-pole and the observers line of sight. To
visualize this, take a globe (with a flashlight representing the beam and magnetic axis) and rotate the spin
axis orientation randomly in space to simulate a set of spin axes oriented randomly with respect to the
observer. Note how the line of sight changes with changing spin axis. We note that a uniform distribution
of simulates a random set of pulsar spin axis orientations with respect to the observer. It is crucial to
note that this procedure is similar, but not the same as randomly placing magnetic axes on the surface the
pulsar, as discussed above in Sections 1 and 2).
It may seem intuitive to find the distribution of the parameter rather than of , where is the angle
between the magnetic axis and the line of sight. (The parameter is shown in Figure 1.) However, while
is distributed uniformly, has a much more complicated distribution, since
= ,

(24)

and and each have their own distributions. (Of course a value of can easily be derived from
Equation 24, once a particular pair of and has been selected.) It can be shown that the cos1 factor
that we discovered in deriving the distribution in Sections 1 and 2 may be applied to in the same way as
we are randomly distributing it across a concentric shell just outside the pulsar. The geometry argument is
identical where the vector represents the radius of a sphere we are trying to randomly distribute. Instead
of a rigorous derivation of the distribution, let us consider another reiteration of the following plausibility
argument: Imagine a sphere (the pulsar) with another shell just outside the pulsar surface. Now imagine
that the line of sight is fixed on the inner sphere. We may uniformly distribute the spin axis around the shell
just outside the pulsar in the same way that we randomly distribute the magnetic axis. This is identical to
a fixed line of sight on the outer shell while the pulsar sphere below it randomly orients itself.

3.1

Aligned vs. Orthogonal Rotators

In an email on September 29, 2015 to Joel Weisberg and Hayden Tornabene, Simon Johnston articulated
the following question when considering randomized pulsar geometries:
There is I suppose the further question of the illumination of the sky which also depends on alpha and which
goes to the question of detectability but not sure what you want to do about that.
The extra issue to my thinking comes from the fact that an aligned rotator only illuminates a tiny fraction
of the celestial sphere whereas an orthogonal rotator illuminates a large fraction. Therefore (all other things
being equal which they may not be) you should detect more orthogonal rotators than aligned rotators.
This also comes down to solid angle so I presume theres some sort of cosine dependence there too.

This problem speaks to the the angle between the spin axis and the magnetic axis, . When = 0, we have
an aligned rotator, or a pulsar where the magnetic and spin axes are the same vector. As the pulsar spins,
the area swept out by the beam is at its smallest for = 0. When = /2, we have what is called an
orthogonal rotator. As increases from 0 to /2, the area swept out by the beam becomes larger, implying
that, if all else is the same, we should see more orthogonal rotators than aligned rotators. Simply consider
the ratio of the areas swept out by the beam.
It seems to us that if both and are randomized uniformly and correctly, then this problem is addressed
in the statistical routine. The above question is simply saying that the range of possible for a detected
aligned rotator is much smaller than the range of possible for an orthogonal rotator. In other words, the
aligned rotator must be pointed right at the earth (one possible ) to ensure detection of the pulsar, whereas
the orthogonal rotator has a much larger range of possible values that ensure detection. Think back to
the shell around the sphere of the pulsar in Section 3, where the line of sight is being randomized. It would
seem that no bias should be made on , that is to say no additional scaling function (like a cos1 ) is needed
given that we are implicitly considering whether or not we see the pulsar when we uniformly distribute .
Any scaling function beyond the cos1 , which ensures a uniform distribution of , would incorrectly bias the
routine designed to produce n random pulsar geometries.

Bibliography
[1] Thanks to K. Hong and his blog post on bogotobogo.com: http://www.bogotobogo.com/Algorithms/
uniform distribution sphere.php . Figures 2 and 3 are taken from this reference.
[2] Fiigure 1 image credit: http://inspirehep.net/record/1095127/plots