865 nm Band for Atmospheric- 
Correction Algorithms. In the currently used 
atmospheric-correction algorithms, it is necessary 
to assume that the ratio of aerosol radiance at 
two wavelengths is independent of position 
(except over waters where the chlorophyll con- 
centration is less than 0.25 ug/l). However, spa- 
tial variations in the type of aerosol will induce 
spatial variations in the aerosol radiance ratio. If 
there were a band at 865 nm (where there will be 
virtually no radiance at the sensor) in addition to 
the band at 765 nm, variations in the aerosol ra- 
diance ratio could be detected by checking for 
spatial variations in the aerosol radiance ratios 
from the 665 and 765 nm bands and from the 
665 and 865 nm bands. The addition of the 865 
nm band would also improve atmospheric cor- 
rection in coastal waters where the total radi- 
ance at the sensor from the 665 nm band is not 
Zero. 
Since atmospheric correction is so vital to the 
accuracy of pigment-concentration estimates, 
this band was selected as the sixth VNIR band for 
the SeaWiFS instrument. 
Bands in the Mid- and 
Long-Wavelength Infrared 
To remain within cost and weight constraints, 
the SeaWiFS design can support only two bands 
in the mid- or long-wavelength regions. Since 
pairs of bands in either region are required for 
sea-surface temperature algorithms, the discus- 
sion focused on a pair of bands in the 3.5 to 
4.0 um region or a pair of bands in the 10.5 to 
12.5 um region. A pair of bands in the long- 
wavelength region would provide continuity with 
the substantial use of AVHRR data from this re- 
gion, separately and in conjunction with CZCS 
data, and would provide these data from the 
same platform. On the other hand, a pair of 
bands in the mid-wavelength region would pro- 
vide a capability for determining sea-surface tem- 
peratures between +30° latitudes where the 
long-wavelength data is inadequate and would 
SYNTHESIS OF REQUIREMENTS 
improve the accuracy of the measurements 
since transmission in this region is greater than 
in the long-wavelength infrared. The panel 
members supporting the mid-wavelength posi- 
tion assume that AVHRR data will continue to be 
made available for measurements in the 10.5 to 
12.5 um region. Additional rationale for both lo- 
cations is presented below. 
The Commercial and Operational Users Pan- 
el recommended the long-wavelength bands 
based on their need to make continued use of 
the algorithms now used to interpret the data 
from the NOAA AVHRR sensors. The Research 
Panel members pointed out that AVHRR-derived 
estimates of sea-surface temperature in con- 
junction with CZCS-derived estimates of pigment 
concentration have been useful in several stud- 
ies of mesoscale processes (e.g., Brown et al. 
1985 and Abbott and Zion 1985). However, 
since the CZCS and AVHRR measurements are 
not simultaneous, constraints are imposed on 
certain studies due to the movement of clouds 
and other ocean features between the satellites’ 
overpasses. Also, scientists must access two 
separate data archives. Since sea-surface tem- 
perature can be used as an indicator of physical 
processes, simultaneous measurements might 
improve estimates of productivity. Finally, simul- 
taneously acquired information on sea-surface 
temperature and the diffuse attenuation coeffi- 
cient (closely related to the pigment concentra- 
tion) may be useful in studying mixed-layer dy- 
namics and surface-transport mechanisms. 
The Research Panel members recognized 
the value of acquiring sea-surface temperature 
data in the tropical latitudes, but pointed out that, 
to obtain precise estimates, these measurements 
would have to be made in the nighttime segment 
of each orbit. This is because daylight-segment 
sea-surface data in the mid-wavelength region 
would need to be corrected for the atmospheric 
backscatter of the solar mid-infrared by aero- 
sols, and this correction would be difficult to 
37 
