S 
t 
\e 
  
sub-spans are used rather than the full span, the azimuth resolution is 
degraded, but the improved image is usually worth the sacrifice. 
There is one further source of image degradation we must consider, 
but to understand it we must take an entirely different viewpoint. Up 
to now, we have explained SAR through analogy with aperture arrays. We 
can also conceive of SAR as a doppler filtering process. Consider 
figure 9. When the antenna is at position x, its distance from the 
terrain point is 
2 
2 2 Ó x 
R° + ~ TI Co 
r = X R 2 R (8) 
Since the antenna moves at velocity v 
> dL al 
F 5 2 1 RY (9) 
The doppler shift of the radiation returned from the terrain point is 
thus. 
n |=» 
t X (10) 
- 
un 
of 
We see that the doppler shift is (to a good approximation) linear with 
azimuth (along track position). The position of zero doppler indicates 
the azimuth position of the target. The SAR is fully analogous to a 
chirp radar, and SAR processing is analogous to doppler filtering. 
Problems arise if the target is moving. If it moves in the 
azimuth (along track) direction, the relative velocity of antenna and 
target changes, and hence the slope of the doppler curve. The target 
will appear out of focus. 
If the target moves in the range (cross track) direction, a 
constant doppler shift is added. This shifts the position of zero 
doppler, and the target position is shifted in the azimuth direction. 
SAR is a valuable tool for remote sensing, as the other papers in 
this symposium will demonstrate. The above caveats must be born in 
mind when viewing and interpreting the SAR images you will be seeing. 
297