3 EARTH AS AN ELLIPSOID

A mathematical figure of the earth known as the spheroid or ellipsoid of revolution generated by rotating an ellipse
about its shorter axis is used to derive equations that would calculate the geodetic position of any point at the surface of the
earth. The terrestrial coordinate system shall be used to locate points at the surface of the ellipsoid.

3.1 Elements of an Ellipse

P
Minor Axis

a b
a
F1 є F2
a
Major Axis

P’

Definitions
a = semi- major axis
b = semi-minor axis
F1, F2 = foci
Є = linear eccentricity

Since the earth is modeled as an ellipse and its major axis runs along the equator has a longer radius better known as
semi-major axis (a) than along the polar axis, semi-minor axis(b).

Geometric elements:
a−b
flattening : f = ( 3.1)
a
linear eccentricity :ε =√ a2−b 2( 3.2)
ε
first eccentricity :e= (3.3)
a
ε
second eccentricity : e' = (3.4)
b

The following relations hold among these quantities:

b 1
=1−f = √ 1−e 2= (3.5)
a √1+e ' 2
( 1−e2 ) ( 1+e ' 2) =1(3.6)
 Gde 323 3: EARTH AS AN ELLIPSE page 2

e 2=2 f −f 2(3.7)

The quantities f, e2, e’2 are small. They can be used as arguments in series expansions:

e2 e4 e6 e8 e' 2 3 e' 4 5 e' 6 35 e ' 8

f= + + + +… … .= + + + (3.8)
2 8 16 128 2 8 16 128

For this course we shall be using the parameters of the Geodetic Reference System of 1980.(GRS 80).
Semi-major axis a=6,378,137 m (exact)
Inverse flattening 1/f = 298.257 222 101
e2= 0.00669438
b =6,356,752.314

In the Philippines the official ellipsoid is the Clarke Ellipsoid of 1866 with parameters:
a = 6,378,206.4 m
e2=0.006768658

3.2 Radii of Curvature of the Prime Vertical (N)

A normal section having an azimuth of 90 or 270 at a point defines the prime vertical plane through it. The
normal N is the prime vertical instantaneous radius of curvature.This radius is the perpendicular line to the tangent line
emanating from a point and terminating at the intersection of the minor axis.

n
φ
N

a
N= (3.9)
√ 1−e 2 s ¿ 2 φ
where:
a= semi-major axis
e=first eccentricity
φ=geodetic latitude

At ϕ= 0 (Equator), N=a; at ϕ=90° (Poles), N=a/ √ 1−e 2 =6,399,593.626

The normal n terminating at the major axis:

n=N ( 1−e 2 ) (3.10)

At ϕ=0 (Equator), n=a(1−e 2); at ϕ=90° (Poles), n=a √ 1−e 2=b

3.3 Radius of Curvature of the Meridian ( M )

 Gde 323 3: EARTH AS AN ELLIPSE page 3

Since the meridian section is an ellipse, its radius of curvature is not constant but changes with increasing latitude. The
radius of curvature M at any point in the meridian section is obtained by taking the first and second derivatives of the
equation of the ellipse(3.10) and substituting those expressions in the general equation for radius of curvature given in Eq.
2.10

M3
M2
M1

Meridian section

a ( 1−e 2 )
M= 3 /2
(3.11)
( 1−e2 si n2 φ )

At ϕ=0° (Equator), M =a ( 1−e 2 ); at ϕ=90° , M =a/ √ 1−e 2

3.4 Arcs
Meridian arc length from one latitude to another is obtained by integrating the equation below between selected
limits.
φ2

Sm =∫ Mdφ(3.24)
φ1

φ2
−3 /2
Sm =a ( 1−e )∫ ( 1−e 2 s ¿ 2 φ )
2
dφ (3.25)
φ1

B C

where:
2
Sm =a ( 1−e )
[
A ( φ 2−φ1 ) − ( sin2 φ 2−sin 2 φ1 ) + ( sin 4 φ2 −sin 4 φ1 )
−D
6
2
E
4
( sin 6 φ 2−sin 6 φ1 ) + ( sin 8 φ2−sin 8 φ 1) −⋯
8
(3.26)
]
3 45 175 6 11025 8 43659 10
A=1+ e2 + e 4 + e + e + e =¿.
4 64 256 16384 65536

3 15 525 6 + 2205 8 72765 10

B= e 2 + e 4 + e e + e ……
4 16 512 2048 65536

15 4 105 6 2205 8 10395 10

C= e+ e+ e + e … …(3.27)
64 256 4096 16384
 Gde 323 3: EARTH AS AN ELLIPSE page 4

35 6 315 8 31185 10
D= e + e + e ……
512 2048 131072

315 8 3465 10
E= e + e ……
16384 65536

693 10
F= e ….
131072

This equation may be written in an alternative form by letting

φ1 +φ2
φ m=
2
and
∆ φ=φ 2−φ1
We can write specific values of

sin 2 φ2−sin 2 φ1 =2cos 2φ m sin ∆ φ

sin 4 φ2−sin 4 φ1 =2cos 4 φm sin2 ∆ φ

sin 6 φ 2−sin 6 φ1=2 cos 6 φm sin 3 ∆ φ

and (GRS 80 parameters)

A = 1.00505250 B= 0.00506311 C= 0.00001063 D= 0.00000002

Sm =a ( 1−e2 ) ¿

Exercise:
Find the meridian arc length of latitudes:

φ A=0 °

φ B=1°

Solution
Find φm, Δϕ

Answer: 110,014.83 m

Find the meridian arc length of latitudes:

φ A=1°

φ B=2°
Answer : 110,016.2 m
 Gde 323 3: EARTH AS AN ELLIPSE page 5

Conclusion: Arcs of 1 degree difference is bigger at higher latitudes

A very accurate expression of meridian arc length reckoned from the equator can be found in Helmert (1962):

a
S= [ a φ−a2 sin 2 φ+a4 sin 4 φ−a6 sin 6 φ+ a8 sin 8 φ ] (3.29)
1+n o

f
n=
2−f
1 2 n4
a o=1 … … .+ n … …+ ;
4 64
3 3
a 2= n … …− n3 ;
2 16
15 15
a 4= n2 … …− n4 ;
16 64
35
a 6= n3
48
315 4
a 8= n
512

For less precise measurement of arc to 0.1 meter

S=¿6,367,444.6[ 1.0000007 φ−0.0025188sin 2 φ+ 0.0000026sin 4 φ ]

3.5 Parallel arcs are parallel to the equator that defines latitude. Its length can be determined by:

S p=N cos φ ∆ λ( 3.30)

Where:
N= radius of prime vertical
Δλ”=λ2-λ1(radians)
 Gde 323 3: EARTH AS AN ELLIPSE page 6

r
φ

N
φ

Let Φ
Δλ= 180 = π

Find s
Solve for N =6,378,780.8

S = 19,735,087.4 m

Φ
=
N=a
S = a π= 20,037,508.34

40,075,016

HW 4:
1. Find the parallel arc at ϕ= 10ᵒN from λ=120 E to 121 E
2. Find the parallel arc at ϕ= 11ᵒN from λ=120 E to 121 E
3. Find the parallel arc at ϕ= 12ᵒN from λ=120 E to 121 E

3.6 Areas at the Surface of the Ellipsoid

The area between meridians designated by λ1and λ2 and parallels desigmated by ϕ1 and ϕ2 is
φ 2 λ2

Z=∫ ∫ MN cos φdφdλ(3.31)

φ 1 λ1

Integrating we have
 Gde 323 3: EARTH AS AN ELLIPSE page 7

φ2
( λ 2 − λ1 )
Z=
[ 2
b
2 sin φ
[2 2
1
+ ln
1+ e sin φ
1−e sin φ 2 e 1−e sin φ ]] φ1
(3.32)

Computing the area on the ellipsoid from the equator to latitude ϕ, completely around the ellipsoid

sin φ 1 1+ e sin φ
Z 0−φ=π b 2
[ + ln
1−e sin φ 2 e 1−e sin φ
2 2
(3.32 a)
]
The area half the ellipsoid (hemisphere)

1 1 1+e
Z 0−90 =π b2
°
[ + ln
1−e 2 e 1−e
2
(3.32 b)
]
Equation (3.32 ) may also be written in a series expansion in the form:
φ2
2 3 4
[ 3 [
Z= b ( λ 2−λ1 ) sin ϕ+ e 2 sin3 φ+ e4 sin5 φ+ e6 sin 7 φ+… ..
2
5 7 ]] φ1
(3.33)

Exercise
Let λ1= 110ᵒ; λ2= 112ᵒ; ϕ1= 7ᵒ; ϕ2 = 8ᵒ

Find the area of the quadrant

❑
( 112−110 )∗π
[
Z= ( 6,356,752.314 )

=244,072.227 has
180
2

❑ 10,000 m2 ]
[ sin 8−sin 7 ] =2,440,722,270m 2∗1 ha =2,441 km 2
( )

3.7 Normal Sections

A normal section is a curve formed by the intersection of a plane that contains the normal at a given point to the
surface of the ellipsoid. A normal section from point P1 to point P2 is on formed by the intersection of a plane containing the
normal at point P1 and that passes through point P2, with the surface of the reference ellipsoid.
A normal section can be viewed when a theodolite is leveled with respect to the normal of the ellipsoid at the point
at which the theodolite is set up and sighting on a distant object, we define a plane that contains the normal at the
observation site, and passes through the observed site. The intersection of this plane with the ellipsoid forms the normal
section from the observation point to the observed point.
 Gde 323 3: EARTH AS AN ELLIPSE page 8

Assume that the points (P1, P2, k1) in figure below span a plane which contains the surface normal of
P1, thus, generating the normal section 1. Assume further that the points (P 1,P2, k2) span another plane which
contains the surface normal at P2, thus generating the surface normal section 2. These two planes intersect
the ellipsoid generally into two distinct normal sections. Exceptions are the cases where the points P1 and P2
lie on the same meridian or parallel. Imagine that one were able to set up the theodolites on the ellipsoidal
surface and level it with respect to the ellipsoidal normal. In that hypothetical case the plane described by the
telescope would intersect the ellipsoid in the normal curves 1 and 2 depending on whether the theodolites
would be set up at P1 or P2. Thus, if normal sections were to be used as reference lines in ellipsoidal
computation one would have to decide which of the two possible normal sections should be used.

2 P2
P1 1

k1

k2

Figure 5.1

• when both of the points are located on the same meridian, then the normal section and the
reverse normal section coincide.
• when both of the points are located on the Equator, than both the normal and the reverse
normal section coincides with the Equator.
• when both of the points are located on the same parallel curve (same latitude), then the
normal section lies not on the parallel curve, but on the opposite sides of the parallel.

Normal section P1 to P2 Normal section P2 to P1

 Gde 323 3: EARTH AS AN ELLIPSE page 9

P1

N1
P2
N2

Reciprocal Normal sections

Normal sections P1 and P2

The normal sections from P1 to P2 and P2 to P1 do not coincide because the normals to the ellipsoid at different latitudes
intersect the minor axis at different places.

io n
S ect
al
rm x
No
d ma
io n
ect
al S
Norm
α s e
ver
Δα Re

3.7.1 Separation Between Reciprocal Normal Sections

The normal section and the reverse normal section has an azimuth separation Δ
:
e2 s 2 2 s
∆=
2
sin α AB( )
NA (
cos φ A cos α AB −
2 NA
tan φ A ( 4.1) )
ϕA= 20
s = 100 k
α = 45
 Gde 323 3: EARTH AS AN ELLIPSE page 10

Δ=?

0.00669438 100,000 2 2 100000 3.6E-7

Δ=
2
sin 45∗ (
6380635.807
cos 20 cos 45− )
2∗6380635.807 (
tan 20 =
sin 1}} =0.07 ¿ ¿ )
The maximum linear separation distance between the normal section and the reverse normal section is:

e 2 s3 s
d max =
8N A 2
2
(
sin α AB cos φ A cos α AB −
2NA )
tan φ A =9.04E-3 m=0.009 m=9 mm(4.2)

When we want to measure the angles of a triangle and connect the nodes of the triangle with normal sections, then the
observed angles are not consequent. -> a different ellipsoidal curve should be used for the representation

Considering numerical examples

αAB =45°; ϕ= 0°
s 50 km 100 km 200 km
Δ” 0.021” 0.085” 0.339”
dmax 0.001 m 0.010 m 0.082 m
αAB =45°; ϕ= 30°
Δ” 0.016” 0.063” 0.251”
dmax 0.001 m 0.008 m 0.061 m

Conclusion:
 Azimuth separation is smaller at higher latitudes with same arc length. Gets bigger at longer arcs.
 Linear separation is smaller at higher latitudes with same arc length. Gets bigger at longer arcs.

3.7.2 Elliptic Arc of a Normal Section

 Gde 323 3: EARTH AS AN ELLIPSE page 11

s B

A
NB

NA

1 1
(
s=N A σ 1+ σ 2 η2A cos2 α AB ( η2A cos2 α AB−1 ) + η2A t A cos α AB ( 1−3 η2A cos2 α AB) σ 3 +… .
6 8 )
2
s 1 s 1 s 3
σ=
NA (6 (
1+ η2A cos 2 α AB 1−η2A cos 2 α AB
NA ( )) 8 ( )
− η2A t A cos α AB ( 1−3 η2A cos2 α AB )
NA
+…
)
η2A =e ' 2 cos2 φ A
t A=tan φ A

To solve for s and σ,

1. initialize the value of σ and compute for s
2. substitute the value of s to compute σ
3. iterate until s and σ does not change any more or convergence.

3.7.3 Radius of Curvature of Normal Section

Given the azimuth of a line for which we are interested in the curvature, the normal section radius of curvature at a point on
the ellipsoid in any azimuth is given by Euler’s theorem

MN
Rα = (3.12)
N cos α + M sin2 α
2

where:
M, N=radius of curvature of the meridian and prime vertical respectively at a certain point
α = geodetic azimuth of the normal section at a point

If  = 0 , R =M; for  = 90, R =N

Exercise. Show proof.

 Gde 323 3: EARTH AS AN ELLIPSE page 12

3.8 The Geodesic

A geodesic is a curve which gives the shortest distance, on a surface, between any two points. If the surface is a plane, the
geodesic is a straight line; if the surface is a sphere, the geodesic is an arc of a great circle. On the ellipsoid is a curve
having a double curvature and is thus not a plane curve. . In most cases the geodesic is found to lie between the two plane
curves and has a reversed curvature. Also the angle by which the azimuth of the geodetic differs from the azimuth of the
plane section is much smaller than can be measured.

We consider the construction of the geodesic on the surface of the ellipsoid. We first level our theodolite with
respect to point A and then aim at a distant point B defining the normal section curve AaB. We then go to B, level the
theodolite , point at A to define the normal section BbA, we then turn the theodolite by 180 and define a new point C and the
normal section BbC. We repeat the operation by going to point C, point D and subsequent points

The separation of the normal section is small and becomes smaller as the separation between the
points decreases. If we let the distance AB, BC, CD, etc become smaller and smaller, a unique curve will be
obtained between the points. This curve is the geodesic.

We could construct the geodesic between two points if we knew the appropriate azimuth of a starting
segment

2
Normal section 2-1

α3
Geodesic

α2
Meridian

Normal section 1-2

α1

1 Parallel

Figure 5.3

3.8.1 Azimuth Reduction

The angle by which the azimuth of the geodetic line differs from the azimuth of the normal section is much
smaller than can be measured. These quantities can be obtained only by computation:
 Gde 323 3: EARTH AS AN ELLIPSE page 13

η2 s 2 η2 t s 3
( )
α 1−α 2=
6 N1 ( )
sin α 1 cos α 1−
24 N 1
sin α 1( 4.3)
2 2 2 3
η s η t s
α =α − ( ) sin α cos α +
24 ( N )
2 1 1 sin α
1 1
6 N 1 1

where:
α1 azimuth of the normal section, at 1, from 1 to 2
α2 azimuth of the geodesic, at 1, from 1 to 2
s = geodetic distance
N= radius of the prime vertical
η2=e ' 2 cos 2 φ A
t=tan φ
To find the value of e ' 2 let
√ 1−e 2= 1 ' 2
√ 1+ e
1
√ 1+e ' 2=
√ 1−e 2
1
e ' 2= −1
1−e 2

e ' 2=0.00681478(GRS 80' )

If we consider only the first term of the equation we can compare it with which gives the azimuth
separation of the normal sections. We conclude that:
1 1
( α 1−α 2 ) ≈ 3 ( α 1−α 3 ) ≈ 3 ∆ (4.4)
1
α 2=α 1− ∆
3
where:
α3 azimuth of the normal section, at A, from B to A

Exercise:
Let ϕA= 9ᵒ αAB= 60ᵒ s = 150 km
Find the azimuth of the geodesic.

Solution:
Solve NA = 6,378,659.5 m
Δ= 7.8 e—7 radians /sin 1” =0.16 “
α2= 60 – 1/3Δ = 59-59-59.95

Exercise
Let ϕA= 40ᵒ αAB= 135ᵒ22’35.4” s = 200 km
Find the
1. azimuth of the geodesic.
 Gde 323 3: EARTH AS AN ELLIPSE page 14

2. Rα
3. Linear separation

e2 200,000 2 2 200,000
∆=
2
sin135 ᵒ 22’ 35.4 ” (
NA ) (
cos 40 cos 135 ᵒ 22 ’ 35.4 ”−
2NA
tan 40 )
NA= 6,386,976.2 m
Δ =-9.81 e -7 = -0.2”/3 = -0.07”
α= 135 ᵒ 22 ’ 35.47 ”

η2 s 2 η2 t s 3
α 2=α 1− ( )
6 N1
sin α 1 cos α 1+
24 N 1 ( )
sin α 1

2 2
η =0.00681478 ( cos 40 ) =0.00399908
α 2=135 ᵒ 22 ’ 35.47 ”
Which says that the geodesic approximately trisects the angle between the reciprocal normal sections,
lying closer to the direct normal section at the given point.

.15 ¿
3.9 Latitudes

z
inscribed circle

P’ (x’, z’)
x
P
tangent line
N
z
n
90º +
x
o D

QUADRANT OF AN ELLIPSE

H
 Gde 323 3: EARTH AS AN ELLIPSE page 15

The geodetic latitude (φ) is the angle measured in the meridian plane between the equatorial plane of the ellipsoid
and the surface normal at point P.

Consider a sphere whose radius is equal to the semi major axis of the ellipsoid. The center of both figures is O.
The parametric or reduced latitude (β) is obtained by projecting the ellipse on the concentric circle having a radius a. The
point P’ is located along the equatorial normal through P
b
tan β= tan φ=√ 1−e 2 tan φ(3.16)
a

The geocentric latitude (Ψ) is the angle between the equatorial plane and a line from the ellipse center to surface
point P:
b 2
tanψ =( 1−e 2 ) tan φ= ()a
tan φ(3.17)

n=( 1−e2 ) N
Exercise
Let φ=34 ᵒ
β= 33-54-39
Ψ = 33-49-18
n = 6,342,080.8
N = 6,384,823.2

3.10 Determining Ellipsoid Dimensions from Two Arcs

The simplest method of determining the a, b and/or e of the ellipsoid is by measuring two meridian arcs. The length
of each arc and the latitudes of the terminal points of each must be measured. If the earth was perfect ellipsoid, the two arcs
would determine exactly the elements of the ellipsoid.

For the computations in a latitude arc measurement, the angles

∆ φ=φ 2−φ1
'
¿ ∆ φ =φ ' 2 −φ ' 1

are formed from the observed geographic latitudes 1, 2, '1 '2. The corresponding meridian arcs

∆ G=s∧∆G ' =s '

are from triangulation networks.The larger the latitude interval Δφ, the more accurate is the computed flattening; whereas,
the accuracy of a depends in particular on the lengths of the meridian arcs.

Lengths of the two meridian arcs are s and s’, and that the differences of the latitudes of their terminals are 
and  respectively.
 Gde 323 3: EARTH AS AN ELLIPSE page 16

s’

M’
b
s
M
φ’2φ’1 φ2 φ1
a
Δφ’ Δφ

a ( 1−e 2 )
M= 3 /2
(3.18 a)
( 1−e2 si n2 φ )
' a ( 1−e2 )
¿M = 3 /2
( 3.18b)
( 1−e 2 si n2 φ ' )
where
1
φ= ( φ1 +φ 2 )
2
1
φ '= ( φ ' 1 + φ' 2 )
2
If the two arcs are regarded as arcs of circles whose radii are to be found, then
s
R= (3.19 a)
∆φ
s'
¿ R' = '
(, 3.19 b)
∆φ

The shorter the arcs, the less the error involved in assuming that they are circular.

Equating the two values of M = R and R’ = M’, we have

s a ( 1−e2 )
(a ) = ( 3.20 a)
∆ φ ( 1−e2 s ¿2 φ )3 /2

s' a ( 1−e2 )
(b ) = (3.20 b)
∆ φ ' ( 1−e2 s ¿2 φ' )3 /2
Equating both equations by a(1-e2)
 Gde 323 3: EARTH AS AN ELLIPSE page 17

3 /2 3/ 2
s ( 1−e 2 s ¿2 φ ) s ' ( 1−e 2 s ¿2 φ ' )
= (3.21)
∆φ ∆ φ'
2/ 3
s ∆ φ'
[ s' ∆ φ ] ( 1−e2 s ¿2 φ )=1−e 2 s ¿2 φ ' (3.22)

2/ 3 2/ 3
s ∆ φ' s ∆ φ'
[ s' ∆ φ ] [ ] − '
s ∆φ
2 2
e s ¿ φ=1−e s ¿ φ '
2 2

2/ 3 2/ 3
s ∆ φ' s ∆ φ'
2
e s¿ φ'− '
2

s ∆φ [ ] 2 /3
2
e s ¿ φ=1−
2
s' ∆ φ [ ]
2/ 3
s ∆ φ'
2
(
e s ¿ φ '− '
s ∆φ
2
[ ] s ¿2 φ =1−
2/ 3
) [ ] s ∆ φ'
s' ∆ φ
s∆φ'
e 2=
1− [ ]
s' ∆φ
2 /3
(3.23)
s∆φ'
s ¿ φ '−2
s' ∆φ [ ] si n φ 2

Flattening can be obtained using (3.5)

f =1− √ 1−e2

The semi-major axis a can be obtained using (3.20 a) or (3.20 b).The value of b may then be found by (3.5)

3.11 Radii of spherical approximation to the Earth

In some applications it is convenient to let the earth be a sphere rather than an ellipsoid. A suitable radius may be defined
that approximates the Earth as a sphere.

3.11.1 Gaussian mean radius R

Gaussian mean radius is defined to be the integral mean value of R taken over the azimuth varying from 0 to 360.
2π
1 MN
R= ∫ dα
2 π 0 N cos α + M sin 2 α
2

Which upon integration yields

a √ 1−e 2
mean R α =√ MN = (3.13)
1−e2 sin2 φ

Values of R will always be greater or equal to M and smaller or equal to N.

3.11.2 Radius of Sphere having the mean of the three semi-axes of the ellipsoid

Let
 Gde 323 3: EARTH AS AN ELLIPSE page 18

1−e2
=a + √
a+a+b 2
Rm =
3 3 3( )
Expanding and using GRS 80 values

e2 e4
Rm =a 1− ( )
− −… .. =6,371,008.8 m
6 24

3.11.2 Radius of Sphere having the same area as the Ellipsoid.

We set the area of the sphere equal to the area of the ellipsoid
2 3
(
4 π R A 2=4 π b 2 1+ e2 + e4 + … .
3 5 )
2 3
(
R A 2=a2 ( 1−e2 ) 1+ e2 + e 4 +… .
3 5 )
1 17 4
(
R A =a 1− e2−
6 360 )
e ∓ … … .. =6,371,007.2 m( 3.14)

3.11.3 Radius of Sphere having the same volume as the Ellipsoid

The volume of the sphere is expressed as:

4
V s = π R3v.
3
The volume of the ellipsoid is expressed as :
4
V e = π a2 b
3

Equating both equations:

3 1/ 6
R v =√ a2 b=a ( 1−e 2 )
Expsnding and substituting GRS 80 values of a and e

1 5
(
RV =a 1− e 2− e 4−… … . =6,371,000.8 m
6 72 )
For most applications the average R may be taken as 6,371 km.