ground coverage or perhaps the use of large, i.e. low resolution, detectors so that the angular sub- 
tense imaged per scan along the aircraft track is large. In either case, however, the limiting 
performance factor is the across track coverage at low altitude. 
For example, if we assumed a sensor had a 120° across track field-of-view, then at an altitude of 
200 ft the across track coverage would be 693 ft. This coverage may be adequate against a known point 
target such as a radar site, but it would certainly not be adequate for location of the mobile targets 
assumed in this discussion. 
An alternative would be to image in a side oblique mode. For the same altitude the coverage would 
be increased as a function of the offset angle, and the apparent V /H would be proportionally decreased. 
This would allow the use of higher resolution detectors, which is in fact desirable for detection at 
an increased slant range to the target. The penalty imposed on anoblique line scanning system, however, 
is substantial distortion in the imaging geometry. 
If we assume a system whose across track field-of-view is offset to the side by some angle à, then 
the across track coverage is given by: 
C = h tan (FOV + a)-h tan à 
ACT 
where h is the altitude, FOV is the angular field-of-view of the sensor and a is the offset angle. 
The across track coverage for an offset angle of 28 degrees and an assumed field-of-view of 60 
degrees at an altitude of 200 ft is 
200 tan (60 4 28)? - 200 tan 28° 
CACT 
5727 — 106 
5621 feet 
or slightly over a mile.  Presumably this mode of imaging would be superior for reconnaissance of 
mobile targets. There are, however, other external environmental considerations which limit the 
oblique imaging advantage. 
The most obvious of these is masking of the target by the terrain. This can be countered by proper 
selection of the aircraft track; imaging from a higher altitude, or a combination of both. 
A more important consideration, however, is that long slant ranges to the target require higher 
sensor resolution so that useful imagery be generated. . The increase in sensor resolution,- however ,-— 
implies a corresponding decrease in thermal sensitivity because the thermal -resolution of an infrared 
detector (Noise Equivalent Temperature Difference NETD) is a function of the square root of detector 
size; all other parameters of the optical system being held constant. This is of particular. importance --- 
because the longer the slant range to the target, the more severe is the atmospheric attenuation of the 
infrared signature generated by the target. In adverse weather situations, such as rain, snow, or fog, 
the atmospheric attenuation of infrared energy is substantial. For example, the approximate average 
percent transmission of 3-5 yum and 8-12.5 um infrared energy through water vapor per km, (the major 
atmospheric loss component) is given below. 
  
  
E 
AX Tom | Tom | ‘50 mm | ho m] 200m 
3-5 um | 0.696 | 0.599 0.48 0.41 0.36 
8-12.5 um. 0.55 4 0.71 0.46 | 0.23 0.06 
  
  
  
  
  
  
  
  
The amount of energy radiated by man-made objects in the 8-12.5 um band is substantially higher 
than in the 3-5 um band. This offsets the more rapid attenuation in the 8-12.5 um band for ranges of 
up to about 5 miles. For example, assuming 25°C air at 70% relative humidity, the inband radiance 
contrast for 8-12 um is substantially higher than that for the 3-5 um when imaging a 305 K tank against 
a 300 K background. 
77