多尺度联合驱动的城市增长模拟建模  

A Multi-scale Synergistic CA Model for Urban Growth Simulation

作　　者：马世发[1] 张婷[1] 李少英[2] MA Shi-fa ZHANG Ting LI Shao-ying(Land and Resources Technology Center of Guangdong Province （Geomatics Center of Guangdong Province）, Guangzhou 510075 School of Geographical Sciences, Guangzhou University, Guangzhou 510006, China)

机构地区：[1]广东省国土资源技术中心/广东省基础地理信息中心,广东广州510075 [2]广州大学地理科学学院,广东广州510006

出　　处：《地理与地理信息科学》2017年第2期19-24,共6页Geography and Geo-Information Science

基　　金：国家自然科学基金项目(41401432;41531176);广东省国土资源科研项目(GDGTKJ2016005)

摘　　要：元胞自动机是模拟土地利用时空演变的重要建模工具,但其经典框架主要根据历史变化规律进行空间"自下而上"驱动,这种模拟机制缺乏对宏观土地供需平衡以及空间发展政策等"自上而下"机制的整合,难以科学解释中长期土地利用变化过程。该文利用金字塔架构解释城市空间增长的复合驱动力,提出了"顶层土地供需平衡-中层空间政策引导-底层元胞状态演化"多尺度联合驱动的土地利用变化模拟框架。以广州都市区为案例,利用改进模型对其2010-2030年的城市扩张过程进行了模拟,并对结果进行了精度对比和案例举证。研究表明,多尺度联合驱动进一步增强了元胞自动机模型对现实世界的仿真能力,多尺度空间协同是地理时空过程建模的发展方向。Simulation of land use changes can help to understand the dynamic process of land use and provide important informa- tion for related decision makings. Cellular Automata （CA） models have been widely used to model land use change due to their abilities to simulate dynamic spatial processes from a bottom-up perspective. Traditional CA models mainly simulate land use dynamics based on the historical local rules. However, land use change is a complex and dynamic process which involves both natural and human systems. The simulation results are difficult to explain the process of land use change in the long term by u- sing traditional CA model. In this paper, the pyramid framework is used to explain the composite driving mechanism of land use change. A Multi-scale Synergistic Cellular Automata（MSCA）model is proposed for land use simulation. In this model, the top layer o{ pyramid framework represents the balance of land supply and land demand, and the quantity of land use change is deter- mined by the driving forces of social and economic developments. The intermediate layer describes the government policy guid ance and reflects the spatial differentiation of land use development. In the bottom layer,the model simulates cellular evolution with local rules. The integration of these three driving layers is carried out by a coupling mechanism of spatial and temporal process. The proposed model has been applied to Guangzhou metropolitan area for simulating the urban expansion process from 2010 to 2030. This model is compared with the classical CA model by using overall accuracy and local space comparison. The re- sults indicate that MSCA model can generate more reliable simulation results than traditional CA models which are merely driv- en by local historical rules. MSCA model appears to be a valuable solution to take into account the human driving forces of land use change at the macro level. The proposed method can further enhance the simulation capabilities of CA model especially ap- plied to the real world. Mu

关 键 词：土地利用/土地覆盖变化 元胞自动机 多尺度 空间规划 广州市 

分 类 号：P208[天文地球—地图制图学与地理信息工程]