pair 1,2.
These results can not be interpreted easily. Clearly, correlation with
the short wave channels is weak because these channels lave a low sig-
nal-to-noise ratio (Figs. 9,10). Since sample number, chlorophyll con-
centration, salinity, and turbidity are correlated (the absolute values of
the correlation coefficients are between 0. 70 and 0. 95), the measurements
probably are done across an area where fresh water mixes with sea wa-
ter, with continuous increase in salinity, and decrease in chlorophyll and
suspended matter concentration. The correlations observed between ra-
diances and sea truth measurements are compatible with that interpreta -
tion. However, a local maximum of radiance between sampling positions
13 and 21 does appear in all channels but not in the chlorophyll or sus-
pended matter measurements.
From this simple correlation analysis notling can be concluded about
spectral separability of chlororhvii and sediments through OCS data be-
cause the two parameters are correlated amoeag each other and probably
the reflectivity is determined mainly by the high sediment content.
There is a general increase of the radiance towards the coast, especially
in the red channels. Even if the sea truth records show that sediment
content is higher near the shore, one should not overlook the possibility
that a part of the measured radiation is light reflected from a bright beach
and scattered into the field of viev of the scanner. Indeed, the contrast
between land and water is strongest for the red channels.
The ship tracks c:iresponding to flight lines 1 and 3 are close to the
border ofthe scanned area, and es»ecially after tide correction, many
sampling points fall even outside thesc areas. Apparently, the frame
of the window at the bottom cf the aircraft was ir ine optical path of the
3canner for large scan angles, causing a strong attenuation of received
radiances which could not be coirccced. For these reasons, the data of
the two flight tracks have not been analyzed.
Hopefullv, analysis of the data of the 29th of June will allow to obtain
more ccnclusive results since sea truth for that day is more complete
and richer, both for amount of data and variation of measured parame-
ters.
si
te
de
Sa
ch.
sa.
se
SO!
