= 1134 =
permit analysis and plot selection from the area covered by the photos.
Total study area is about 6 million acres. Photo coverage is about
960,000 acres. Cost of photography to the project is $5,000.
Large-Scale Photography
Experience in the field has shown that observations and analyses of
various types of surface resources and vegetative components from the
ground level are generally limited to a relatively short distance from
the observer. Even with intensive field work it is difficult to pinpoint
a precise location on the map and the LANDSAT imagery or to obtain a
comprehensive and accurate description of the area covered by a 1l.l-acre
(pixel) signature. Variations in vegetative components; changes in
slope, aspect, and soil types; and stresses of low nutrients, poor water
supply, or drought are difficult to assess on an acre basis from a ground
perspective. Consequently Wear and Wheeler believed that elevating the
perspective by large-scale photography would enhance the comprehensive
overview of the area, increase the accuracy of the observations, and
pinpoint the location of each area of interest. The large-scale photo-
graphy, covering selected plots, would also provide greater cost-
effectiveness. Ground surveys alone would require more intensive and
extensive inspection techniques at greater total cost to obtain compar-
able "ground-truth" of individual pixels.
Decisions for scale of photography to be used were based primarily on the
relative size of the vegetative elements to be evaluated. Agricultural
plants are generally smallest, range shrubs next largest, and forest
trees largest. Selected scales were 1:3,000, 1:6,000, and 1:8,000,
respectively.
A series of photo strips of variable length were selected on the LANDSAT
composite 4x photo enlargements in the three vegetative classes
throughout the survey area. A wide variety of sample sites were selected
in each class to provide a broad base for identification and comparison.
Selected strips were transferred to the 1:62,500 USGS contour maps for
