基于粗糙集与突变级数法的土地利用景观分区研究

Landscape Zoning of Land Use Based on Rough Set Theory and Catastrophe Progression Method

基于江苏宜兴市土地利用现状数据库,利用土地景观格局指数,采用粗糙集与突变级数法相结合的方法对宜兴市进行土地利用景观分区。分区结果表明:区域地貌特征影响的土地利用方式是土地利用景观格局分异的主要因素,随着水域、林地面积比重的增加,人类对土地系统的改造逐渐减弱,斑块破碎化程度降低,斑块形状复杂化程度增强,景观多样性减弱;结合粗糙集论通过知识约简挖掘数据中隐含的内在规则,依据属性重要性对评价因子进行重要性排序,无需确定权重,最终通过多指标集成的突变级数法进行土地利用景观分区。

Based on the currently land use database in Yixing of Jiangsu Province,landscape pattern indices were firstly calculated.According to these computed landscape pattern indices,Yixing was divided into four landscape zones of land use by the use of rough set theory and catastrophe progression method.The zoning result reflects that land use under the influence of regional geomorphology characteristic is the main factor of the landscape pattern＇s variation of land use.With the increase of the area of water and forest proportion,the human disturbance to land system will gradually weaken,patch fragmentation will reduce,patch shape complexity will enhance,and landscape diversity will decrease.The hidden rule under the raw data is simply obtained by knowledge reduction of data in rough set theory.Without the determination of the weight,it just needs to arrange indicators according to the computed importance of attribute.The landscape zoning of land use,finally,is made by multi-indicator integrated catastrophe progression method.The catastrophe progression method,in a large extent,avoids the subjectivity in the process of determining the factor＇s weight.