The Study And Application Of Remote Sensing Techniques For The Distribution Information Extraction Of Tidal Flat Resources

Coast zone is zone of transition between terrestrial environments and marine environments where terrestrial factor and marine factor are co-action. Now, coast zone has been becoming an important area for the international competition and exploitation owing predominant geographic location and abundant resources itself. Intertidal zone, located between MSHTL and MSLTL, is a sensitive belt of sea and land interaction. As the tidal flat is characterized as muddy surfaces, densely covered and frequently varied inlets, the routine topography survey method has been greatly restrained. However, remote sensing technology is large-scale, fast synchronized response and high frequency and so on, we can take advantage of those to investigate and monitor the intertidal flats.As the development of IT and RS technology, it's not difficult to obtain the high quality remote sensing imagine. Therefore, it has been the primary problem how to extract information from the abundant RS data to update the GIS database for the investigation and monitor of the intertidal flats.This paper is supported by "The tidal flats resource information service platform", which is a sub-projects of "Study on sustainable development of tidal flat of Shanghai" Sponsored by Science and Technology Committee of Shanghai. The aim of this paper is to get an operable method of automatic extraction of information of the tidal flats.The experimental area of this paper is East mud flat of Chongming island. The source data is a Landsat5 TM imagine as well as a SPOT pan imagine. In this paper, a few of questions have been researched:(1)The Choice of Best Detecting Band for Remote Sensing on tidal flats of Shanghai;(2)The Comparative research and choice on algorithm of differrent resolution merge;(3)The Comparative research and choice on different classifiers for remote sensing imagine on tidal flats of Shanghai;(4) Taking into consideration of the information of vegetation zoning, reflectance spectral signatures of land cover and field investigations, the estuarine and coastal wetland distribution information of Shanghai region is extracted from multiple Landsat-5 TM images on the basis of the determination of pixel threshold for land cover classification. Terrain beneath water between 0m～-5m is retrieved from digital sea charts. With ArcGIS, spatial data of the wetlands both of tidal flat and 0m～-5m are integrated. Therefore the wetlands distribution in the past and present and changes are thoroughly analyzed.