We have used digital Landsat data to create land use 
maps for several siting studies. Software was obtained 
from Pennsylvanis State University (*ORSER package) and 
modified to run on a commercial computer system. With ten 
major land use categories, identification was generally 
better than 90$ accurate. At the highest resolution 
(v56 x 79 m = 1 alphanumeric) a 1/24,000 scale alphanumeric 
map can be produced (see Figure 5). 
  
Visual interpretation from aerial photography along with 
field verification is employed to delineate sub-areas used 
in training the computer. Signatures correlating to the 
land use categories represented by these sub-areas are then 
applied to the Landsat digital data for the entire area of 
interest. 
Data Management 
To synthesize the large volume of remote sensing data 
along with other data compiled on any given project, a Geo- 
graphic Based Information Management System is frequently 
used. This interactive computerized system provides flex- 
ibility in assigning values or weights to the available data 
base. The alphanumeric printouts, or color coded analog 
displays assist in the decision-making process for site se- 
lection. 
Example - Fulton Power Station 
Figures 1 and 2 show a portion of the 200 mile radius 
investigated for the Fulton Generating Station of the Phila- 
delphia Electric Company, State of Pennsylvania, eastern 
coast, U.S.A. The first step in the investigation was to 
identify all those features which possibly were faults or 
were fault controlled. Actually at this stage of the investi- 
gation such an assignment i.e. fault or fault controlled, 
was only possible in a few cases. In these cases clear 
evidence of structural discontinuity could be observed i.e. 
topographic or stratigraphic offset or alignment. More often 
the control of a particular feature was obscure. This was 
especially true for those features that were tonal discon- 
tinuities on the Landsat imagery. 
  
*Office of Remote Sensing of Earth Resources