the scan and the track directions), the sample 
spacing along track is also taken to be 1.13 km 
at nadir. This results in a nominal square instan- 
taneous field of view (IFOV) of 1.6 mrad. 
The requirement for a proven-technology 
approach leads to a choice of moderately sized 
detectors, particularly for the thermal-infrared 
(TIR) region. Existing instruments, such as the 
VAS/VISSR, and detectors developed for several 
Santa Barbara Research Center IR&D programs 
use long-wavelength infrared (LWIR) HgCdTe 
detectors in sizes ranging from 100 to 125 um. 
Hence, the initial detector size selected was 
100 um. However, in optimizing the optics and 
mechanical-packaging design, a 20% increase in 
the nominal detector size was required, resulting 
in 122 um detectors. This fixes the optical focal 
length at 7.64 cm (for an IFOV of 1.6 mrad). 
The very low reflectivity of the ocean creates 
a requirement for high signal-to-noise perfor- 
mance in the visible and near-infrared (VNIR) re- 
gion that drives the design towards concepts 
SAMPLE 1 CENTER 
SAMPLE 2 CENTER CENTER TRACE 
SAMPLE 3 CENTER 
A ateasge 6 
a> ON = So =| NOS 
= 
x 
F 
[s) 
< 
c 
Ee 
So 
z 
° 
a 
< 
WwW 
S) 
z 
< 
Ee 
2 
a 
. 
= 
00 
-500 0 500 
DISTANCE ALONG SCAN, KM 
1000 
ee 
IMPLEMENTATION PANEL 
that use either long dwell times or time-delay in- 
tegration (TDI). The only practical method for re- 
alizing long dwell times is to reduce the required 
scan rate by using more than one detector per 
band along the track direction. For example, if 
there are two detectors along track rather than 
one, then the required scan rate for contiguous 
scan swaths would be three scans per second 
instead of the six scans per second required for 
one detector. Since the scan rate is halved, the 
detector dwell time and, therefore, the available 
signal integration time, is doubled. Unfortunately, 
the very wide scan angles of the SeaWiFS instru- 
ment cause severe bowtie distortion when multi- 
ple detectors are used along track, i.e., the off- 
axis angle and the Earth's curvature cause the 
spacings between sample centers along track to 
become larger and larger on the ground as the 
scan angle increases. This effect is shown in 
Figure 19. In order to avoid this geometric dis- 
tortion, the SeaWiFS design was constrained to 
using a single detector along track. 
2.57 KM AT EDGE OF SCAN 
y Figure 19. 
Bowtie distor- 
tion from more 
than one 
detector per 
band along 
track and wide 
scan angle. 
49 
