Monitoring water quality of Lake Taihu from HJ-CCD data using empirical models

HJ-CCD data has a great potential in monitoring water quality of inland waters. To test the HJ-CCD data application results of inland water quality monitoring, we carried out a sampling experiment in Lake Taihu. When HJ-1 satellite passing, we obtained synchronously 30 samples of remote sensing reflectance spectra, the concentration of total suspended matter (TSM), transparency and turbidity data. Regression analyzing with 27 sampling stations of water quality parameter data and the water remote sensing reflectance obtained in Lake Taihu, we established the empirical retrieval models of concentrations of TSM, transparency and turbidity. Based on these models we obtained the TSM, transparency and turbidity distribution maps in Lake Taihu using the HJ-CCD image. Finally testing retrieval results with the reserved three sampling stations of the water surface quality parameters, the average relative error is less than 0.2. This result indicates the retrieval accuracy of inland water quality parameters from HJ-CCD data satisfied basically the application requirements.