Participant Sample spacing Coeff. for straight line approx. 
Ax meter of terrain spectrum 
a E 
A 30 2.80 2.56-107? 
B 30 2.95 1.591073 
D 5 2.58 8.12107? 
E 5 2.64 6.00*107° 
  
  
  
  
  
Table 3, Drivdalen, Norway. 
Predicted standard deviations for DEM at 4 different grid sizes 
  
Participant Ax=10 meters  Ax-20 meters  Ax-30 meters  Ax-50 meters 
  
A 9.56 1.04 1.50 2.38 
B 0.53 1.04 1.55 2.54 
D 0.76 1.32 1.82 2.73 
E Q.71 1.25 1.74 2.64 
  
  
  
  
Table 4. Drivdalen, Norway. 
If the terrain has a spectrum L°C? = 2.56-1073- A?:??! as assumed in the 
case of participant A, and a digital elevation model was measured by points 
spaced 10 meters apart in a grid, the standard deviation between the ter- 
rain and the DEM equals 0.56 meters. It is, however, assumed that no measu- 
ring error is present in the data points. If the grid size is 20 meters, 
the standard deviation equals 1.04 meters, if Ax - 30 then S = 1.50 meters. 
The standard deviation is directly related to the part of the spectrum 
having wavelengths smaller than twice the distance between the data points. 
In case A and B lower values for the spectrum are estimated for the short- 
wave part than in case D and E. Consequently a smaller standard deviation is 
predicted in the case for D and E. 
Bohuslän, Sweden. 
Profiles from the test area Bohuslän have been measured by 4 groups: A,B, 
D and E. In figure 4, spectra derived from profiles, measured by B and 
by D are shown. The profiles from B were sampled with a point distance of 
30 meters, while the profiles from D were sampled with a spacing of 5 me- 
ters. The profiles from A,B and D were measured at scale 1:30.000, while 
the profiles from E were sampled at scale 1:5.300.