Synthetic Analysis for Extracting Information on Soil Salinity Using Remote Sensing and GIS: A Case Study of Yanggao Basin in China

/ This paper reports the experience of extracting information on the salinity of soil and offers a method of synthetic analysis. The experimental areas for analysis are located in Yanggao Basin, Shanxi Province, China. The types of soil are mainly meadow soil and salinized meadow soil. The method of synthetic analysis of salinity uses a geographic information system (GIS) as a tool, building a basic saltwater analysis model of saline soil and adjusting the result with expert experience after computer processing. The method of feature extraction has been used for remotely sensed data. An optimum combination of features has been determined and, after comparing several combinations in the Yanggao region, an improved result has been obtained after Kauth-Thomas (K-T) transformation. For precise quantitative analysis of the salinization, not only Thematic Mapper (TM) remote sensing data, but also two forms of non-remote-sensing data are needed: depth of groundwater and mineralization rate of groundwater according to the theory of genesis of soil. For the analysis of synthetic compounded multisources, a generalized Bayes classification is used after overlay, matching, and related coefficients have been determined. On the premise that various information sources are independent, global membership functions with probability are used to combine various pieces of information in order to apply them directly to the pixels and classifications of soil salinity. The experiment indicates that this analytical method is sound because of the increased speed of processing and its simplicity and improved precision of classification of salinity. Finally, it is necessary to examine and adjust the factors using expert intelligence. The experiment shows that synthetic analysis using the geographic information system can raise the precision of quantitative analysis of salinity, which has advantages for environmental monitoring and management.KEY WORDS: Salinity; Remote sensing; Thematic Mapper; Geographic information system; Classification