基于约束性CA的大都市郊区城镇增长的情景模拟与管控——以武汉市江夏区为例

Scenario simulation and control of metropolitan outskirts urban growth based on constrained CA:A case study of Jiangxia District of Wuhan City

大都市郊区是快速城镇化进程中空间演变最为频繁、人地矛盾最为突出的区域,尤其在中国加快推进"就地就近"城镇化战略的背景下,把握大都市郊区小城镇土地利用时空变化过程及其演变机制,对制定科学合理的管控政策和优化都市空间结构具有重要的现实意义。约束性元胞自动机(constrained Cellular Automata,constrained CA)能够通过简单的规则模拟复杂的城市动态演化过程。本文将土地利用总体规划指标、城镇空间发展战略布局、土地利用开发适宜性等,嵌入约束性CA的转换规则中,采用Logistic逐步回归法分析土地利用空间影响因素,对严格约束下的武汉市江夏区2020年土地利用进行情景模拟分析,并提出城市增长管控手段。结果表明:(1)研究时段内,江夏区城镇用地呈低效外延式扩张,土地利用集约节约程度较低,其人口规模并未有较大增长,对主城区人口的分散作用尚未真正形成;(2)约束性CA在模拟大都市郊区演化方面具有较高的可靠性,能够真实反映近郊小城镇的未来空间布局与结构,模拟结果与土地利用规划和城市规划较为契合;(3)将规划目标导向与现实发展趋势下的模拟结果进行叠加分析,可确定城镇增长需求与规划指标调控间冲突的空间分布,从而划定土地督察的重点监测区域,为加强大都市近郊区的违法用地监查和土地利用管控提供先验的预警知识。

In the process of rapid urbanization in China, metropolitan outskirts are areas where land use change most frequently occurred and human-nature relationship most complex, and it is very important to understand changes in land management and urban planning in these areas. Constrained Cellular Automata （constrained CA） can simulate complicated urban dynamic changes in accordance with certain simple rules and has become an important tool in the study of spatiotemporal changes of urban land use. Using land use data of Jiangxia District of Wuhan City collected in 2007 and 2011, this study incorporated considerations such as overall land use planning index, strategic arrangement of urban spatial development, and land use constrains into the transition rules of Cellular Automata. Then the logistic regression method was employed to determine the influencing factors of city spatial development and a constrained CA model was established to simulate scenario urban construction of Jiangxia District in 2020. The findings were compared with the prediction of forecast-CA. The simulation results show that： （1） Jiangxia District gradually absorbs some of the functions and population of the city center during the study period, but the urban expansion and land use is of low efficiency. The population of the study area will not grow significantly and the expansion of the area will not ease the high population pressure in the city center. （2） Constrained CA is able to depict a vivid picture of the layout and structure of small towns in metropolitan outskirts and is suitable for land use planning and urban planning. （3） Overlay analysis of the results of scenario simulations can identify the key monitoring areas of land use supervision and provide warning messages for the supervision and control of illegal land use.