CIP A 2003 XIX th International Symposium, 30 September - 04 October, 2003, Antalya, Turkey 
720 
2. LOCAL REFERENCE SYSTEM 
2.1 General Strategy 
According to the agreed standards of the geodetic community 
surveyors do their best to bring all measurements of a project 
into a common geodetic reference frame. The normal way to 
achieve this is to connect the results of the local survey to the 
official state system. This, however, was not possible in our 
project due to the fact that no official co-ordinates of any points 
in the project area have been available until today. That is why 
is was decided to define one physically identified point, which 
is known to be part of the official Turkish reference system 
(point no. 1000, see Figure 2) to be the fundamental point of a 
local reference system. 
Figure 2. Points of local reference network 
To avoid negative co-ordinates in the project area this point was 
given local co-ordinate values in East direction E = 10000,000 
m and in North direction N = 10000,000 m in the same way. A 
Turkish 1/5000 scale map was available on which the 
fundamental point was identified. To make the new local 
system fit as well as possible to the height system of that map, 
the fundamental point was given the rounded up height value H 
= 1130,000 m above sea level which was very close to the 
annotated point height value of 1129,987 m in the map. By that 
way of establishing the reference system it was made sure that 
the height system fits with the official system and that no 
confusion between the currently used local co-ordinate values 
for Eastings and Northings and any other co-ordinate system 
which will eventually be used in future can occur. 
2.2 GPS Measurements 
A set of points was established to serve as the geospatial base 
for all subsequent measurements. The co-ordinates of these 
points were determined by GPS baseline measurements with 
two Trimble 4700 system receivers in static mode and by 
calculation of co-ordinate values in postprocessing mode. The 
results of the baseline measurements were processed with 
Terrasat’s software package GeoGenius® which delivered a set 
of geocentric co-ordinates in the WGS 84 reference frame and, 
on the basis of the Earth Model JGM3/OSU91, which is a 
composite of the JGM3 and OSU91A models (Merry, 2003) a 
set of geographic co-ordinates and orthometric heights for all 
measured points. Table 1 shows the list of plane and height co 
ordinates after shifting it to the predefined co-ordinates of point 
no 1000. As can be seen, the standard deviation (sE, sN, sH) of 
the point co-ordinates resulting from a free network adjustment 
are all in the 1 centimetre range thus establishing a highly 
accurate base reference frame. This set of co-ordinates serves as 
the georeferential base for all subsequent work. 
Pt. No. 
Easting 
Northing 
Height 
sE 
sN 
sH 
1000 
10000,000 
10000,000 
1130,00 
0,0 
0,0 
0,0 
2000 
9853,615 
9570,622 
1092,82 
3,2 
2,8 
7,8 
3000 
9768,449 
9419,180 
1080,17 
3,2 
2,8 
8,1 
4000 
8790,385 
8922,818 
1108,18 
2,8 
3,1 
7,2 
6000 
9292,813 
10415,763 
1221,69 
2,5 
1,9 
4,9 
7000 
9441,799 
11058,075 
1278,81 
2,1 
1,7 
4,5 
8000 
10296,378 
10332,056 
1096,51 
1,9 
1,7 
4,6 
9000 
10272,821 
9833,549 
1095,35 
1,5 
1,2 
3,1 
10000 
9723,471 
10454,293 
1162,28 
2,2 
2,3 
4,7 
11000 
11221,836 
9928,892 
1046,45 
6,9 
5,3 
14,3 
12000 
9480,751 
10308,714 
1193,73 
2,0 
1,5 
3,4 
13000 
10153,024 
10080,383 
1104,21 
1,6 
1,2 
_A2 
Table 1: Reference point co-ordinates with their relative 
standard deviation values (all metric values given in unit metre) 
Above that, the geographical co-ordinates as obtained from the 
GPS measurements established an absolute geodetic reference 
frame within the accuracy which can be obtained from absolute 
GPS measurements. 
2.3 Quality Checks 
The quality of the local area network was checked in several 
ways. Firstly, the height of the local fundamental point no. 1000 
which was obtained from the world wide Earth model 
JGM3/OSU91 was compared with the height which was taken 
from the map. The difference was in the 4 metre range, which 
seems to be within a reasonable scale. Secondly, a set of 
distances was measured by electronic distance measurements 
between some of the reference points. The comparison of the 
independently measured distances and the distances as 
calculated from co-ordinates showed that the difference in all 
cases was below 2 centimetres (see Haas, 2002). 
3. LARGE SCALE MAPS 
In-depth archaeological field work often needs detailed large 
scale maps for the georeferenced documentation of findings, 
etc. In several campaigns parts of the landscape of the research 
area were captured by extensive tacheometric surveys. A 
sample of highly detailed 1/500 scale maps was generated (see 
Figure 4) within a predefined map sheet framework (Figure 3).