306
AVHRR
UOR
0.2 -r-----------,
192 track tow 13 18.6.87
R550 tow 13 18.6.87
128
J\
'vi"
r
. V
0.1
64~
o I---r--,----r-~ 0.0 +n"TTT1"TTT1"TTT1TfT1CTTT1crrrrrrr1
o
90
180
0 30 60 90 120 150 180
'--~f
0.25
192
23.6.87
R550 tow 26 28.6.87
0.20
r
128
64
0.15
0
0.10
,
I
0
45
90
0
30
60
90
0.2
track tow 29
R550 tow 29 1.7.87
192
.
23.6.87
1ft
128
IJ:.
0.1i>
64
\+~
N"
~
0.0
,
,
75
150
0
30 60 90 120 150
distance (km) approx.
distance (km)
Figure 3. Comparison of along-track AVHRR radiances (arbitrary scale) and UOR reflectance at 550nm for
three tows in the NE Atlantic, June 1987 (from Aiken and Bellan, 1989).
In general, increased backscatter by non-absorbing particles reduces the ratio of upwelling
light at the wavebands used to determine chlorophyll (blue to yellow or green to yellow for
the CZCS), so that phytoplankton biomass will tend to be underestimated. Bricaud and Morel
(1986) measured high chlorophyll-specific scattering coefficients for suspensions of E. huxleyi
cells compared to those for other species of phytoplankton, but the ratio of backscatter to total
scatter was too low to explain the high satellite reflectances for natural blooms, supporting the
earlier suggestion that the detached coccoliths cause most of the backscatter. Gordon et al.,
(1988) developed a model for predicting the upwelled radiance at the sea surface as a function
of light absorption and backscatter by phytoplankton pigments and coccoliths. However,
validation of the model was not possible without values for the spectral backscattering
properties of the coccoliths and other particulate matter in the water.