for profile 1 (Figure 3) is 2.1 m, for profile 2 (Figure 4) 3.2 m. Profile 1 was obtained
from an altitude of 780 m, profile 2 from an altitude of 480 m. The standard deviation
expressed in percent of flying height is 0.3 and 0.7 respectively. The difference in
results between line 1 and 2 is due, at least in part, to further problems with the field
survey identified while in the field. Time, however, only permitted the correction of
line 1. Corrections for line 2 were requested from the local surveying contractor, but a
response has not yet been received. In spite of these difficulties, the results look
excellent and are well within the initial accuracy requirements of 2 and later upgrade to
1% of flying height.
The photogrammetric profiles (Figure 6) were not only useful for evaluating
the tree canopy profile, but they also showed that the absolute orientation values
provided by the radar altimeter and the tilt indicator are sufficiently accurate to permit
photogrammetric restitution of several adjacent models.
With one exception, the ground profiles produced by the two means (photo-
grammetric and radar) deviated by less than 5 m. Near the 500 m point along the profile,
the photo ground profile rose about 8 m above the radar counterpart. This was mainly
due to the difficulty of seeing through the vegetation to the ground on the photos. In any
case, under the best conditions, the photogrammetric profile is likely to vary by 2 or 3 m
around the true position.
On average, the forest profiles produced by the two methods are very close.
However, the photo profile is subject to much greater variation than the radar
counterpart, frequently swinging both above and below the radar profile. The reason for
this is that the photo profile is based on point measurements made directly off the
surface vegetation of individual trees along the profile whereas the radar profile, in
effect, is based on an average covering a considerable area. First, the radar altimeter
sample is based on a pulse of energy that is radiated over a circular area which can be
regarded as a cone (Figure 7). The angle of the cone would depend upon the antenna
FIGURE 7. Sketch representing area illuminated by radar roughly in relation
to size of photograph.