基于修正的转移概率方法进行城市景观动态研究——以南昌市区为例

利用南昌市1988～2000年期间5个时段的TM卫星影像数据编制了景观组分类型图，并通过叠图分析统计了全部组分类型在4个比较时段的转移概率矩阵。在此基础上构建了组分转入、转出贡献率和特定转移过程贡献率等3个动态分析参数，对南昌市的景观动态变化特征和驱动机制进行了研究。结果显示，城市化引发的建设用地规模急剧膨胀是工作区内景观结构在研究时段内急剧调整的主要原因；农业经营的稳定性需要作为一种重要驱动因素，在农业用地大量流失的同时，使其它景观组分向农田转移成为优势性转移过程类型；地形和水文变化对于林地和水体等组分变化具有次要约束性影响，但对于一些特定优势转移过程的出现则具有决定性作用。研究结果还证实了所提出的3种基于组分转移概率矩阵的量化参数在景观动态变化特征和驱动机制研究中的合理性和有效性。

Dynamic analysis of urban landscape using a modified conversion matrix method:A case study in Nanchang City

In this paper, we used TM satellite images of Nanchang City from five different periods between 1988 and 2000 to conduct a dynamic urban landscape analysis. Using a modified conversion matrix method, we calculated the conversion probability matrix for all the landscape types. We rely primarily on three parameters conversion in rate, conversion out rate, and specific conversion process contribution rate to analyze the landscape changes and the driving mechanism behind these changes. It was found that dramatic changes in landscape structure occurred in Nanchang City during the study period. Because of the rapid urban sprawl, there was an increase in urban land use and concomitant decline in forest land, farmland, and water bodies. Dominant process analysis shows considerable conversion from agricultural land and natural landscapes to urban ones, with complicated inter conversion among the other landscape types. In particular, we also noticed the dominant conversion to farmland as an important process to landscape dynamics. Analysis on the combination of conversion out and conversion in rates indicates that farmland has kept high conversion in rate as well as high conversion out rate. Water body conversion is affected by both urban development and agriculture management, and is strongly constrained by hydrological changes. Because of the unique topography of the study area, forest land is the most stable among all the landscape types. Grassland and shrub have always had high turnover rates. Conversion of bare land is exceptional, as it is strongly constrained by hydrological changes as well as large scale urban development activities. The main dynamic changes in the study area were primarily conversion from farmland to built up areas and from other landscape types to farmland. Urban development is the driving force behind the dramatic change of the landscape structures in the whole study area. Needs for the stability of agricultural management is the direct reason for the dominant conversion processes from other landscape elements to farmland. Natural environment conditions like topography and hydrological changes are secondary constraint forces upon the conversion of forest and water body, but have primary effects upon some specific dominance conversion process, e.g. from farmland to water body. The modified conversion matrix analysis is an effective method for studying landscape dynamics. But it should be complemented by some supplementary analysis methods that can reflect the importance of basic landscape types and their conversion processes. Only in this way, can the various conversion processes in the dynamic landscape change be effectively characterized, and in turn facilitate the subsequent analysis on the driving mechanisms of landscape dynamics.