Environ Geol 2006.pdfVIP

Introduction Karst rocky desertification is a progressive process of land degradation where soil is seriously or thoroughly eroded. Bedrock exposure is widespread, productive capability of land declines and finally, a landscape sim- ilar to desert appears under human impact on a vul- nerable eco-geo-environment (Yuan 1993; Yang 1995; Xiong 2002). This major type of desertification results in extensive exposure of bedrock. In the southwest moun- tain area of Guizhou, the karst area covers about 620,000 km2, which is the largest known karst area in the world (Cai 1994). Much research has been done on the process of karst rocky desertification including distribution (Su 2002; Xiong 2002), the ecological–environmental effects (Wang 2003), cause analysis (Wang and Li 2003; Zhang 2001; Zhang et al. 1999) and integrated rehabilitation (Cai 1994; Cai and Meng 1999; Huang 2000). Yet how to assess karst rocky desertification is still difficult be- cause there are many factors to weigh such as percentage of vegetation, percentage of area with slope 25
and annual precipitation. This study uses the GIS technique to extract spatial data, train data to identify the coefficients, and assess the 82 counties’ karst rocky desertification levels in Guizhou Province using the resulting RBFN model. The result is in conformity with actual conditions, which indicates that GIS coupled with RBFN model is a flexible and powerful approach for land assessment. The aim of this study is to devise an efficient and simple method for quickly classifying karst rocky desertification by means of GIS and RBFN model. Study area The study area, Guzihou province, at approximatly 24
– 29
N and 103
–109
E, is in the southwest of China (Fig. 1) and covers about 176,000 km2, of which karst Huang Qiu-hao Cai Yun-long Assessment of karst rocky desertification using the radial basis function network model and GIS technique: a case study of Guizhou Province, China Received: 14 July 2005 Accepted: 21 November 2005