10 
"High-Stage“ Aerial Photography and ERTS Data-Based Systems 
High-Stage Aerial Photography : A resource survey using (for the first 
time in Montana) 1:80,000 scale quad-centered peak-of-green CIR aerial photo 
graphy as a data base is progressing on schedule. The firm accomplishing the 
interpretation and mapping contract has expressed its satisfaction with the 
ability of this film/season/scale medium to provide the necessary resource 
classifications. In some cases, for example, surprisingly small vegetation 
communities and/or associations are identifiable by species or species group. 
ERTS Data Applications : Results obtained with broad resource feature 
mapping with ERTS data(color-combined transparencies projected onto a rear 
projection screen at a scale of 1:125,000) on the BLM test site north of the 
Missouri River had been fairly successful (Meyer and Gerbig, 1974). Five 
specific surface resource categories had been quite accurately identified: 
grassland, grassland/sagebrush, forest/woodland, bare shale slopes associated 
with sagebrush/greasewood coulee bottoms and water. 
Application of the same techniques to the Decker-Birney test site in 
volved preliminary examination and testing of all available ERTS bulk imag 
ery coverage -- including variations between dates of coverage, single bands 
(4,5,6, 7) and various band/color/season combinations. As a result of these 
tests, the May 19th Band 5 (green filter) and Band 7 (red filter) combination 
was selected, projected onto the rear projection screen at a scale of 1:125, 
000, and mapped by an experienced photo interpreter on the basis of 7 visible 
image color differentials: red, pink, reddish-brown, tan, light tan, white, 
blue. These 7 categories were then compared singly, and in various combina 
tions, with the 5 previously-mapped ground truth categories (Pine-Woodland, 
Ponderosa Pine, Grassland, Broadleaf (streamside) & Agriculture, Sagebrush- 
Shrub) . 
Those ERTS-mapped color differentials (or combinations thereof) which cor 
related best with specific ground truth categories are illustrated in Figure 1. 
It will be seen that the ERTS-generated reds and pinks corresponded best to 
the Broadleaf-Agriculture category; tan corresponded to the combined Grassland/ 
Sagebrush-Shrub categories; and reddish brown corresponded to the combined Pine 
Wood 1and/Ponderosa Pine categories. 
For an overlook at the general area and configuration of these vegetation 
types in large areas, this could be of some use as a first-stage over-view at 
the start of a resource survey. But one must raise practical questions as to 
its repeatability -- i.e., given the same or a similar area one or more years 
from now, at a slightly different date with respect to vegetation peak-of-green 
and/or at a different moisture level (and possible plant stage and vigor level) 
either during the year under study or in previous years, etc. — would the 
differentials provided by color combination reflect actual ground vegetation 
variations to the same degree?