or -20°. The continuous 360° scan allows refer- 
ence sources to be viewed during the non-scene- 
viewing part of the scan, as well as allowing a 
deep space view for a zero reference just before 
the scan begins. There is also a solar diffuser that 
can be inserted into the field of view of the sensor 
for calibration against the input solar radiance. 
Expected Sensor Performance 
Estimates of three key performance parame- 
ters were made: radiometric performance, modu- 
lation transfer function (MTF), and polarization 
sensitivity. The expected radiometric perfor- 
mance, shown in Table A-2, meets the perfor- 
mance goals established by the user groups. 
The MTF was calculated at the Nyquist angu- 
lar frequency of 0.313 cy/mrad. The results indi- 
cate that the VNIR MTF will be 0.36 in the scan di- 
rection and 0.57 in the track direction; TIR MTF 
will be 0.34 and 0.53 in the scan and track direc- 
tions, respectively. These values correlate well 
with the nominal rule for sensor design that the 
MTF should equal or exceed 0.3 at the Nyquist 
frequency; again, there is performance margin rel- 
ative to the user requirements. 
For the SeaWiFS, the polarization sensitivity 
depends on the scan angle, since the major con- 
tributor is the half-angle mirror. Figure A-2 shows 
the estimated polarization sensitivity of the Sea- 
WIFS as a function of scan angle at a wavelength 
of 443 nm and that it always remains within the 2% 
desired limit. Moreover, the 443 nm data repre- 
sent the worst case; polarization effects are small- 
er in all other bands. 
Table A-2. SeaWiFS Estimated Radiometric Performance 
Saturation 
Radiance! 
(nW/cm2-sr-um) 
Signal 
1 
2 
3 
4 
5 
6 
if 
8 
1 
2 At 300K. 
Vili 
Radiance! 
(nW/cm2 -sr-um) 
Noise-Equivalent 
Temperature 
Difference 2 (K) 
Signal-to-Noise 
Ratio 
H. R. Gordon 1987: personal communication. 
