不同地貌区耕地集约利用水平空间分异及驱动机制——以江西省为例

摘要

Based on global and local spatial autocorrelation analyses of exploratory spatial data, the spatial disparity and driving forces of cultivated land intensive utilization at the county level were discussed by using Arc View GIS and Geoda software. The results show that:Global Morans I value was 0. 16 and 0. 30. There was an obviously temporal increase of Moran's I value from time Ⅰ to Ⅱ. The global spatial autocorrelation for intensive utilization levels of cultivated land was significant. The spatial clustering phenomenon about regional intensive utilization levels of cultivated land appeared on the whole. The intensive utilization levels was almost the same in some region by analyzing the Grid figure of Local Moran's I. The character of spatial clustering about regions of high value and low value was significant. The quantity of counties with the positive association was higher than that with the negative association. The regions with the "high-high" correlation were mainly located in the north of Jiangxi Province. However,the regions with the " low-low" correlation were distributed in hilly area of northeast, northwest and south. The result of study conforms to the facts basically and has certain feasibilities. The main driving forces influencing cultivated land intensive use in different regions were different.%在构建衡量耕地集约利用水平的P-S-R模式指标体系基础上,以县域为基本空间单元,借助GIS空间分析软件,利用ESDA等方法,对江西省各地貌区耕地集约利用水平空间差异及其驱动机制进行了研究,针对各地貌区制约耕地集约利用因素,分别提出了今后耕地利用过程中应注意的问题.结果表明:两时段基于综合度的Global Moran's Ⅰ分别为0.16和0.30,在显著性水平α=0.05下呈空间正相关,耕地集约利用水平在地域上呈现出较明显的空间集聚特征与区域差异性,且有不断加强趋势.总体上平原区的集约水平高于丘陵与山区的,经济发展较快地区的相对较高；各研究实体Local Moran'sI及Grid图,反映了耕地集约利用水平在局域空间上的均质性,即水平高、低区域各自集聚；LISA显著集聚四种类型中,正空间自相关的“H-H”、“L-L”占主体部分,时段Ⅰ二者约占总数的73.33％,时段Ⅱ则提高到80.95％.其中,“H-H”关联的区域主要分布在北部平原区、经济实力较强的县(市),而赣东北、西北及南部的丘陵山区则表现出明显的“L-L”关联特征；平原、丘陵、山区三大地貌区影响耕地集约利用水平的因素各异.