日本落叶松人工林立地质量的灰色定权聚类  

Site Quality of Larix kaempferi Plantations Based on Grey Fixed Weight Clustering

作　　者：马友平[1] 

机构地区：[1]湖北民族学院,恩施445000

出　　处：《东北林业大学学报》2010年第11期58-62,共5页Journal of Northeast Forestry University

基　　金：湖北省自然科学基金项目(2008CDB049);湖北民族学院博士基金项目

摘　　要：为了研究日本落叶松的立地质量,外业测定了339个样点的土壤厚度值,并用手持GPS72记录其坐标位置;收集了43个样地资料;收集了1:10000北京54地形图6幅;收回专家评分表63份。内业采用逐步回归法得出:土层厚度、海拔、坡度、坡向4个因子对立地质量影响显著,可作为立地质量评定的主导因子;在ArcGIS9.1中生成数字高程模型(DEM),并获取了海拔、坡度、坡向的数据,精确到每个像元;对339个土壤厚度值检验服从正态分布,采用普通克立格法在ArcGIS9.1中生成分布图;4个因子的权重采用层次分析法确定,得出海拔、土层厚度、坡度、坡向的权重分别为0.0616、0.1737、0.2910、0.4737;立地质量的量化评定采用灰色定权聚类法,将立地质量分为好、中、差3类,构造出4个因子不同立地质量时的白化权函数,并计算出每个像元点的灰色定权聚类系数,绘制出全场的立地质量图,立地质量好的面积为354.60hm2(占总面积的39.26%)、中等的为429.75hm2(占总面积的47.58%)、差的为118.95hm2(占总面积的13.17%)。An experiment was conducted to evaluate the site quality of Larix kaempferi according to data of soil thickness of 339 sample points and their coordinates measured by portable GPS72 instrument,data of 43 sample plots,six pieces of 1/10000 topographic map of Beijing 54 Coordinate System and 63 pieces of expert evaluation forms.Results of the stepwise regression analysis showed that soil thickness,elevation,slope gradient and aspect could be regarded as the dominant factors for evaluating site quality of L.kaempferi due to their significant effects on site quality.An accurate digital elevation model was produced by using ArcGIS9.1,and the processing and analysis were carried out to acquire the data of elevation,slope gradient and aspect,with the accuracy of pixel.Soil thickness values of 339 sample points were proved to obey normal distribution,and a distribution map was generated by using Ordinary Kringing method in ArcGIS9.1.The weights of elevation,soil thickness,slope gradient and aspect were determined by using Analytic Hierarchy Process（AHP）,which were 0.061 6,0.173 7,0.291 0 and 0.473 7 respectively.Quantitative assessment of site quality was conducted by grey fixed weight clustering method,and site quality was classified into three levels（good,moderate,and poor）.A whitening weight function of different site qualities evaluated by the four factors was established to calculate the grey fixed weight clustering coefficient of each pixel and to draw a panoramic map of site quality.The areas of good,moderate,and poor site quality were 354.60,429.75 and 118.95 hm2 respectively,accounting for 39.26%,47.58% and 13.17% of the total area.

关 键 词：日本落叶松 立地质量 灰色定权聚类 克立格法 ARCGIS 层次分析法 

分 类 号：S758.57[农业科学—森林经理学] S791.223[农业科学—林学]