Abstract: Land use change is a dynamic process and it is affected by the interaction of multiple factors. At present, it has become a global environmental change and sustainable development content, while, regional land use spatial pattern has become the key research of LUCC content. Based on the remote sensing image in the year of 2000 and 2010, combined with digital elevation model, water system, railway, highway, rainfall and temperature data, binary logistic regression model is used to select the best analog scale on loess platform areas, and on the basis of it, the land use types of loess platform areas are set to simulate land use spatial pattern. The results indicate：（1）In the land-use pattern of simulated ten spatial scales, the correlation characteristics of land-use changes in spatial pattern and its driving force factor exist between certain scale. （2）The ROC values of three main land use types in ten spatial scales are shown a increasing trend at first, then decreasing, and its turning point is 400 m. It indicates that 400 m×400 m is the best analog scale among the effect of scale effect and scale conversion in this area during the land-use pattern optimization of analog scale. （3）In the best analog scale of 400 m, the distribution patterns of grassland and woodland are most significantly with the variables of per capita GDP and terrain composite index. The most obvious impact variable on the distribution of arable land is terrain composite index.

Key words: loess platform areas, Binary Logistic, land use, spatial simulation, simulation scale

摘要： 土地利用变化是一个受到多重因素相互影响的动态过程。目前，已经成为全球环境变化和可持续发展的重要内容，而区域土地利用空间格局模拟已成为LUCC研究的关键内容。以2000年以及2010年的TM遥感影像解译数据以及数字高程模型、水系、铁路、公路、降雨量和气温等数据为基础，运用二元逻辑斯蒂回归模型对黄土台塬区的土地利用最佳模拟尺度进行了选择，并在此基础上对研究区的各种土地利用进行了空间格局模拟。研究结果显示：（1）在土地利用格局模拟的十个空间尺度上，土地利用变化空间格局与其驱动力因子之间存在着一定的尺度相关性特征；（2）黄土台塬区耕地、林地、草地的ROC值在十个空间尺度上均呈现出先增加后减少的趋势，转折点在400 m尺度附近，说明黄土台塬区的土地利用在尺度效应和尺度转换的效应下，400 m×400 m是此区域土地利用格局优化的最佳模拟尺度；（3）在400 m最佳模拟尺度上所模拟出的草地和林地的分布格局都与人均GDP和地形综合指数两个变量显著相关，而对耕地的分布影响最为明显的变量则是地形综合指数。

关键词: 黄土台塬区, Binary Logistic, 土地利用, 空间模拟, 模拟尺度