基于模糊神经系统与GIS的区域土壤侵蚀快速评价

土壤侵蚀调查与评价是水土保持工作的重要组成部分。该文对区域土壤侵蚀评价进行了尝试性研究，提出一种基于模糊神经系统和GIS技术的快速评价方法。研究中利用模糊神经系统技术从地面监测调查数据中提取土壤侵蚀强度的评价规则，建立能够全面、客观地反映特定区域范围内土壤侵蚀自然特征及规律的评价标准，在此基础之上结合GIS技术以实现对整个地区的快速评价。采用该方法进行湖北省土壤侵蚀普查，结果表明全省侵蚀面积约占国土总面积的30.1%。其中，鄂西南、鄂西北高山区土壤侵蚀情况最严重，鄂东南、鄂东北低山丘陵区次之，鄂北岗地中度侵蚀面积较大，而江汉平原基本无明显侵蚀。经验证评价结果与实际情况具有较好的一致性。

Rapid assessment of regional soil erosion based on fuzzy neural system and GIS

Reliable soil erosion investigation and assessment are important parts for implementing effective soil and water conservation measure. In order to assess soil erosion status at regional scale rapidly and effectively, this paper presented a assessment approach integrating fuzzy neural system and GIS technique. By using that approach, five environment factors of soil erosion were selected as evaluating indicators, and proper basic assessment units were then created with GIS overlay analysis, after that comprehensive erosion assessment rules for basic assessment units were extracted from field survey data with neuro-fuzzy system. Thus it was capable to establish evaluating criteria, which could reflect the inherent characters between the multiple environment factors and the soil erosion patterns in the specific areas. Supported with GIS technique, it was easily to scale up erosion assessment from field-plot scale to regional scale. Application of the approach to Hubei Province, the results showed that about 30.1% of the total land area in Hubei Province was affected by different levels of soil erosion problem. In the western and eastern mountain areas soil erosion damages were very severe, and the northern hills mainly suffered from moderate erosion and chiefly slight in the central plains. Comparison of the results with actual conditions, it indicated that they were in good conformity, proving that the approach was reasonable and applicable.