基于CLUE-S模型和Markov模型的城镇土地利用变化模拟预测——以江苏省常熟市辛庄镇为例

Simulation of Land Use Change in Xinzhuang Town under Different Scenarios Based on the CLUE-S Model and Markov Model

基于高分辨率遥感影像解译得到江苏省常熟市辛庄镇4期土地利用历史数据,首先对CLUE-S模型的预测精度进行验证,然后利用CLUE-S模型和Markov模型相结合的方法对研究区未来土地利用变化进行3种不同情景的模拟预测,并系统分析了各情景下土地利用变化的时空特征。结果表明,CLUE-S模型在乡镇小尺度的土地利用变化模拟预测效果较好,精度较高。各情景下,建设用地的快速增长仍将在未来的土地利用变化中占据主导地位,且均以消耗大量的水田为代价,河流密布、交通便捷和经济发达的区域将是未来耕地非农化的"热点"区域。生态保护情景和基本农田保护情景对未来土地利用变化的调控效果较好,林地、水体、水田和旱地受到了更好的保护,并在一定程度上抑制了建设用地的肆意蔓延。

As an important research field of global environmental change and sustainable development, land use change has drawn much attention. Model simulation is the key in predicting future land use change. Land use models are an important technical means to better understand the process of land use change, driving mechanism, and ecological effect. Simulation and prediction of land use change under different scenarios would be of great significance for scientific decision making on land use planning and management. Based on historical land use data extracted from high-resolution satellite images acquired in 1980, 1991, 2001 and 2009 of Xinzhuang Town, the accuracy of predictions of CLUE-S model was first verified. Then, future land use change of the study area was predicted on the basis of the combination of the CLUE-S model and Markov model, with systematically analyzing the spatiotemporal characteristics of land use change under three different scenarios （historical trend, ecological protection, and basic farmland protection）. Results showed that the predictability of the CLUE-S model at town scale seems to be satisfactory, which means that the CLUE-S model may have potential to be well applied at relatively small scales. Conversions from paddy fields to construction land, from paddy fields to fish ponds, and from water area to construction land, were the three major types of land use change. Rapid growth of construction land would remain in the future under all scenarios, especially under the historical trend scenario, with considerable decreases in paddy land concurrently occurring. Areas for rivers, convenient traffic, and developed economy would increase, which is referred to as farmland non-agriculturization. Forest, water bodies, and other types of ecological land would be well protected under the ecological protection scenario, with a total area of only 75.92 hm2 being transferred to other types, far less than the other two scenarios. Under the basic farmland protection scenario, paddy land and dry land would be well protected, with only 297.36 hm2 being converted into other land use types. In short, the regulatory effects of the basic farmland protection and ecological protection scenarios are obvious. Land use change in the future can be better controlled due to more restriction and preservation of forest land, water bodies, paddy land, and dry land. Rapid sprawl of construction land could therefore be restricted to some extent.