MCE法在土壤侵蚀危险评价中的应用

Application of MCE method in soil erosion risk evaluation

影响土壤侵蚀的因素很多,有自然因素(如地形、植被、土壤、降雨等)和人为因素,这些因素中既有量化指标如坡度、降雨等自然因素,也有土地利用、土壤母质等不可量化的指标。这些因素的共同作用,使得单纯在GIS下利用叠加的方法对土壤侵蚀潜在危险进行判别带来一定的困难(如人为因素的量化问题等)。多标准评价法(Multi-Criteria Evaluation,MCE)可以将难以量化的指标进行量化处理,并综合考虑多种因素对目标的影响,通过加权线性合并及布尔叠加的方法,在众多因素或是相互矛盾的客观实际中确定一种折中的优化方案,来对目标作出客观的评价。主要介绍了多标准评价法的概念及其方法原理,并以江西兴国县为例,介绍了该方法在土壤侵蚀评价中的实现过程。在土壤侵蚀评价因子的选用上,先取了植被指数(LAI)、土地利用、坡度、土壤母质、高程、道路、降雨及人口密度八个因子,根据实际情况确定单个因子在不同指标水平下可能造成的潜在土壤侵蚀危险等级;然后将因子组成矩阵,通过两两比较,来判断各因子的权重;把权重与相应的因子图相乘得到带权重值因子的栅格图,这些栅格图再进行统一的标准化处理,最后将标准化处理的栅格图通过叠加生成潜在侵蚀危险分布图。结果表明,兴国县大部分是微度侵蚀危险地区,呈连片分布,中、低度侵蚀危险主要分布在平原及沟谷地带,呈零星态,二者呈相互交错。微度危险、低度危险及中度危险区域分别占兴国县总面积的37.9%、29.2%及30.4%。三者之和为97.5%,高侵蚀危险及极高侵蚀危险的面积只占2.5%,且呈零星分布。这与兴国县的土壤侵蚀分布现状基本一致,多标准评价法应用在土壤侵蚀危险评价上是切实可行的。

Soil erosion is induced and controlled by a number of factors. Factors said include both natural ones, such as landform, vegetation, soil as well as rainfall, and human activities. Some, such as slope and rainfall, can be quantitatively measured, but others, such as land-use and parent material, can only be qualitatively described. The complexity of factors makes it difficult to evaluate potential erosion risks by solely using the overlapping method in GIS environment, because some problems, such as the quantifica- tion of human activities, remain to be solved. Fortunately, Multi-Criteria Evaluation （MCE） method, which take impacts of multi-factors on objects into consideration, enables the quantification of some factors which are usually difficult to be quantified, and thus the evaluation for objects by selecting a compromised and optimum scheme among numerous factors or complicated facts, using Weight Linear Combine（WLC） ＆ Boolean Overlapping methods. The present paper deals with a case study, conducted at Xingguo County, Jiangxi Province, China, on MCE and its application in soil erosion risk evaluation. Firstly, 8 factors, namely leaf area index （LAI）, land-use, slope, parent material, altitude, road, rainfall and population density are chosen, and the potential erosion risk class of each factor at various criteria levels is rated on the basis of actual situation. Then, weights of various factors are judged by formu- lating a factor matrix and comparing factors with each other, and weighted factor grid maps are resulted from multiplying those weights with corresponding original factor maps. Finally, after an unified standardization of those weighted grid maps, a Potential Erosion Risk Map of the County is compiled by overlapping said standardized maps. Results show that the major part of the County＇s territory is of slight erosion risk class, and most of the rest parts, sparsely distributing on plains and in valleys are of low and moderate risk classes. Areas of slight, low and moderate erosion risk classes account for 37.9%, 29.2% and 30.4%, respectively, of the total territory of the County. By contrast, area of both high and extremely high risk classes accounts only for the other 2.5%. It is concluded that it is effective and feasible to apply MCE method in soil erosion risk evaluation, for results of the case study are identical with the real distribution of soil erosion in the County.