基于信息熵和模糊物元模型的城乡结合部农地质量评价--以湖北省荆州市为例  

Evaluation of Cropland Quality in Jingzhou's Urban Fringe

作　　者：聂艳[1] 周勇[1] 钱建平[2] 黄建武[1] 

机构地区：[1]华中师范大学城市与环境科学学院,武汉430079 [2]国家农业信息化工程技术研究中心,北京100089

出　　处：《资源科学》2008年第6期919-924,共6页Resources Science

基　　金：国家自然科学基金:“可持续耕地分等定级模型、方法与信息系统集成”(编号:40671078);湖北省科技攻关计划:“湖北省土地利用/覆盖变化及环境效应研究”(编号:2007AA401B13)

摘　　要：城乡结合部是城市与农村相互结合的特殊经济地理单元,是城市快速扩张的重点区域。目前城乡结合部范围不明确,确定方法以定性为主,土地评价偏重土壤质量。针对此点,文章以信息熵理论为基础,通过遥感解译TM影像获取研究区土地利用类型图后,构建土地利用信息熵模型,计算500m×500m网格单元的熵值后借助突变点检测法确定城乡结合部内外边界,内外突变点位熵值为30;然后确定农地质量评价指标体系,通过叠置法和多边形法相结合来获取评价单元及其属性数据,借助地理信息系统(GIS)和模糊物元贴近度聚类分析模型计算评价单元贴近度,划分研究区农地质量等级,并采用野外调查和相关系数法检验结果。结果表明:研究区农地质量中等偏上,以二级,三级为主,占评价总面积的73.35%;尽管受多因素及RS栅格数据特点的影响,城乡结合部带有锯齿状和犬牙交错,但本研究结合了行政界线完整性的原则,划分方法科学、适用,其结果为荆州市确定城市扩展方向、合理利用城乡结合部土地资源提供了重要参考依据,同时也能够为类似地区城市扩展方向的确定提供一种较好的思路。Land use change and its influence on the natural environment have received widespread attention. Comprehensive evaluation of the quality of cultivated land, especially in peri-urban areas, has become even more important for current and future land resource management practices. Since they are affected by both rural and urban development processes, areas on the urban fringe are economically and geographically distinctive. Demarcation of the urban fringe is qualitative and uncertain, and land evaluation lays particular stress on soil quality. With the city of Jingzhou in Hubei Province as a case study, this paper describes the urban fringe area and evaluates cropland quality. We created land use maps using supervised classification for Landsat images from October 2000. A 500 x 500-meter grid was established for analysis, and an information entropy model was developed. By calculating the entropy of every grid cell and analyzing the distribution characteristics, we found that the range of entropy values of Jingzhou＇ s urban fringe was around 30. The range of urban fringe was identified as 124.4 km2, including 25 villages. In order to reduce subjectivity, a fuzzy matter-element proximity cluster analysis model was introduced. After confirming the indexes related to cultivated land quality, the overlay method and polygon method were used to divide evaluation units and acquire attribute data. Based on membership functions of evaluation factors and the fuzzy matter-element model, we calculated the weight of evaluation factors and proximity of cells, then sorted cultivated land into various quality levels. 120 samples were randomly collected for verification. The results indicate that cropland in this area is of medium quality and can be sorted into 5 levels, which are roughly in accordance with the actual situation. The results of this study support the use of this methodology as a preferred approach for this type of research.

关 键 词：信息熵 模糊物元贴近度聚类 城乡结合部 农地质量 

分 类 号：F301[经济管理—产业经济]