- 1548 - 
The paramount importance and limitation imposed by environmental 
modulation is generally acknowledged. Formalized as the environmental 
modulation transfer function (EMTF) it may be expressed by the formula: 
E = f (Cp,Sp Tg Ag,. + + Xp) 
where Cp = number of distinct land use or mapping categories (the more 
detailed the classification the more complex an environment 
will be), 
Sp = size and spatial arrangement of categories comprising an 
area, 
Tp = a measure of the time-variable nature of the environment, 
Ag = a measure of atmospheric constraints on sensing, 
Xp = other aspects of the environment such as topography, 
technology of a society, settlement history, climate, and 
soils (after Everett and Simonett, 1976). 
Simonett and Coiner (1971) found that the level of complexity varied 
considerably within the United States as well as overseas. Similarly, 
Schwarz and Gaydos (1975) have reported on the physical and cultural 
environment as determining factors in regional interpretability varia- 
tions of land use using satellite data in visual and digital forms. In 
general, however, surprisingly little research has been devoted specifi- 
cally to the EMTF. It may be alluded to obliquely as a limiting factor 
or referred to in theoretical terms, but unfortunately it is too often 
slighted in the discussion of research findings. This study is an 
empirical examination of two broad, diverse environments of the United 
States and a discussion of the interrelationship between radar system 
parameters and environmental modulation for small scale land use mapping. 
It is an attempt to mitigate the perceived void in EMIF research. 
Methodology and Study Area 
Two extensive strips of K-band Westinghouse radar imagery of 
diverse environments in the United States served as study areas for this 
investigation. One traversed over 1392 km. of the northeastern United 
States at a scale of 1:225,000 (Study Area I) while the other covered 
approximately 2400 km. of the Central and Western portions of the country 
at a scale of 1:180,000 (Study Area II) (Figure A). Owing to problems 
with gain control only the HH polarization was employed in analysis. 
Physically, the Northeast study area contained segments of steep 
slopes, dissected plateaus and plains, level drift plains, glacial 
valleys, rolling hills, mountains, nearly level lowlands, lakes, rivers, 
and swamps. Only four major land uses were present (cropland, pasture, 
forest, and urban),,but the combinations and relative percentages varied 
considerably forging a collage of land use patterns. Agricultural 
activities, generally in the form of small land holdings, included 
dairying, truck farming, and orchards. Feed and forage crops were almost 
    
   
  
   
  
   
  
   
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
   
  
  
  
  
  
   
   
  
  
  
  
  
  
   
  
  
  
  
  
  
   
never i 
land. 
woods , 
mented 
lumber: 
small, 
mounta. 
and coi 
southe: 
T 
contin 
From e; 
plains 
narrow 
plains 
walls; 
stream 
dissec 
mounta 
use wa 
reflec 
the ea 
devote 
of woo 
corn a 
woodla 
compri 
consis 
and ca 
were d 
presen 
and sh 
and sh 
drylan 
of the 
in fre 
these 
what m 
precis 
from m 
grazin 
U 
Anders 
consis 
as to 
parame 
patter 
were a 
detect 
themat 
radar