Fig. la 
Fig. 1b 
SPOT PAN. X ENS 
  
  
Goult 
Not to Scale 
  
profile 
—_—__ 
  
  
  
Lamanon 
  
  
profile 
  
  
profile 
  
Salon 
  
Aix-en-Provence 
  
  
  
  
  
  
  
  
  
Fig.2: 
Test sites for grid and 
measurements 
profile 
  
306 
REE Si 
SPOT FAN. £ 
  
Area Size | No of| Zmin | Zmax |Std.dev. 
(km) | pts. (m) | (m) (m) 
  
AIX 5x5 603 
GOULT | 5x5 612 
LAMANON| 6x6 761 
SALON | 6x6 736 
-57.6:|- 50.21. 22.6 
252.4 1553.44 55293 
-62.4 78.47 30.6 
-49.7 55:41 21.7 
  
  
  
  
  
  
  
  
Table 5 ; Summary of grid computations 
  
Area Prof.| No of| Zmin |Zmzx Std.dev. 
(km) | pts. (m) (m) (m) 
  
AIX 36 174 -46.0 | 47.6 18.5 
GOULT 36 152 =57.3 | 55.3 27.9 
LAMANON  |17 122 -60.1 43.21 21.0 
SALON 24 93 -56.9 50.0] 23.0 
  
  
  
  
  
  
  
  
Table 6 : Summary of profile computations 
Both grid and profile measurements in the TM/SPOT 
are burdened with systematic errors in some areas 
due to a radiometric imbalance between the THM 
image (May 1988) and SPOT image (Oct. 1989). 
Outliers had to be rejected; therefore tolerance 
values were set to 2.7 of the standard deviation. 
grid measurements of  LAMANON, 
shoving very large height discrepancies, ve can 
accept an average standard deviation of 22 m, 
defining tolerance values of + 60 m. The outputs 
Disregarding the 
of the height difference models, previously 
corrected for bias have been visualized in two 
ways: 
- by grid stereograms obtained from two pers- 
pective views (fig. 3) 
- by profiles showing the direction of the scan 
lines in the transformed TM image (fig. 4).