This action might not be possible to undo. Are you sure you want to continue?
, 55 (2011), 501÷514 501
© 2011 Inst. Geophys. AS CR, Prague
On the gradient of curvature of the plumblines
of the Earth’s normal gravity field and its isocurvature lines
GERASSIMOS MANOUSSAKIS AND DEMITRIS DELIKARAOGLOU
Department of Surveying Engineering, National Technical University of Athens, Iroon
Polytechniou 9, Zografos 15780, Athens, Greece (gmanous@survey.ntua.gr, ddeli@mail.ntua.gr)
Received: September 30, 2009; Revised: June 15, 2010; Accepted: February 25, 2011
ABSTRACT
This paper presents an approach to determine the gradient of curvature of the normal
plumblines at a point P above the ellipsoid and introduces a new geometrical object
which is the isocurvature line. The assumed facts are the coordinates of the point P and
the formula for the normal gravity potential U. For the determination of the gradient of
the normal plumbline curvature k at the point P we define a small circle on the meridian
plane of P whose center is at the point P. The circle has the radius of one meter and
interior D. In this circle we construct a curvature replacement function to approximate
the curvature function k. This replacement function is a quotient of polynomials hence it is
easy to find its partial derivatives at the point P. For the construction of replacement
function we make the assumption that in the interior of the circle D the first order partial
derivatives of U behave linearly and the second order partial derivatives have constant
values which equal their value at the point P. Then we set the gradient of the curvature
function to be equal with the gradient of the aforementioned replacement function at P.
An isocurvature line of the normal gravity field passing through a point P is a curve such
that the value of the function of the plumblines’ curvature k is constant and equals k(P).
We give a formula to find the direction of the isocurvature line on the meridian plane and
we prove that there are infinitely many isocurvature lines passing through the point P and
they all lie on a special surface, the isocurvature surface.
Ke ywor ds : plumblines, curvature, normal gravity field, isocurvature lines,
isocurvature surface
1. INTRODUCTION
In Geodesy the term “isocurvature line” defines a curve along which the functional
value of the plumblines’ curvature is constant. The term “isocurvature” is commonly used
in Physical Cosmology, frequently as “isocurvature fluctuation”, or as “isocurvature
perturbation” e.g., Moroi and Takahashi (2004), Abramo and Finelli (2001). Some
examples can be found in the study of models with an inhomogeneous primordial baryon
distribution. The previously mentioned models were  and still are used for the description
of the development of the early universe and the effects of the cosmic microwave
background. We will start our study of the isocurvature lines by determining those two
G. Manoussakis and D. Delikaraoglou
502 Stud. Geophys. Geod., 55 (2011)
which are on a parallel plane (which contains P) to the equatorial plane and the meridian
plane of the point P accordingly. To find the direction of the isocurvature lines on the
meridian plane we need the partial derivatives of the curvature function k of the
plumblines, i.e. we need to know the gradient of the function k. To determine grad k at
a specific point P in Cartesian coordinates it is necessary to determine the partial
derivatives of the normal potential U up to degree three. For example if (X, Y, Z) are the
coordinates of a global Cartesian equatorial system whose origin is situated at the Earth’s
center of mass, the Zaxis being the Earth’s mean axis of rotation, the Xaxis is the
intersection of the equator’s plane and the meridian plane of Greenwich and the Yaxis
makes the system righthanded. We can evaluate the first order partial derivatives at point
P using the following expression from Heiskanen and Moritz (1967):
2
sin
cos cos sin cos
cos
cos
cos sin sin sin
cos
0
sin cos 0
u f f
U U
f f f
X u
U u f f U
Y f f f
U
f u
Z
f f
 
 
 
ì
 ì  ì

ì
 ì  ì
 
 
(
÷ ÷
( c c (
(
(
(
(
c c
(
(
(
(
c ÷ c (
(
=
(
(
(
c c
(
(
(
(
( c
(
(
(
(
c
¸ ¸
¸ ¸
(
¸ ¸
, (1)
where
( )
( ) ( )
2 2 2 2 2 2
0
2 2 2 2 2
1 1 1
, arctan sin cos ,
2 3 2
, , sin .
GM E q
U u a f
E u q
f f u u E f f u u E
 
 e  e 
 
 
= + ÷ +

\ .
÷ = + ÷ = +
(2)
Another formula can be found in Grafarend and Ardalan (1999) and Ardalan and
Grafarend (2001). The second order partial derivatives, at point P, are determined by the
following expression:
2 2 2
2 3 3
2
2
1
2
cos cos cos sin
cos cos cos sin sin
0
, ,
sin cos sin sin cos
0
cos cos cos sin
U U E U E
X Y
u f f
U U U U
f f u
X Y Z
u
U U u U u U
u X f Y f Z
U U
f f
X Y
 ì  ì
 ì  ì 


 ì  ì 

 ì  ì
(
c c c
÷ ÷
(
c c
c
(
(
c c c c
(
+ + +
(
c c c
c
(
(
=
(
c c c c
(
÷ + ÷
(
c c c c c
(
(
(
c c
(
+
c c (
¸ ¸
M ( )
2
2
2
2
2
2
2
2
2
U
X
U
Y
U
Z
U
X Y
U
X Z
U
Y Z
ì
(
c
(
c
(
(
c
(
(
c
(
(
c
(
c (
(
c
(
(
c c
(
(
c
(
c c
(
(
c
(
( c c
¸ ¸
, (3)
Curvature of the plumblines of the Earth’s normal gravity field
Stud. Geophys. Geod., 55 (2011) 503
where ( )
1
, , u  ì M is 6 × 6 matrix (see Appendix A). To find the third order partial
derivatives of the normal potential we need an additional ten dimensional transformation
of even more complicated form. To avoid additional derivations of the higher order
partials of the function U we shall formulate an alternative way to determine ( ) grad k P
without using the third order partial derivatives of the normal potential U.
2. DETERMINATION OF THE GRADIENT
OF THE CURVATURE FUNCTION
Suppose that P is a point above the ellipsoid with geodetic coordinates ( ) , ,
P P P
h ¢ ì
and we want to determine the value of ( ) grad k P without using the third order partial
derivatives of U. As a start we introduce the following transformations:
2
2
cos cos sin sin cos cos cos
cos sin cos sin sin cos sin
0 cos sin
sin
P
Q
X N x
Y N y
Z h
b
N
a
 ì ì  ì  ì
 ì ì  ì  ì
 

(
(
÷ ÷
( ( (
(
( ( (
(
= + ÷
( ( (
(
( ( (
(
¸ ¸ ¸ ¸ ¸ ¸
(
¸ ¸
, (4)
where N is the principal radius of curvature in the prime vertical at the projection (along
the vertical) of point P on the ellipsoid, and
1
1
1
0 1 0 0
0 0 1 0
0 0 1
P
x x
y y
h h h
( ( ( (
( ( ( (
= +
( ( ( (
( ( ( ( ÷
¸ ¸ ¸ ¸ ¸ ¸ ¸ ¸
, (5)
where the origin of ( ) , , x y h system is on the ellipsoid, as shown in Fig. 1.
The first transformation described by Eq.(4) defines a local Cartesian system ( ) , , x y h
such that the xaxis is tangent to the local parallel (positive to east), the yaxis is tangent to
the local meridian (positive to north), and the haxis is the vertical normal line to the
ellipsoid passing through P. The second transformation, from Eq.(5), defines a local
Cartesian system ( )
1 1 1
, , x y h which is a parallel transport of the previous one and its
center is at the point P. Let D be the interior of a circle on the meridian plane of P which
has its center at the point P and radius ε = 1 m. The equation of this circle is
2 2 2
1 1
y h c + = . (6)
If
( ) ( ) ( ) ( )
1 1 1
0, , a t y t h t = ,  
0
0, t t e (7)
G. Manoussakis and D. Delikaraoglou
504 Stud. Geophys. Geod., 55 (2011)
is the vector equation for the segment of a plumbline passing through a known point of the
set D then for the tangent vector and the prime vertical of this segment of the plumbline
holds
1
1
dy U
dt y
c
=
c
,
1
1
dh U
dt h
c
=
c
(8)
and
2 2 2
1
2 2
1 1 1 1
1
d y U U U U
y y h h
dt y
c c c c
= +
c c c c
c
,
2 2 2
1
2 2
1 1 1 1
1
d h U U U U
y h y h
dt h
c c c c
= +
c c c c
c
. (9)
The curvature of the plumbline k is given as
2
2
3
da d a
dt
dt
k
da
dt
×
= , (10)
whereas alternative expressions for it can be found, for example, in Grafarend (1997).
From Eqs.(8) and (9) it is easily seen that it is possible to determine the value of the
plumbline curvature at a specific point without knowing the vector equation ( )
1
a t . In
local coordinates ( )
1 1 1
, , x y h the point P has coordinates ( ) 0, 0, 0 . On D we define the real
functions
1a
y' ,
1a
h' ,
1a
y'' ,
1a
h'' and
1a
k by considering Taylor expansions and neglecting
higher order terms:
Fig. 1. The local Cartesian systems used for the definition of the gradient of curvature function.
Curvature of the plumblines of the Earth’s normal gravity field
Stud. Geophys. Geod., 55 (2011) 505
( )
( )
( ) ( )
2 2
1 1 1 1 1 1
2
1 1 1
1 0,0,0
0,0,0 0,0,0
, ,
a
U U U
y x y h y h
y y h
y
c c c
' = + +
c c c
c
, (11)
( )
( )
( ) ( )
2 2
1 1 1 1 1 1
2
1 1 1
1 0,0,0
0,0,0 0,0,0
, ,
a
U U U
h x y h y h
h y h
h
c c c
' = + +
c c c
c
, (12)
( )
( )
( )
( )
( )
( ) ( )
( ) ( ) ( ) ( )
2 2
1 1 1 1
2
1 1 1 1
1 0,0,0 0,0,0
0,0,0 0,0,0
2 2
2 2
1
2
1 1
1
0,0,0 0,0,0
2 2 2 2
1
2 2
1 1 1 1
1 1
0,0,0 0,0,0 0,0,0 0,0,0
, ,
,
a
U U U U
y x y h
y y h h
y
U U
y
y h
y
U U U U
h
y h y h
y h
 
c c c c

'' = +

c c c c
c 
\ .
(
   
(
c c
 
+ +
(
 
c c
c  
(
\ . \ .
¸ ¸
 
c c c c

+ +

c c c c
c c 
\ .
(13)
( )
( )
( )
( )
( )
( ) ( ) ( ) ( )
( ) ( )
2 2
1 1 1 1
2
1 1 1 1
1 0,0,0 0,0,0
0,0,0 0,0,0
2 2 2 2
1
2 2
1 1 1 1
1 1
0,0,0 0,0,0 0,0,0 0,0,0
2 2
2 2
1
2
1 1
1
0,0,0 0,0,0
, ,
.
a
U U U U
h x y h
y h y h
h
U U U U
y
y h y h
y h
U U
h
y h
h
 
c c c c

'' = +

c c c c
c

\ .
(
c c c c
(
+ +
(
c c c c
c c
(
¸ ¸
(
   
(
c c
 
+ +
(
 
c c
c  
(
\ . \ .
¸ ¸
(14)
It should be clarified that
1a
y' ,
1a
h' ,
1a
y'' , and
1a
h'' here represent real functions of
three variables and the primes do not denote derivation. Eqs.(11) and (12) imply that the
first derivatives
1
U y c c and
1
U h c c change linearly on D. Hence, the last two terms in
Eqs.(13) and (14), are easily derived by substituting Eqs.(11) and (12) into the
corresponding Eq.(9), and manipulating the resultant expressions for the right hand side of
Eq.(9). We define the following function
( )
( ) ( ) ( ) ( )
( ) ( ) ( ) ( )
1 1 1 1 1 1 1 1 1 1 1 1
1 1 1
3 2
2 2
1 1 1 1 1 1
, , " , " ,
,
, ,
a a a a
a
a a
y y h h y h y y h h y h
k y h
y y h h y h
' '' '' ' ÷
=
(
' ' +
(
¸ ¸
, (15)
which after some manipulations becomes
( )
( )
2 2
1 1 2 1 3 7 1 8 1 1 9 1
1 1 1
3/ 2
2 2
4 5 1 6 1 10 1 11 1 1 12 1
,
1 2
a
c y c h c c y c y h c h
k y h
c c y c h c y c y h c h
+ + + + +
=
(
+ + + + +
(
¸ ¸
, (16)
G. Manoussakis and D. Delikaraoglou
506 Stud. Geophys. Geod., 55 (2011)
where c
i
(i = 1, 2, …, 12) are real numbers. The full expressions for computing the
coefficients c
i
are given in Appendix A. With Eq.(16) it is now possible to determine the
curvature of a plumbline passing through an arbitrary point B in the set D at the specific
point. While constructing Eq.(16) we made the assumption that the change of the values
of all second order partial derivatives of the normal potential in the set D is negligible i.e.
their values are approximately equal to their value at the point P. This statement holds
within an accuracy adopted in the paper. Since Eq.(16) is a differentiable function in the
set D it is possible to determine its partial derivatives, hence (assuming that the
polynomial in the absolute value is positive at the point P) we have that
( )
( )
( )
( )
1 1 7 1 8 1
3/ 2
2 2 1
4 5 1 6 1 10 1 11 1 12 1 1
2 2
1 1 2 1 3 7 1 8 1 1 9 1 5 10 1 12 1
5/ 2
2 22
4 5 1 6 1 10 1 11 1 12 1 1
2
1 2
3 2
,
1 2
a
k c c y c h
y
c c y c h c y c h c y h
c y c h c c y c y h c h c c y c h
c c y c h c y c h c y h
c + +
=
c
(
+ + + + +
(
¸ ¸
+ + + + + + +
÷
(
+ + + + +
(
¸ ¸
(17)
( )
( )
( )
( )
1 2 8 1 9 1
3/ 2
2 2 1
4 5 1 6 1 10 1 11 1 12 1 1
2 2
1 1 2 1 3 7 1 8 1 1 9 1 6 11 1 12 1
5/ 2
2 2
4 5 1 6 1 10 1 11 1 12 1 1
2
1 2
3 2
,
1 2
a
k c c y c h
h
c c y c h c y c h c y h
c y c h c c y c y h c h c c h c y
c c y c h c y c h c y h
c + +
=
c
(
+ + + + +
(
¸ ¸
+ + + + + + +
÷
(
+ + + + +
(
¸ ¸
(18)
1 1
0
a
k x c c = . (19)
Subsequently, instead of differentiating the function k, we differentiate the function
1a
k which is not a complicated function. On D within an adopted accuracy we set
1 1 1
1
1 1 1
grad grad , ,
a a a
a
D D
k k k
k k
x y h
  c c c
= =

c c c
\ .
(20)
and, recalling that at P we have y
1
= h
1
= 0,
1
1
0
a
P
k
x
c
=
c
, (21)
1 1 3 5
1 4
3
a
P
k c c c
y c
c ÷
=
c
, (22)
1 2 3 6
1 4
3
a
k c c c
h c
c ÷
=
c
, (23)
Curvature of the plumblines of the Earth’s normal gravity field
Stud. Geophys. Geod., 55 (2011) 507
i.e.,
( )
( )
1 1 1
, ,
1 3 5 2 3 6
1 1 1 4 4
3 3
grad , , 0, ,
x y h
D
P
c c c c c c k k k
k P
x y h c c
    ÷ ÷ c c c
÷ =
 
c c c
\ . \ .
(24)
in the ( )
1 1 1
, , x y h system. But the system ( )
1 1 1
, , x y h is a parallel transportation of the
system ( ) , , x y h hence
( )
( )
, ,
1 3 5 2 3 6
4 4
3 3
grad , , 0, ,
x y h
D
P
c c c c c c k k k
k P
x y h c c
    ÷ ÷ c c c
÷ =
 
c c c
\ . \ .
. (25)
Finally, if we want to find the coordinates of ( ) grad k P in the global Cartesian system
( ) , , X Y Z , we apply the following matrix formula
1 3 5
2 3 6
0 sin cos 0
3 sin cos sin sin cos
3 cos cos cos sin sin
P
P P
P P P P P
P
P P P P P
P
k
X
k
c c c
Y
c c c
k
Z
ì ì
 ì  ì 
 ì  ì 
(
c
(
c
(
÷ ( (
(
c
( (
(
÷ = ÷ ÷
( (
(
c
( ( ÷ (
¸ ¸ ¸ ¸
(
c
(
c
(
¸ ¸
. (26)
The above 3 × 3 matrix is invertible (if 2
P
¢ = ±t ) and hence from Eq.(26) the
coordinates of ( ) grad k P can be determined.
3. ISOCURVATURE LINES
Suppose that
( ) ( ) ( ) ( ) ( )
, ,
is is is is is
a a t X t Y t Z t = = (27)
is a curve in the three dimensional space, ç is a conservative vector field and
( ) , , k k X Y Z = is the curvature function for ç , i.e. it describes the curvature for its
integral curves for a specific point ( )
0 0 0
, , X Y Z . For example, the integral curves for an
electric field or magnetic field are known as field lines, and integral curves for the
velocity field of a fluid are known as flow lines. In the case of the gravity field, the
integral curves are the plumb lines. Let k
is
be the curvature function for
is
a ,
( ) , ,
P P P
P P X Y Z = is a point lying on
is
a and ( ) a a t = is the equation of the plumb line
of the field ç passing through P. We say that the curve
is
a is an isocurvature line of ç
G. Manoussakis and D. Delikaraoglou
508 Stud. Geophys. Geod., 55 (2011)
with a  field strength ( ) ( ) , , : , ,
a P P P P P P
R X Y Z k X Y Z = if the following relation holds
(the symbol “ ” means the composition)
( ) ( ) ( ) , ,
is P P P
k a t k X Y Z = . (28)
We are going to determine two (among the many) isocurvature lines of the Earth’s
normal gravity field passing through a point P above the ellipsoid. The curvature of the
normal plumb line passing through the point P is equal to ( ) , ,
P P P
k X Y Z . Due to the
rotational symmetry of the field all the plumb lines which pass from points with
coordinates ( ) , ,
P P
h ¢ ì will have the same value of curvature. The locus of these points
is a circle lying on the plane
( )
2
2
sin sin
P P P P N P
b
Z N h f
a
  
 
= + = 

\ .
(29)
and it has the following vector equation
( ) ( ) ( ) ( ) ( ) ( )
cos cos , cos sin , sin
I
is P P P P P P N P
a N h N h f ì   ì   ì  = + + , (30)
where the superscript “I” is related to the first set of isocurvature lines.
Since the plumb lines of the normal gravity field are planar curves (lying in a meridian
plane) the second isocurvature line of field strength ( ) , ,
P P P
k X Y Z is a planar curve and
lies on the meridian plane of the point P. If we introduce the local Cartesian coordinate
system ( )
1 1 1
, , x y h (see transformations (4) and (5)) then the equation of this isocurvature
line in the set D will be the solution of the equation
( ) ( )
1 1
, k y h k P = , (31)
which is obviously a very complicated algebraic equation and cannot be solved
analytically. But from the solution of the initial value problem it is possible to determine
the direction of the isocurvature line in the set D. If
( ) ( )
1 1 1
II II
is is
a a y h y = ÷ (32)
is the solution of Eq.(31) then the direction of the isocurvature line at P is given by
1
1
1
II
is P
P
P
k
y
da
dy
k
h
c
c
= ÷
c
c
(33)
and a point Q
1
close to P of the isocurvature line will have position vector
( )
1 1 1
1
1 1 1 1
1
0, , 0, ,
P
Q Q Q
P
k
y
x y h h h
k
h
o o
 
c

c

= = ÷

c


c
\ .
, with 0 < δh
1
< 1 m , (34)
Curvature of the plumblines of the Earth’s normal gravity field
Stud. Geophys. Geod., 55 (2011) 509
where the constraint 0 < δh
1
< 1 m is imposed because of the way we have defined D in
Section 2.
Repeating this procedure it is always possible to determine the direction of the
isocurvature line at the next points. Because of the fact that with this procedure we cannot
locate the points of the isocurvature line exactly, after the nth iteration step we have to
apply a correction to the coordinates of the point Q
n
. The steps of this correction are the
following:
a) We determine the value of curvature at the point Q
n
, ( )
n
k Q and subtract it from
( ) k P .
b) Since the difference ( ) ( )
n
k P k Q ÷ is very small the new position vector of Q
n
will be
( ) ( )
1 1
1
0, ,
new
n n
n
n
n
Q Q
Q
Q
k P k Q
x y h
k
h
 

÷ 
= +

c


c
\ .
. (35)
The second isocurvature line
II
is
a is vertical to the first and has two end points on the
surface of the ellipsoid. The first statement is obvious since the first isocurvature line lies
on a plane parallel to the equatorial plane and the second is situated on the meridian plane
of the point P. For the second statement, let point P be situated outside the ellipsoid and
( ) k P be the value of R
a
at the point P
. As the isocurvature line approaches the equatorial
plane it intersects plumb lines whose curvature becomes smaller and tends to zero. The
plumbline which is on the meridian plane ì = ì
P
and has the point ( ) ( ) , 0,
P
¢ ì ì = as
a “starting” point is a straight line, and hence it has curvature equal to zero. Furthermore,
since the isocurvature line intersects plumb lines with nonzero curvature (and the field is
stronger as we move towards the surface of the ellipsoid), this means that it approaches
the ellipsoid and one of the last points will be on the surface of the ellipsoid. The same
situation holds as we approach the North Pole, and therefore the isocurvature line has the
second of its end points closer to the North Pole. Due to the symmetry of the ellipsoid the
situation depicted in Fig. 2 is produced (the figure is drawn schematically).
4. BREAKING AN ISOCURVATURE LINE
The line segments d
1
and d
2
connecting the two end points with the point P are called
diameters of the isocurvature line. As the geometric height of P tends to infinity (¢
P
and
ì
P
remain constant), the two end points approach asymptotically the equator and the
North (or South) Pole respectively. If the point P “reaches infinity” then the isocurvature
line “breaks” in two pieces, one along an axis x' (which is the intersection of the
meridian plane of P and the equatorial plane of the ellipsoid) and the other along the axis
G. Manoussakis and D. Delikaraoglou
510 Stud. Geophys. Geod., 55 (2011)
Z (as shown in Fig. 3). The first piece is a straight line vertical to the ellipsoid and lies on
the equatorial plane and the second is a vertical line to the ellipsoid at the North (or South)
Pole. Considering that these straight lines are plumblines of the normal gravity field we
can further state the following:
Fig. 2. The isocurvature lines of the normal gravity field on the meridian plane.
Fig. 3. The diameters of an isocurvature line.
Curvature of the plumblines of the Earth’s normal gravity field
Stud. Geophys. Geod., 55 (2011) 511
a) In case of R
a
= 0 the isocurvature lines are also plumblines of the normal gravity
field.
b) Finite diameter length of the isocurvature line means nonzero curvature for the
plumblines.
c) As the curvature of the plumblines tends to zero then d
1
and d
2
tend to infinity.
5. ISOCURVATURE SURFACES
Suppose that
( ) ( ) ( ) ( ) ( ) ( )
2 3
1
: : , , , , , , ,
is is is is is
s D u v s u v X u v Y u v Z u v 9 ÷ 9 ÷ = (36)
is a vector equation for a coordinate patch (i.e. a smooth part) of a surface S in the three
dimensional space. Let ç be a conservative vector field, ( ) , , k k X Y Z = the curvature
function and ( ) , ,
P P P
X Y Z is a point on S. We say that the surface S is an isocurvature
surface of ç with a  field strength ( ) ( ) , , , ,
a P P P P P P
R X Y Z k X Y Z = if the following
relation holds
( )( ) ( ) , , ,
is P P P
k s u v k X Y Z = . (37)
For the normal gravity field holds that the isocurvature surface passing through the
point P is a surface of revolution. Therefore, since the field is symmetric, let P be on the
XZ plane. The isocurvature line which lies on the meridian plane of P has a local
parametric representation of the form
( ) ( ) ( ) ( )
1 2
, 0,
II
is
a f f    = . (38)
In other words, the isocurvature surface passing through a point P  due to the
symmetry of the field  is generated from the rotation of the isocurvature line on the
meridian plane about the Zaxis. Therefore a parametric representation for a coordinate
patch of the isocurvature surface passing through the point P in ( ) , ¢ ì coordinates has the
form
( ) ( ) ( ) ( ) ( )
1 1 2
, cos , sin ,
is
s f f f  ì  ì  ì  = . (39)
In summary:
a) Since we shoved that there is one isocurvature surface passing through any point P
above the ellipsoid, it means that there are infinitely many isocurvature lines
passing through the point P, all lying on the surface S.
b) In turn, since these isocurvature lines can be thought of as covering (at least
locally) the surface S, for any direction there is always an isocurvature line.
Mathematically, this means that the family of these isocurvature lines passing
through the point P can be described from a second order system of ordinary
differential equations.
c) In addition, since the isocurvature line on the meridian plane is the generating
curve for the isocurvature surface we define the isocurvature line on the meridian
plane as the mother isocurvature line.
G. Manoussakis and D. Delikaraoglou
512 Stud. Geophys. Geod., 55 (2011)
6. CONCLUSIONS
We have outlined a method for determining the gradient grad k of the curvature of the
plumblines of the Earth’s normal gravity field at a point P without using the (ordinarily
required) third order partial derivatives of the normal potential U. With this method extra
derivations and inversion of complicated matrices are avoided. The hypothesis about the
second partial derivatives of U was that they do not change in the interior of a unit circle
(i.e. of radius c = 1 m) whose center was the point of interest P. This methodology is
useful for the determination of the isocurvature lines passing through any point P. Along
these lines the afield strength R
a
of the normal gravity field is constant and its value
( )
a
R P is equal to ( ) k P . We shoved that there are at least two isocurvature lines passing
through a point P, one on the meridian and the other on the parallel plane of P, and hence
they are perpendicular to each other. Both isocurvature lines represent a family of
isocurvature lines as P changes its position above the ellipsoid, with the only restriction
that P is situated on a meridian plane. The first family represents circles which lie on
planes which are parallel to the equatorial plane and the second family represents curves
on the meridian plane which have two end points on the surface of the ellipsoid.
The second family is more interesting because, as the geometric height of the point P
tends to infinity, the isocurvature line tends asymptotically to the x' axis on the equatorial
plane and the Zaxis, and subsequently “breaks” into two pieces. These two pieces have
the special property that they are plumb lines of the normal gravity field and they have
a vanishing value of curvature.
Finally we shoved that there are infinitely many isocurvature lines passing through any
point P and they lie on a special surface which is called an isocurvature surface of the
normal gravity field. The isocurvature lines and isocurvature surfaces are new geometric
objects whose geometric properties may reveal new properties for the normal gravity field
and thus contribute to a better understanding of the geometrical properties of the Earth’s
normal gravity field.
APPENDIX A
The matrix M
1
in Eq.(3) is equal to
2 2
2 2 2
2 2
2
2 2 2
2 2 2
2 2 2 2 2
2 2
2
cos cos cos sin 2 sin 2 cos
sin 2 cos sin 2 sin 2 cos sin cos
2 2
cos sin 2 cos cos 2 sin 2 sin
2 2
sin cos sin sin 2 sin 2 cos
sin 2 sin 2 sin 2 cos 2 cos
4 2
u u u
f
f f
u u u
f
f
u f
u
f f u f
f f
u f
 ì  ì  ì
 ì  ì   ì
 ì  ì  ì
 ì  ì  ì
 ì  ì 
 
÷ ÷ ÷ 

\ .
÷ ÷
÷
÷ ÷
2 2 2 2 2
sin
cos sin cos sin 2 0 f f
ì
 ì  ì
÷
¸
Curvature of the plumblines of the Earth’s normal gravity field
Stud. Geophys. Geod., 55 (2011) 513
2
2 2 2
2
2
2 2 2
2
2 2 2 2 2
2
2
2 2 2
cos sin sin 2 sin sin
sin 2 sin cos sin sin sin 2
2 2
cos sin 2 sin 2 cos 0
2 2
sin sin sin 2 sin cos
sin 2 sin 2 cos cos 0
4
cos cos 0 0
u u
f
f
u u u
f
f
u f
f u f u
f
u f
f
 ì  ì 
 ì   ì 
 ì  ì
 ì  ì 
 ì  ì
 ì
(
(
(
(
 
(
÷ ÷ 
( 
\ .
(
(
(
(
(
÷
(
(
÷
(
(
(
¸
. (A.1)
The real number terms c
i
(i = 1, 2, …, 12) which are encountered in Eqs.(17) through
(19) are given by the following expressions:
2 2 2 2 2 2 2
1
2 2 2 2
1 1 1 1 1 1 1 1
1 1 1 1
2 2
2 2 2 2 2
2 2
1 1 1 1 1 1 1 1 1
1 1
,
U U U U U U U U U
c
y y h y h y h y
y h y h
U U U U U U U U
h y h y h y y h h
y y
   
c c c c c c c c c
= + + +  
 
c c c c c c c c
c c c c
\ . \ .
¸
(
 
     
c c c c c c c c
( 
 ÷ + ÷ +  
  (  
c c c c c c c c c
c c 
\ . \ . \ .
(
\ .
¸
(A.2)
2 2
2 2 2 2 2
2
2 2
1 1 1 1 1 1 1 1 1
1 1
2 2 2 2 2 2 2
2 2 2 2
1 1 1 1 1 1 1 1 1
1 1 1 1
,
U U U U U U U U
c
y h y h y h y y h
h y
U U U U U U U U U U
h y h y h y y h h
y h h y
 
     
c c c c c c c c

= + + +   
   
c c c c c c c c c
c c 
\ . \ . \ .
¸
\ .
(    
c c c c c c c c c c
÷ + ÷ + (  
 
c c c c c c c c c
c c c c (
\ . \ .
¸
(A.3)
2 2 2 2
3
2 2
1 1 1 1 1 1 1 1 1 1
1 1
U U U U U U U U U U
c
y y h y h h y y h h
h y
   
c c c c c c c c c c
= + ÷ +  
 
c c c c c c c c c c
c c
\ . \ .
, (A.4)
2 2
4
1 1
U U
c
y h
    c c
= +
 
c c
\ . \ .
, (A.5)
2 2
5
2
1 1 1 1 4
1
1
2
U U U U
c
y h y h c
y
 
c c c c
= + 

c c c c
c
\ .
, (A.6)
2 2
6
2
1 1 1 1 4
1
1
2
U U U U
c
y y h h c
h
 
c c c c
= + 

c c c c
c
\ .
, (A.7)
G. Manoussakis and D. Delikaraoglou
514 Stud. Geophys. Geod., 55 (2011)
2 2
2 2 2 2 2 2 2 2
7
2 2 2 2
1 1 1 1 1 1 1 1
1 1 1 1
U U U U U U U U
c
y h y h y h y h
y y h h
(  
     
c c c c c c c c
( 
= + ÷ +   
   ( 
c c c c c c c c
c c c c 
\ . \ . \ .
(
\ .
¸ ¸
, (A.8)
2 2
2 2 2 2
8
2 2 2
1 1
1 1 1
U U U U
c
y h
y h h
 
     
c c c c

= ÷ +   
   
c c
c c c 
\ . \ . \ .
\ .
, (A.9)
2 2
2 2 2
9
2
1 1 1 1
1
2 2 2 2 2
2 2 2
1 1 1 1
1 1 1
,
U U U
c
y h y h
h
U U U U U
y h y h
h y h
 
   
c c c

= +  
  
c c c c
c 
\ . \ .
\ .
¸
(
 
c c c c c
÷ + ( 

c c c c
c c c (
\ .
¸
(A.10)
2 2
2 2
10
2
1 1 4
1
1 U U
c
y h c
y
 
   
c c

= +  
  
c c
c 
\ . \ .
\ .
, (A.11)
2 2 2 2
11
2 2
1 1 1 1 4
1 1
1
2
U U U U
c
y h y h c
y h
 
c c c c
= + 

c c c c
c c
\ .
, (A.12)
2 2
2 2
12
2
1 1 4
1
1 U U
c
y h c
h
 
   
c c

= +  
  
c c
c 
\ . \ .
\ .
. (A.13)
All the above partial derivatives are determined at the point P. Note also that Eq.(A.5)
corresponds to the wellknown expression for the square of the normal gravity, i.e.
2
¸ .
References
Abramo R.L. and Finelli F., 2001. Attractors and isocurvature perturbations in quintessence models.
Phys. Rev. Lett. D, 64, 1÷10.
Ardalan A. and Grafarend E., 2001. Somigliana  Pizzetti gravity: the international gravity formula
accurate to the subnanoGal level. J. Geodesy, 75, 424÷437.
Grafarend E., 1997. Fields line of gravity their curvature and torsion, the Lagrange and the Hamilton
equations of the plumbline. Annali di Geophysica, XL, 1233÷1247.
Grafarend E. and Ardalan A., 1999. World Geodetic Datum 2000. J. Geodesy, 73, 611÷623.
Heiskanen W.H. and Moritz H., 1967. Physical Geodesy. W.H. Freeman and Co., San Francisco.
Moroi T. and Takahashi T., 2004. correlated isocurvature fluctuation in quintessence and suppressed
cosmic microwave background anisotropies at low multipoles. Phys. Rev. Lett. D, 92, 1÷4.