73
areas which
amount of
¡erpretation
led by range
ictive means
mndaries
i most generalized
mvironment when
stratification
Ling procedures
Lgh altitude
for many
:s of resource
i two to five
;ography.
ike generaliza-
lieve a higher
jtset.
igh manual
grazing,
tfet and dry
formation; for
the sensitive
gh other sources,
meadow types
i and the date
>rpes (according
regions
tion. The
arvoirs which
ii had dried
^h resolution
identifying
ces the mapping
ethod in large,
study area and
le alternative;
napping cells.
perpendicular
ement of soil
a within each
properties
(7.5 X 15km,
r one plot per
ate due to
in the study area.
Patterns evident on both the ERTS imagery and large scale photography were
similar in all phases of the mapping effort. Evaluation of the small scale
aerial photography (1:120,000) proved to be of little value since more detailed
information could be gained on the relatively larger scale photography (1:30,000).
The use of an ERTS image, aerial photographs (1/30,000) and ground data in
a multistage analysis proved to be a valuable combination for regional inventory
of the soil resource.
Land units most easily interpreted were the units consisting of shallow,
eroded soils of upland plateaus with Juniper-low sagebrush community, and those
with deep, uneroded soils of upland basins with dense Rabbitbrush community.
Interpreters had the most difficulty in distinguishing sandy, dune soils with
dense Big sagebrush-Bitterbrush vegetation type from the silty, lacustrine
basin soils with low density Rabbitbrush or Greasewood (Sarcobatus vermiculatus).
Land units with high surface rock density were easily identified due to the
homogeneous dark gray tones on the imagery. Occasionally, however, this rock
cover transgresses adjacent land units making it difficult for interpreters to
map the unit boundary accurately.
When mapping on a regional level, soils must be evaluated and examined
with a certain degree of bias because they vary within and between vegetation-
terrain types. Previously, the only certain way to detect and document this
variability was through highly detailed survey procedures. When surveying
on a regional level, however, detailed examination based on conventional
techniques succeeds in only expending valuable time, money, and personnel. What
is desired for reconnaissance level mapping is the grouping of soils which will
behave similarly under certain management programs. In the past these groupings
have been made from completed, detailed surveys by combining smaller mapped
units of similar types of soils into larger units, associations. The procedures
discussed in this paper which are based on mapping land system units and
documenting the major characteristics of that land unit, eliminate the need for a
detailed survey before one can group similar types of soils. Satellite imagery
though low in resolution proved to be of sufficient resolution for grouping
similar soils, vegetation and land types by distinguishing large units and
coalescing the small inclusions which tend to confuse the interpreter on high
resolution large scale photography. By utilizing a strong multistage analysis
approach, ranging from ERTS imagery to aerial photos to strategically located
field profile plots, soil resource evaluation and mapping is quite feasible on
a regional level at a considerable savings in time, money, and personnel, with
no sacrifice of the accuracy needed for resource management objectives at this
level.
Ground data collection efforts for the project were hampered by lack
of accessibility to plot areas and resulted in a portion of the interpreted
area not being verified on the ground. This portion of the vegetation mapping
effort was classified through photo interpretation by using the convergence
of information gained in the previously examined regions of the study area and/
or through data sources present prior to the study. Time was the most critical
constraint during the project. The total time allocated and needed to complete
the mapping efforts, however, using remote sensing data products reduced the
time and costs considerably from the conventional methods of rangeland inventory
without use of photography. The total vegetation mapping effort performed on
color infrared aerial photos (1/30,000), including interpretation time,