Merging Ocean Color Data From Multiple Missions

In this paper, the background of data merging and the methods applied to ocean color remote sensing are reviewed. The technical specifications, algorithms and products of SeaWiFS and MODIS sensors are also introduced.The comparison of the global availabilities of the chl-a data between SeaWiFS and MODIS in 2001 has been carried out, and the feasibility and necessity of merging these two sensors data is also discussed. Losses of data near the solar declination as a result of the change of the tilt of SeaWiFS can be compensated by MODIS; conversely, mid-latitude ocean coverage losses of MODIS due to sun glint can also be compensated by SeaWiFS. In addition, the difference between their orbits, and that the swath of MODIS is broader than that of SeaWiFS, result in the coverage of merged data is broader than that of individual satellite data.The 9 km SeaWiFS chl-a products and 4km MODIS products have been merged using wavelet transform and the results show: 1) the oceanic coverage of the merged data is increased by 7 percent in comparison with that of each satellite sensor data; 2) the marine features of high resolution are well maintained; 3) Comparison of chl-a from SeaWiFS, MODIS, merged and in situ data show the bias and standard deviation of merged data against in-situ data is 0.17 and 0.38 respectively(The bias and standard deviation of SeaWiFS against in-situ is 0.34 and 0.50,The bias and standard deviation of MODIS against in-situ is 0.16 and 0.40). The merged data keep better accuracy.