)r transporta- 
tectability. 
dated the 
able contrast 
ingly more 
> of environ- 
s is summarized 
ed employing 
al. 1976 
east one of the 
n both study 
e Northeast owing 
also more easily 
nilar settlements 
dily identifiable 
ut with more 
d smaller, 
forest. 
fication proved 
on, was found 
one and texture 
ar imagery posed 
e and judgement 
heast to avoid 
(6) Wetland 
absence of 
surrounding 
e approached 
s extant in 
ng activities. 
W. 
| the rather 
. and precise 
ies impossible. 
| differentiated 
rge urban 
ices, Industrial, 
or transportation 
Residential Land 
large as well 
listrict and 
were manifest 
other urban 
amount of 
- was not possible 
ries in the 
categories : 
ries, and 
- 1555 = 
Ornamental Horticultural Areas; Confined Feeding Operations; Other 
Agricultural Land. As discussed above, Cropland and Pasture could be 
identified quite easily in Study Area II but was more difficult in Study 
Area I owing to different cultivation patterns and field sizes. For 
example, compare Figures 1 and 4 with 5 and 7. Confusion of pasture with 
forest regrowth was also a problem in the Northeast. Only Groves of the 
second category were visible and present in the West and Midwest. In 
the Northeast, Groves frequently merged with forest precluding detection. 
While a few small orchards did exist a lack of contrast with other 
vegetation prevented detection. It is suggested that orchards would be 
detectable when present in the form of large commercial operations. 
In the West and Midwest study area Rangeland was identifiable and 
divided into improved and natural pastures based on tone and texture 
clues and such surrogate data as field borders, size, shape, presence 
of farm ponds, etc. (see Figures 4 and 8). While this land use may be 
comparable to the Herbaceous Rangeland category no positive correlation 
can be made at present. Shrub and Brush Rangeland in arid and semiarid 
environments is characterized by xerophytic vegetative types. In Study 
Area II rangeland distinct from improved pasture and natural grasses was 
observable but the exact composition was not determined. In the Northeast, 
brushland typically consists of former cropland or pastureland which are 
in the process of regrowth and may be grazed by livestock. Given the 
complex field size, shape, and location parameters in this area and the 
resolution of the radar system it was not possible to distinguish pasture 
from rangeland activities. Ground checking proved that the Mixed Range- 
land category did exist in Study Area II, but it was not visible as a 
distinct category on the radar imagery. However, from observations in 
both study areas it is suggested that two if not all three of the 
"official" rangeland categories extant in Study Area II could be 
classified if interpreted by vegetation specialists. The complexity of 
land use and related environmental factors excludes a similar optimistic 
outlook for Study Area I. 
The three categories of Forest Land (Deciduous, Evergreen, and 
Mixed) were present in both study areas but reliably detectable only in 
Study Area II. Here, stands were more extensive in area and identifica- 
tion was aided by topographic, soil, and climatic inferences (Figure 8). 
In the Northeast most forest was mixed and identification of specific 
evergreen and deciduous stands more suspect (Figure 6). 
The unique return of Water on radar imagery made positive identifi- 
cation quite easy in both study areas. The only limitations were posed 
by radar shadow and the resolution of the system for small water bodies. 
However, the contrast with the surrounding terrain did facilitate 
detection in Study Area II. 
. As stated previously, Non-Forested Wetlands were readily detected 
in the West and Midwest owing primarily to the contrast with adjacent 
land cover. In the Northeast this contrast was absent and detection much 
more arduous. Forested Wetlands were extant only in Study Area I. When 
concealed by the vegetation canopy this land use category was not visible 
on the radar imagery.