Preliminary research of water classification from TM images in Yangtze estuary

Classification of water body helps to understand relationships between different properties inside a certain class and quantify variations between different classes. The paper concentrates on studying classification of water in Yangtze estuary based on in site investigations of spectral reflectance properties and water color parameters. By our researching, we found eight optical classes of water bodies. We put forward an algorithm of reflectance spectrum slope in order to distinguish different spectrum reflectance types. The algorithms applied to water body classification of a TM image. From TM image, we detected six water classes. Spatial distributions of different water classes are consistent with the facts. The results of study are beneficial to monitor dynamic distributions of different water classes and provide a basis for regional algorithms to retrieve biogeochemical parameters concentrations.