Ocean Chlorophyll-a Derived from Satellite Data with Its Application to Red Tide Monitoring

The signal-to-noise ratio of ocean color channels of MODIS sensor is higher than SeaWiFS sensor and the bandwidth is narrower than SeaWiFS, thus MODIS data are more suitable for the retrieval of chlorophyll-a in ocean. Many standard empirical algorithms and semi-analytical algorithms are supplied to calculate ocean chlorophyll-a by the SeaDAS software which is distributed from NASA. Considering the characteristic of MODIS sensor, the two standard empirical algorithms SeaDAS OC2 and OC3, two semi-analytical algorithms SeaDAS Clark and NSMC-CASE2 are used to calculate ocean chlorophyll-a based on MODIS data, the correlation of MODIS chlorophyll-a and ten stations measured chlorophyll-a is analyzed in Fujian coastal red tide monitoring region in 2004. The statistic result of MODIS chlorophyll-a and ten stations measured chlorophyll-a shows that the correlation index of OC2 standard empirical algorithms is 0.787, and OC3 is 0.839, and Clark semi-analytical algorithms is 0.510, and CASE2 is only 0.133. Although the ocean chlorophyll-a calculated by SeaDAS based on MODIS data is high compared with the chlorophyll-a measured in locale, the chlorophy ll-a variational curve of SeaDAS OC2 and OC3 and Clark is similar to the chlorophyll-a variational curve measured in locale. However, the chlorophyll-a counted by CASE2 semi-analytical algorithms is much higher than the chlorophyll-a measured in locale, and the chlorophyll-a variational curve of CASE2 is very much different to the chlorophyll-a variational curve measured in locale. The red tide occurring in Fujian coastal region at one time with the MODIS chlorophyll-a during 2002—2005 is analyzed also, and the possibility of monitoring red tide based on 250 m and 500 m resolution ratio MODIS visible light data is discussed. The ratio of MODIS red (250 m)and green (500 m) channel obviously changes in the whole course of red tide disaster, the ratio is much higher when the red tide disaster occurs than before and after the red tide disaster occurs. The results show that there is inconsistence high estimate for chlorophyll-a derived from two standard empirical algorithms and two semi-analytical algorithms. On the opposite, SeaDAS OC3 standard empirical algorithms are suitable to calculate ocean chlorophylla in Fujian coastal region based on MODIS data. On the basis of accumulating the chlorophyll-a data measured in locale, the relevant inversion mode for quantitative monitor of remote sensing data of chlorophyll-a on Fujian coastal water will be established with the 250 m and 500 m resolution ratio of MODIS visible light data, and atmospheric correction process will be carried out in MODIS data, thus it is possible that the change of chlorophyll-a on Fujian coastal water will be accurately predicted .