中文摘要：

对南昌地区城市发育约束因素的分析的基础上,基于非城市像元所处的城市化梯度、距离最近建成区斑块的距离、最近建成区斑块的规模、边缘像元的邻域水平等4个城市景观格局因子和道路及地形等两个社会经济和自然影响因子,构建了广义转移概率模型,并开展了相关的动态模拟研究。表明,与传统模拟方法相比,该模型具有比较理想的精度水平,模拟结果能够很好地反映南昌地区今后城市建设用地发展趋势,提出的模型手段改进设计获得了令人满意的效果,是对现行城市景观动态模拟研究工作的一个重要的方法论补充和完善。

英文摘要：

In the traditional research about the spatial dynamic modeling of urban built-up area, there are some unavoidable deficiencies derived from the limited image datasets for a certain time interval and the subjectivity in the determination of transitional threshold. In order to solve the aforementioned problems, the author established a general-conversion probability model which can consider the transition of all the pixels and improve the existing urban landscape dynamics models. In the case study of Nanchang City in Jiangxi Province, the author established a general transition probability model and conducted related dynamic modeling by taking into consideration of 4 urban landscape pattern factors including the urbanization gradient of non-urban pixel, the distance to the closest built-up patch, the size of the closest built-up patch, and the neighborhood level of peripheral pixel, and 2 socio-economic and natural factors including road and topography, which all together can strongly control the urbanization process in the region. The simulation result indicates that this model is more precise compared to the traditional models. It can simulate the expansion of urban built-up areas of Nanchang perfectly, and it is an important methodology supplementary of the research about the current spatial dynamics modeling of urban landscape.