= 1550 — 
Analysis and Results 
System Parameters 
Radar imagery is quite similar in appearance to black-and-white 
photography. However, as Coiner and Dellwig (1972) have pointed out, 
the interpreter must be aware that the causes of variation in tone, 
texture, shape, and other photo interpretation clues are distinct and a 
function of radar system related parameters. In conjunction with diverse 
environmental conditions these factors affect and alter the land use 
information obtainable. 
The amount of radar shadow is a function of topographic relief and 
look angle. Although less prevalent in Study Area I it tended to pose 
more problems than in Study Area II. Land use activity was most intensive 
in valleys and river plains but radar shadow often prevented analysis or 
led to confusion with water. The combination of tree vegetation and 
shadow concealed settlement and transportation arteries and hindered 
interpretation. In Study Area II radar shadow enhanced field borders in 
level to gently rolling terrain by highlighting riparian vegetation, 
vegetation along field borders, and the drainage network. In the western, 
more arid and rugged segments of this study area shadow was more frequent 
but a less significant factor. Land use was more extensive (range and 
grazing) and settlement less dense than in the Northeast. Consequently, 
there was less danger in having vital landscape features hidden. 
Tone and texture contrasts were also much more apparent and distinct 
in Study Area II. The combination of climatic changes and a regular 
field pattern composed of large fields presented rather precise tone and 
texture variations to the interpreter. Patterns were easily discerned. 
In the Northeast a mottled, more uniform appearance was dominant. Less 
contrast in vegetation and between natural and man-related features 
produced few gray tones and textures. Whereas bright returns indicative 
of farmsteads, settlements, and transportation linkages were character- 
istic of Study Area II, the complex, concentrated activity in the 
Northeast produced a more homogeneous landscape on radar with fewer 
variations. 
Undoubtedly, the overriding system element affecting detectability 
was resolution. The fifteen-sixteen meter resolution of the K-band 
system was sufficient for thematic land use mapping in Study Area II. 
Extensive land use patterns and the 'blooming'' of certain cultural targets 
were easily visible. In the Northeast, many of the cultural elements 
(farmsteads, roads, settlement, fields) were less than the resolution 
of the system. In combination with radar shadow and/or the nearly omni- 
present tree vegetation these features were frequently averaged into the 
cell producing a "composite" return. This system-environment generaliza- 
tion reduced the landscape detail apparent in the more open land use 
pattern components in the Midwest and West (for example, compare Figure 1 
and 2). That one resolution would not be applicable to all environments 
is common knowledge, but it was believed that for small scale thematic 
land use mapping purposes K-band radar should prove adequate. For one 
environment (Study Area I) this was simply not the case. The modulation 
of physical and cultural features and patterns handicapped the capability 
of rad: 
Howeve1 
landsc: 
enviror 
General 
Ar 
be cre: 
surface 
settlen 
1975). 
to envi 
figurat 
diverse 
was mar 
the rac 
i.e., 1 
change: 
An exar 
is les: 
landfoi 
aside ! 
lowlanc 
by vege 
physic: 
Study / 
Ni 
supplie 
consis: 
ubiqui! 
the We: 
form o: 
landsc: 
frequei 
patten 
and pr 
as ide) 
all we 
by adj. 
possib 
tion e 
Where : 
detect 
areas, 
firmin: 
and an 
Study . 
and ag 
of the 
F 
less)