Construction and Realization of the Estuarine Water Environment Evaluation System Based on SOM

In order to explore a reliable method of evaluating the quality of the estuarine water environment, with SOM(Self Organizing Neural Network)networks, we developed a quality evaluation system on estuarine water with ComGIS. Embedded with SOM neural network model, the system can classify the estuarine water quality on the basis of quality index data. Considering the real needs, the standard control data are divided into two kinds of types. When the standard reference data are input, the results coincide with the standard data and display the corresponding level of the water quality. AS there is lack of reference data, the result would be a simple clustering classification, which can only display stations of the same category and hardly provide the status of water quality. The system was tested by the in situ water quality data collected from Daliao River estuary and the classification results were proved intuitive, reasonable and achieved the requirement of the visual effect. The system not only can be applied in estuarine water environmental quality assessment or other water environment quality assessment, but also provide data support and management for the river basin authorities.