Optical Classification of Coastal Water Body in China using Hyperspectral Imagery CHRIS/PROBA

Abstract

Coastal saltern and aquaculture are vital components of human-made coastal areas and they have immense influence on the coastal water environment in China. Hyperspectral space-borne remote sensing is a significant technology in remote sensing, enabling in-depth identification and discrimination of the spectra from water features on shore. The study uses CHRIS/PROBA images to identify water-bodies and classify six optical water types from artificial and natural water in three sites across regions along China’s coastline. Most of the offshore waters are affected by the inflow of land source materials. The fluorescence peaks appearing at 675 nm can be observed by 5 water types. In addition to Class 1 water which is far from the shore and the Class 4 water whose signals are mainly affected by the algae at its bottom, the other 4 kinds of water show backscattering peak after 800 nm, suggesting that particles are playing an important role in coastal water. For coastal waters in Liushagang, the mangrove forest will transfer the water body signal received by the satellite into the vegetation signal; pools in coastal saltern have rather shallow water during production period when the benthic signal will seriously interfere the reflected signal. In the intertidal area, the hyperspectral characteristics in the same sample also change periodically due to the ebb and flow. The Class 2 and Class 3 water bodies alternately occur in the intertidal zone in the Lianyungang research area. Therefore, it is concluded that optical classification approaches reflect the advantages of remote sensing from Satellite-bone, and our study can be helpful and conductive for follow-up sensors.