Geographical Information Systems (GIS) are powerful tools for the manipulation of spatial objects, and Remote Sensing (RS) data has been used historically to depict and assess ocean environment, and that data can be translated directly into maps, image maps, GIS data layers or text reports. The application of GIS/RS technology to ocean and coast area use has continued to be a major international growth area, with the launching of new high-resolution satellites providing added impetus in last decade. The submarine sand ridges in the South Yellow Sea are located between 32°00N-33°48'N and 120°40'E-122°10'E, off the coastline of Jiangsu Province, China, and the area is about 22470km2. They are the biggest submarine sand ridges in the world, and also have a unique radial shape. Since it is very hard to collect in situ data in this area for the labyrinthine navigation, there are only few of time series of map could be imported to GIS for analysis. So the remotely sensed imagery, e.g. AVHRR, LANDSAT, and RADARSAT imagery, is an important part to determine the recent movement of submarine sand ridges. However, by now, the interface between GIS and remote sensing systems is functional but week. Each side suffers from a lack of critical support of a type that could be provided by the other. For this project the GIS has a continuing need for timely, accurate updates of the various spatial data entities, whereas remote sensing systems could benefit from access to highly accurate ancillary information to extract more useful information from the imagery to determine the movement of sand ridges. There are many things to be considered when performing change detection analyses, especially, processing procedures in GIS and RS are obviously different. This project examine how the submarine sand ridges of South Yellow Sea have changed in an high resolution on both GIS and RS, using TM at 30m and SAR at 12.5m, over the period from 1968 to 1995 - almost thirty years, and predict the change in next 20 years.in this project, the GIS database is managed in ESR's workstation ARC/NFO (version 7.2) software and the remote sensing procedures are supported by ERDAS MAGNE (version 8.2) and PC (version 6.3) software, running on a network of UNX workstations and WNDOWS NT 4.0.

In this project address the technology of GIS and remote sensing and seek to initiate a predictive modeling of the submarine sand ridges by multi-layer feed forward neural networks.