5B-1-4 
horizontal repeatability of same point was 40cm±20cm, which is 
same as the distance difference of GPS observations. Sign of 
residual in almost all the lines are minus for X, minus for Y and 
plus for H. It is possible to achieve results with higher accuracy if 
new method of correction is established. 
Test field 2 (long loop area) 
The following tables are the result of deviation of orthometric 
height between the direct leveling and laser profiling. 
Table 5 Height difference between the leveling and profiling 
GSI 
Point name 
Profiling 
Direct leveling 
deviation 
11231-1 
25.79 
24.861 
0.93 
11231-2 
25.87 
24.929 
0.94 
11231-3 
27.04 
26.10 
0.94 
11231-4 
25.81 
24.87 
0.94 
High - energy research center 
11228-1 
28.91 
29.882 
-0.97 
11228-2 
29.04 
30.002 
-0.96 
GSI 
11231-3 
25.12 
26.10 
-0.98 
11231-4 
23.87 
24.87 
-1.00 
Height difference between the direct leveling and laser profiling is 
around lm in each point. However, they are opposite in sign at 
GSI and high-energy research center. Since the loop was started 
and ended near GSI, it was able to confirm the coordinate 
difference at same point. It confirmed that there are large 
differences. The differences were 2.2m, 0.3m and 1.9m along X, 
Y and H directions respectively. It was found that the GPS 
analysis is responsible for those large differences. In the present 
case, it can be imagined that a data discontinuity was occurred 
after flying over GSI due to an unknown reason. Hence, it was 
unable to have continuous kinematics processing over whole loop. 
Therefore different ambiguities have been used for different data 
sets. Fig.6 shows the standard deviation of GPS data observed by 
helicopter between GSI and high-energy research center. It found 
obviously that there was a cycle slip during flying. 
0080 
/ 
¿0 060 
| 0 .040 
0020 
1. 
02:4800 (182880) 
2 «// 
Fig. 6 The cycle slip during Hying 
On the other hand, relative value between closer points within the 
same observation lines is tallied. For example, when the height 
difference between 11231-1 and 11231-2 is concerned, profiling 
value is 0.08m and the corresponding value of direct leveling is 
0.07m. They differ only by 0.01m. Similarly, the difference 
between 11228-1 and 11228-2 is 0.01m. This implies that there 
will have possibilities to use this system in public surveys after 
making the height corrections if there are known fundamental 
points in the near vicinity. 
According to the above results, there may have some systematic 
errors in the ranges where there are no cycle slips. This systematic 
error tends to be large if there are cycle slips. Such phenomena 
were not occurred in the present test field 1. These reasons are not 
hold. Hence, it is necessary to develop a method to calculate the 
ambiguities within the ranges where there are cycle slips. As the 
most reliable method, ground control point system can be 
introduced. 
5 CONCLUSION 
The accuracy of horizontal coordinate based on GPS static survey 
was about 40cm ± 20cm. and standard deviation of height on one 
ground control point in one survey line was about 10cm in the 
case of test field 1. However, the residual of height between laser 
profiling and direct leveling was about 20cm for test field 1, and 
in the case of test field 2(long loop area), deviation was about 2m. 
It was obvious that the cause is cycle slip of GPS data. One of 
purpose was to investigate the drift of gyroscope in long loop 
survey, but it was not able to detect the drift, because of cycle slip. 
When the center of the ground control point is decided by 
profiling, a deviation of more or less 10cm can be seen. This can 
be happened due to wrong centering by human eyes and due to 
careless selection of grid data for contour lines. However, this 
does not affect to accuracy, as long as large-scale mapping is 
concerned. However, problem is arisen where the deviation is 
more or less 40cm. As in the case of laser profiling, there should 
have to be made corrections for the differences of several meters. 
In the present study, investigation has been done using ground 
control points and the differences between results are same in sign 
but differ in size. Hence, there are some parts where correction is 
not sufficient only for horizontal displacements. As the swinging 
of aircraft effects to aerial photographs, results may also be 
fluctuated. Hence, correction to the ground control points is 
remained as a problem to be solved in the future. 
REFERENCE 
[Takada, 1998] Takada, K., and Y. Akiyama, 1998. The 
possibility of water surface slope measurement with airborne laser 
profiling method. ISPRS Commission V Symposium, Hakodate, 
Japan. pp284-289.