- 1554 -
of forest expanses. Trees may have lined a field border or transporta-
tion artery but in the process enhanced that feature's detectability.
In the Northeast the pervading forest cover virtually inundated the
entire landscape. There was simply a reduction in observable contrast
on the imagery and the interpreter had to rely on increasingly more
subtle nuances. The importance of vegetation and the role of environ-
ment per se on the visibility of land use pattern elements is summarized
in Table II.
Comparability of Land Use Classification
Effects of environmental modulation were also examined employing
Level I and Level II categories outlined by Anderson, et.al. 1976
(Table I). Seven Level I categories were present in at least one of the
study areas. (1) Urban or Built-up Land was detectable in both study
areas but boundaries were more difficult to delimit in the Northeast owing
to masking by vegetation. Small towns and villages were also more easily
and consistently visible in the West and Midwest while similar settlements
in the Northeast were not. (2) Agricultural Land was readily identifiable
in Study Area II and could be delimited in Study Area I but with more
difficulty. As described above, the latter area contained smaller,
cluttered, and irregular shaped fields interspersed with forest.
(3) Rangeland was extant only in Study Area II and identification proved
facile. (4) Forest Land, although of different composition, was found
in both study areas and easily identified by its unique tone and texture
contrasts. (5) The black, no-return tone of water on radar imagery posed
no identification problems with one minor exception. Care and judgement
had to be exercised in delimiting some rivers in the Northeast to avoid
confusion with radar shadow in steep-sloped river valleys. (6) Wetland
was more easily identified in Study Area II owing to the absence of
extensive forest vegetation and better contrast with the surrounding
terrain. In both study areas, accuracy diminished as size approached
the resolution of the imagery. (7) Barren Land per se was extant in
Study Area II and in both study areas in the form of mining activities.
A more specific discussion of this category is found below.
The complexity of urban land use in combination with the rather
coarse resolution of K-band radar imagery made consistent and precise
identification at Level II extremely difficult and at times impossible.
Residential Land could be identified in Study Area II and differentiated
from downtown core activities. However, there were no large urban
concentrations precluding analysis of Commercial and Services, Industrial,
Mixed Urban, and Other Urban and Built-up Land. Only major transportation
arterials were detectable. In Study Area I instances of Residential Land
could be identified but not the extent. This was true of large as well
as small urban areas. The presence of central business district and
major commercial and services activities along arterials were manifest
as light gray to white toned areas, but detection of all other urban
functions including transportation, was dependent on the amount of
forest vegetation. Regardless of the size of the city it was not possible
to consistently detect and delimit urban land use categories in the
Northeast.
At Level II, Agricultural Land is divided into four categories:
Cropland and Pasture; Orchards, Groves, Vineyards, Nurseries, and
Ornamen
Agricul
identif
Area I
example
forest
second
the Nor
While a
vegetat
detecta
In
divided
clues a
of farm
compara
can be
environ
Area Il
observa
brushla
in the
comple»
resolut
from re
land c:
distin
both s!
"offic:
classi:
land u:
outlool
TI
Mixed)
Study .
tion w
In the
evergr
T
cation
by rad
Howeve
detect
A
in the
land c
more a
concea
on the