Remote Sensing Detection And Diurnal Variation Research Of Green Macro-algae Bloom By Geostationary Ocean Color Imager

Green macro-algae bloom (GMB), which outbreaked in the East China Sea andYellow Sea from2007to2013, had caused serious impact on the marine environment,fisheries and ecosystem of China. Nowadays satellite remote sensing had played animportant role in GMB monitoring, but the diurnal change and movement of GMBwas still a research problem as a result of the coarse temporal resolution of traditionalpolar-orbiting satellites. The occurence of Geostationary Ocean Color Imager (GOCI)provided a reliable weapon for us to solve this problem, but GMB informationextraction technology for GOCI still need futher development. The macroalgaespecies identification technology for GOCI has not yet been developed, and thediurnal movement characteristics of GMB in the East China Sea and Yellow Sea hasnot yet been analyzed systematically.To solve the above problems, the GMB information extraction technology, algaespecies identification technology and diurnal movement characteristics of GMB hadbeen analyzed in this paper. The main contents and conclusions are as follews:(1) GMB information extracting technology for GOCIThe GMB detection capabilities of6major GMB remote sensing algorithm(NDVI, RVI, EVI, FAI, NDAI, IGAG) and their different band combinations wereevaluated, taking GOCI bands as reference. The results showed that the GMBdetecting capability of NDVI was significantly better than OSABI, KOSC, RVIand EVI, which indicated that NDVI could be regarded as the preferred algorithmfor GOCI green macro-algae monitoring application. The IGAG algorithmshowed good potential, but it had significant uncertainty. When monitoring thegreen macro-algae with the6above algorithms, choosing between GOCI band 5(660nm) or6(680nm) as red band was flexible, while results between GOCIband7(745nm) or8(865) as NIR band showed noticeable variability. On this basis,the new algorithm was proposed, which utilized the information from2NIRbands and2red bands simultaneously.(2) Macoralgae species identification technology between Enteromorpha andSargassumBased on the GOCI GMB data of Enteromorpha and Sargassum, the spectraldifferences of these two species of GMB have been analyzed. The results showedthat at blue-green bands (GOCI band1to band4), the reflectance of Sargassumincrease with wavelength, while that of Enteromorpha decreases. The reflectanceof Sargassum was obviously higher than Enteromorpha at red bands (GOCI band5and6), while slightly smaller at NIR band (GOCI band7and8). On this basis,the macroalgae species identification technology by Support Vector Machine(SVM) was proposed, and the identification accuracy was84%.(3) Diurnal movement character of GMB in the East China Sea and Yellow SeaBase on the GOCI data which had8images per day, the diurnal variabilityinformation of GMB was extracted, and the driving mechanism was analysed withutilization of ocean current and wind data. The results showed that the area ofGMB increase then decrease within the time span from8:30am to15:30pm,reached maximum at1:30pm. Then the hourly drift trajectories of GMB wereextracted using maximum correlation coefficient (MCC) algorithm, and the resultsshowed that the drift of GMB in the Yellow Sea has significant changing features.