The Study Of Cloud Classification In The Northwestern Pacific Ocean Using The Data Of FY-2C

Clouds are closely related with people's life. It has great influence on weather and climate's changes, it should be concerned when predicting weather and climate, and it's also playing an important role in transporting water vapor and energy on earth. As different clouds act differently, to do research on cloud classification is of great meanings. Upon oceans, there's not much routine research tools, satellite is an important way. So this thesis, using the data from FY-2C, researched on the classification of clouds in the Northwestern Pacific Ocean.The FY-2C satellite data at channels of infrared split windows and water vapor were used. And research area is the ocean of 7°N～42°N, 105°E～140°E. First, in data preprocessing, the Earth location of satellite data is particular discussed. then, the clouds were classified above the ocean. This is a primary work in the thesis. In the method of cloud classification, the Nephanalysis and dynamic cluster classification were used. And the conclusions are:In the Earth location, this thesis researched on the FY-2C satellite's earth location, using the Earth location algorithm of the Japanese geostationary meteorological satellites. As latitude and longitude is given, we calculated the location of the ranks of satellite image pixel, and by calculating 31 points from the sea (these points were given in simplified block grid intersection spacing of 5 degrees latitude and longitude), the conclusion is that the row calculation error is 0.125, the column calculation error is 0.122. the navigation error less than 0.13 scan line which is enough for the cloud classification analysis. In the cloud classification, the Nephanalysis uses the data of infrared split windows channel and water vapor channel and confirmed different cloud models or clear, using the threshold method. And then this thesis further on talked about details of each model. Eventually, the clouds were divided into Clear, St, Cu, Cg, Ci +Cu, Ci+Cm, Cm, Ci, Dense and Cb. In order to improve the spatial resolution, this thesis classified every pixel point, using dynamic cluster analysis. And the clouds were divided into Ocean, Sc-Ac, Mixed, As-Ns, Cs, Dense and Cb.The results of two cloud classification methods were compared to FY-2C NOM. We found: the cloud types of the Nephanalysis is more than FY-2C NOM. And the middle/low cloud class is verified. The result of dynamic cluster classification was better than the result of FY-2C NOM. And the result was consistent with the actual situation.The thesis of the cloud classification has some referenced value for the study of marine weather and typhoon's dissipation process. And some study (such as the cloud mask in the SST inversion) can be comparison information.