MAPPING THE CHEMICAL AND MINERAL PROPERTIES OF TOTAL SUSPENDED MATTER IN PEARL RIVER WATER BY MULTISPECTRAL OPTICAL REMOTE SENSING

In the past decades, optical remote sensing data were widely used to retrieve water quality parameters including total suspended matter (TSM) and Chlorophyll a (Chla). However, the chemical and mineral properties of TSM are not detected and characterized. In this study, feature-orientated principal component analysis based on Crosta method using multispectral optical remote sensing data was conducted. The results show that this technique can retrieve not only routine parameters including TSM and Chla, but also the chemical and mineral properties of TSM such as Fe-bearing level and clay/carbonate mineral variations. The mapping results in Pearl River and inland lake/reservoir show good agreements with the observation and measurement. Combined the analysis of chemical and mineral correlation of catchment soil/rock outcrops, the source of TSM and Chla can be tracked. And it also implies that the water quality such as TSM and Chla as well as their chemical and mineral properties can be retrieved from the long-term historical image data of multispectral optical remote sensing, which can provide useful information for regional water pollution prevention and environmental protection.