Profiles 7 and 9 were categorized as long profiles. Profile 71 was 
categorized as short profile, supplemented with its subdivisions and 
those of the long ones: 71-1, 71-2, 7-1, 7-2, 7-3, 7-4, 9-1, 9-2, 
9-3and 9-4. Profiles of varying sample spacings were derived from the 
long ones to produce new profiles at 8, 12, 16 and 20 m sample 
spacings. 
3. PROCEDURE 
All sample profiles were processed using all four programs: SPECTRA, 
LOGKV, LINEAR and RF. Input data to each program included: 
a) Profile header/identifier 
b) Error budget 0, ne (M) 
c) Sample spacing Ad (m) 
d) Profile data X,Y,Z (m) 
The error budget oy... Was determined as follows (see the Appendix, 
11/51 
= 2 a uA zz x 2 
Int (c spec c AT g setup e a) 
For the Raymond profiles: 
Gels ocuan ion 043. a 0. 15m (2) 
  
o 
specs 1.64) 2(1.64) 
o = 0 s 0.03m, mean of RMS's of height residuals after 
samp setup 
exterior orientation of the three models measured. 
1/2 
2 2 : 
Sy (RMS pp + RMS c») (3) 
~ (0.032 + 0.062)}/2 = 0.07 m 
1/2 
d, ar" (0.152 7 0.072 - 0.03? - 0.037) 4 0.12 m 
For the Crowsnest Pass profiles: 
~ I =0.31 m 
2(1.64) 
  
G 
spec 
o = 0 ~ 0.05 m, RMS of height residuals after exterior 
samp setup 
orientation of this model. 
1/2 
c.. = (0.062 + 0.117) 950.12 m 
AT 
/2 
m 2s 2% 2:3 2V Leni 
S nt (0.31 0.12? - 0.05 0.05 ) 0. 28m 
4. RESULTS 
4.1 Long profiles 
These profiles are arranged in decreasing order of roughness factors 
(see Table 1). It can be noted that: 
a) The mean sampling intervals (14) are increasing with decreas- 
-d3-r