Comparison of CZCS-derived pigment concentration by standard atmospheric correction
method with sea-truth data in the waters around Japan shows an underestimate by satellite to a
differentdegree. In the range of pigment concentration from 0.2 to 6.0 mg/m3, the CZCSobser
vation is lower than in-situ data from about 50% to a factor of 6. Assuming a constant Angstrom
exponent during atmospheric correction stage has been found to account for this bias, provided
that bio-optical algorithmis valid in those waters. An improved atmospheric correctionmethod
which can estimate the Angstromexponent pixel by pixel has been developed based on Gordon
ocean color radiance model, and has been evaluated by comparison with these in-situ data. The
results demonstrate that this method can increase the accuracy of CZCS-estimated pigment con
centration to the level within the system error (30-40%) of CZCS in-water algorithm.
An information onocean surface wind speed in remote-sensed image is required for accurate
application of atmospheric correction algorithm for ocean color remote sensor. A method for
retrieval of surface wind speed with Sun glitter intensity in visible range is proposed with theaid
of probability distribution of wave facet which varies with change in wind speed. The retrieval
accuracy is estimated with the error in the retrieval of wind speed, water-leaving radiance and
phytoplankton pigment concentration obtained by atmospheric correction using retrieval wind
speed in a model atmosphere-ocean system. This method is applied to ground, vessel and air
borne radiance data and the result showed the successive correlation between retrieved wind
speed and observed one.
It is possible to elucidatea lot of chemical, physical and biological phenomena in the oceanby
obtaining the high quality and reproducibility nutrient concentration data sets. Therefore, we
need conferences to talk about calibrating and getting the common standards, replicating
samples, settingup the reference stations, correcting the sample concentrations, plotting the ver
tical profiles and so on. From now on, these checks for completeness of correction are necessary.
After that proceeding mentioned above, we would be able to get and accumulate the excellent
data sets globally.
Some resent works in Japan on the standardization of the measurement of total carbonate
(TCO2) arereviewed. Themeasurement method and thestandard areneeded to setconsistent and
traceable among each labs to keep uncertainties of the data within few μmol/kg in worldwide
scale. A experimental work to make a standard of total carbonate based on natural sea water is
introduced. The results of the time series measurement improves that the sea water standard
keeps its concentration ofTCO2 over 80 days with the variability of ± 1.4 μ mol/kg. Some ex
amples of the intercalibration works in Japanese labs show that there are complex and unique
systematic offsets between each lab’s data. Thus it is recommended that a consistent program of
the intercalibaration study for total carbonate is needed to put forward.