Study Of Remote Sensing Algorithm And The Temporal-spatial Distribution Of SSD In Bohai Sea And Yellow Sea

Sea Surface Density?SSD?is one of the important parameters of the physical properties of seawater.The accurate inversion of SSD plays an important role in global marine environment and ecological processes.Selecting Yellow and Bohai Seas as study area,combining with in situ data,based on the band setting of GOCI,an empirical remote sensing model is developed.Then application of the SSD model to satellite data leads to spatial and temporal distribution product.The main research contents are as follows:The mapping between SSS and Rrs is established based on the in situ data sets of several voyages and different band combination forms are trained to choose the best band combination,then two inversion forms are established,including taking SSS as the intermediate quantity and directly establishing the empirical algorithm of SSD.The R² of verification samples is 0.72with MAPE=3.27%,RMSE=0.87(kg m^-3).The error sensitivity analysis of the model showed that the fluctuation of MAPE variation was less than 3%,and the model was not sensitive to error,which proved that the model was stable.Applying this algorithm for GOCI satellite data and long-term spatial and temporal distribution of SSD was completed.The results indicated a significant fluctuation in SSD variations.The highest values of SSD were observed in the Bohai Sea,middle values in the North Yellow Sea,and the lowest values in the South Yellow Sea,the basic trend is to increase with latitude in;Obviously seasonal variations were observed with a high SSD values existing in summer,while relatively low values in winter.In spring and fall,the values are the transition interval.The eight-year average results show that SSD has a certain interannual change,but it is not obvious.Qualitative discussion on remote sensing application based on the optical characteristics of SSD and the relation with optical factor?backscatter of seawater?.The results show that the optical characteristics of SSD are closely related to the properties of sea water in the study area and remote sensing reflectance can be used to invert SST in secondary water bodies with complex optical characteristics and more substances in seawater.According to the correlation coefficient of stepwise regression,all areas can be divided into3 different kinds.By using the in situ data of environmental state parameters,SST and SSS,correlation analysis and stepwise regression analysis were carried out and the results showed that SSS contributes more to SSD changes.The correlation between SSS and SSD in the feature region was analyzed with long time series data.The results showed that there are significant temporal and spatial differences in the correlation between them in the feature region.In terms of temporal,the correlation is the highest in autumn,and in terms of spatial,the correlation between the south yellow sea and the north yellow sea is the higher.