151
cosßcosy cos a sin y + sin a sinß cos y sinasiny-cosasinßcosy
-cosßsiny cosacosy-sinasinßsiny sin a cos y + cos a sin ß sin y
sin ß -sinacosß cosacosß
days, where 68 points were observed. The purpose of this
campaign was the enlargement the EUREF reference frame into
the continental part of Croatia, and to establish dense
geodynamic network covering the area around the Adriatic sea.
At this campaign 5 EUREF points in Slovenia and 10 points in
Croatia were observed once more.
TRANSFORMATION OF COORDINATE SYSTEMS
With the availability of high precision coordinate systems the
transformation of three dimensional coordinate systems is
becoming frequent computation. There are several ways of
transforming one coordinate system to another. At this place we
will only mention two of them. The most common
transformation is similarity transformation in which the scale
factor is the same in all directions. This transformation
preserves shape, but the distances and point positions may
change. The second is so called orhogonal transformation in
which the scale factor is unity. This transformation preserves
the distances and the shape, but the positions of points do
change on transformation.
Similarity transformation
If we define x=(jc| ,,jc 2 ,,^3,/), i=l,...,n the coordinates of n>3
points with respect to {O.bi.b^b^} and x'=(jc 1>/ '^c 2 ,/’^3,/'), 1=1,.-»«
the coordinates of the same points with respect to {0,b\',b 2 ',b 3 '}.
The 7 similarity transformation parameters (t',oc,(3,y,dm) can be
determined with the 7 coordinates given in both coordinate
systems. In the case when the number of common coordinates
given in both systems is larger than 7 we get the overdetermined
linear (linearized) system of unknown parameters which are
estimated by least squares solution.
Transformation written in the form of equation (1) where m= 1
is an orthogonal transformation. It is used when the presumption
of equal scale factor for both coordinate systems can be
accepted. In this case only six transformation parameters relate
both coordinate systems (t',a,(3,y).
ASTROGEODETIC NETWORK OF SLOVENIA
AND ETRS 89
Similarity transformation is usual transformation procedure in
geodesy, where coordinate systems established by different
measuring techniques are orthogonal but of different scale.
Let S={0,b l ,b 2 ,b ? ,} and S’={0,b\,b 2 ',bT,'} be Cartesian coordinate
systems, and let x=(x!,x2,x 3 ) and x'=(xi',x2',x 3 ') be a Cartesian
coordinates of a point with respect to the coordinate systems S
and S' respectively.
In the general form a similarity transformation between
coordinate systems is written:
x'=/»Rx + t' (1)
where R is an orthogonal transformation matrix RR'=I,
m=(l+dm) is the scale factor, t'=(ti,t 2 ,t 3 ') is the vector of the
coordinates of the coordinate system's S origin in the coordinate
system 5'. Every orhogonal matrix R can be described by three
parameters, where there are several possibilities to choose these
three representation parameters. Usual representation of these
parameters in geodesy are Euler- or Cartan- angles. Here we
mention Cartan- angles, where every orthogonal matrix can be
represented as:
R = R 3 (y)R 2 (p)R,(a)
As we already mentioned, the astrogeodetic network of
Slovenia is classical triangulation network. In classical
triangulation networks angles are determined with much higher
accuray than baseline lenghts and in general all such networks
exhibit quite large scale errors. Some previous investigations
(Jenko, 1986) confirmed the hypothesis that the situation is the
same also in the astrogeodetic network of Slovenia.
For the 27 of first order triangulation points, besides the official
coordinates in the national coordinate system, also the positions
in ETRS 89 coordinate system are available. Both sets of
coordinates of these points enable some comparisons of
coordinate systems. Points with positions given in both system
are in Figure 2 marked with the filled circles and squares.
For the transformation procedure Cartesian coordinates of the
points in the astrogeodetic net have to be computed. These
coordinates were computed using the following relations:
X
(N +h) cos 9 cos À
y
-
(N + /i)cos9sin^
z
[(1 - e 2 )N + /r]sin 9
e'=2 f-f 1
where N = . ^ — ,
yj 1 - e 2 sin 2 9
(2)
with:
R,(y)
R,(a)
cosy
siny
0'
cosß
0
-sinß
-sin y
cosy
0
, R,(ß) =
0
1
0
0
0
1
sinß
0
cosß
'1
0
0
'
0 cosa sina
0 -sina cosa
The orthogonal rotation matrix R is decomposed into three
consecutive rotations around the b\, b 2 and ¿3 axes. In the
general form of the Cartan- representation of an orthogonal
matrix is:
R = R(a,ß,y) =
Parameters of official reference elliposid i.e. the Bessel
reference elliposid are a=6377397,155m and
l//=299,1528128533. Ellipsoidal height /? of each point in the
net is computed from h = H + N, where the geoidal height N is
computed with the existing geoid model (Colic et al., 1992a)
Transformation between national coordinate system and
ETRS 89
We have performed the 7-parameter similarity and 6-parameter
orthogonal transformations of ETRS 89 coordinate system to
the national coordinate system. This way the estimation of some
attributes of the official national coordinate system became
possible.
The 7-parameter transformation is the most common procedure
when combining different coordinate systems. But in general it