A geological model for describing the spectral emittances of rough- 
textured rock and mineral surfaces which accounts for birefringence, 
multiple mineral constituents and varying particle diameters is being 
developed by Vincent [16] based on earlier theories [17, 18]. In the 
case of monomineralic rocks, it theoretically accounts for birefringence 
effects, and for rocks composed of several minerals it calculates an 
effective complex refractive index. For both cases, the indices are 
substituted into equations of a model for particulate media derived 
from Mie (single particle scattering) and radiative transfer (multiple 
Scattering) theories. The resulting spectral emittances can be used to 
determine the effects of textural variations on rock-type discrimination 
techniques. 
Implicit in the application of both the 
geologic models for predictive studies is the ‘act that they can be 
calculationally combined with atmospheric radiative transfer models to 
provide estimates of atmospheric effects necessary to test the adequacy 
aircraft and satellite 
sensing under various conditions. 
Electromagnetic signals received by airborne or Spacecraft sensors 
are affected by the intervening atmosphere between the Earth's surface 
and the sensors. The atmosphere scatters and absorbs the radiation from 
the surface materials and adds extraneous radiation (path radiance) to 
the received signals through scattering and emission by particles in 
the atmospheric path. Molecular absorption and emission are the most 
important considerations for infrared and microwave radiation. However, 
for visible radiation, the principle concern of present multispectral 
studies, the light scattering properties of the atmosphere are of 
greatest importance. The scattering properties are most affected by the 
aerosols that are present. The vertical distributions of these aerosols 
can vary substantially from time to time and place to place, and cannot 
be measured as easily as those of molecular absorbers such as water vapor 
and carbon dioxide which are determinable from radiosonde data.  Further- 
more, the measurement and calculation of the effects of molecular 
absorption on a radiation signal is somewhat simpler than performing 
similar work on the effects of atmospheric scattering. 
It is through the development and exercise of a radiative transfer 
model that one hopes to obtain a bettei understanding of the extent of 
atmospheric scattering effects on multispectral sensor signals and to 
use that knowledge in developing techniques for overcoming them. Such 
models have developed to the point where they can be used in calculations 
representing a wide variety of real atmospheric conditions with sufficient 
accuracy for remote sensing studies. The intent here is both to calculate 
specific corrections to individual data sets to account for atmospheric 
scattering and also to use the knowledge gained to develop feature enhancing 
preprocessing techniques more powerful in information extraction. 
The study of radiative transfer in the atmosphere is not new but, on 
the other hand, until recently the scientific community had not produced 
a simple, practical and unified computational method for use in modeling 
the real atmosphere and its transfer of radiation. Classical radiative 
transfer theory was developed many years ago and has undergone quite a 
few refinements since then. From a semi-quantitative point of view,