Regardless of terrain, at these scales and wavelength, the most important factor in 
settlement detectability was the high return or cardinal effect caused by corner 
reflectors. This condition was particularly apparent in environments where settlements 
were located in meadows separating forest stands. The high return from settlements 
provided a sharp contrast from the low return of the surrounding grasses and darker, 
textured areas of the woods. However, in a volcanic mountain area of vineyards (i.e. the 
Kaiserstuhl) near F reiburg, this high return contrast between settlement and background 
was not as positive. Here, the stepped terraced vineyard slopes presented a high signal 
return similar to settlements when the former were oriented perpendicular to the look 
direction. In such cases the interpreter had to carefully observe the location of the return 
as well as the areal extent and texture, if any. Settlements were more apt to be situated 
in valleys and/or at the foot of the mountain as opposed to terraces which were located on 
the mountain slopes. 
Pass 3, flown near orthogonal to the first two passes, employed an incidence angle 
(AO.9 degrees) between that of the other two passes. This pass produced the highest per 
cent of settlements detected (85.0 per cent) as well as the highest accuracy (76.A per cent) 
of the three passes (Table II). In mountainous areas and wooded areas the detectability of 
settlements, compared to other terrain areas of the pass, was noticeably low. Relief, 
slope angle, and orientation relative to look direction all contributed to poor visibility 
(e.g., the Black Forest area). However, the settlements located in the Kaiserstuhl were 
more easily detected and in greater accuracy that on Pass 2. It is suggested that this 
increase in accuracy was a result of the change in incidence angle and look direction. 
A similar occurrence was noted in the Rhine valley near Freiburg. Here, settlements 
were confused with similar bright returns from wetlands and fields perched with water on 
Pass 2, but were distinctly different in tone and texture on Pass 3. In this case the cause 
is believed more attributable to incidence angle, suggesting that the point where signal 
return is dominated by surface roughness rather than by a combination of roughness and 
signal penetration of the canopy lies between AO.9 degrees and 52.3 degrees. This 
possibility, however, requires considerably more evidence to validate. It was noted that 
settlements were easily identified in the plains of northeastern France. Several bright 
spots around Verdun, later identified as old forts and battlements, were also detected but 
not labeled as settlements due to their variance in pattern and size. 
Pass 1 and Pass 2 were virtually parallel in look direction and varied only in 
incidence angle. In this case the difference in accuracy was attributed to change in 
incidence angle as the results correlated with those found by other researchers (Wise and 
Trinder, 1987; Drake and Patton, 1980). Although Pass 3 recorded the highest accuracy of 
the three passes the poor detection in mountain and forest areas and high visibility of 
settlements in plains posed a question as to the consistency of settlement detection among 
the three passes over similar terrain. The question was lent further cause when the 
change in settlement visibility on the Kaiserstuhl was considered. These observations 
pointed to look direction as a possible explanation.