多源开放数据支持下的广州市城市扩张模拟
Urban Expansion Simulation Supported by Multi Opening-source Data: A Case Study of Guangzhou City
作者信息：
朱江 华南理工大学建筑学院博士研究生，广州市城市规划勘测设计研究院国土与空间规划设计所所长，教授级高级工程师
李翔 广州市城市规划勘测设计研究院国土与空间规划设计所助理工程师
尹向东 广州市城市规划勘测设计研究院国土与空间规划设计所副所长，高级工程师
姚江春 广州市城市规划勘测设计研究院国土与空间规划设计所副总工程师，高级工程师
摘要：以微博签到、夜间灯光和高德地图等开放型数据为基础,构建城市扩展指标体系。计算广州市全域空间单元城市扩展潜力,并利用生态重要性评价结果修正扩展潜力,然后基于元胞自动机模型分区模拟广州市城市扩展过程。研究发现:(1)2013至2017年扩展模拟过程中,模拟结果 Kappa系数为0.91,精度较高;(2)辖区之间模拟结果精度存在差异,中心城区辖区的模拟结果精度较高,Kappa系数为0.9左右,外围辖区模拟结果精度相对较低,Kappa系数为0.8左右;(3)2018至2035年,广州市主城区城市增量有限,以填充式扩展为主,外围区新增城市规模较大,主要集中在花都区、白云机场空港片区以及南沙新区,存在填充和蔓延两种扩展方式。研究将互联网开放数据与城市规划结合,实验结果说明,充分利用开放型数据能有效提高规划的科学性和准确性,同时,模拟结果有助于预判城市格局的演变,辅助战略预留区划定和重点项目布局。
关键词：多源开放数据; 城市扩张模拟; 元胞自动机; 广州市;
Abstract：Rapid urbanization in China has turned large amount of non-construction land into construction land (urban and rural construction land take up the major),providing abundant carrying capacity for population and economy. However, improvident development pattern has caused many problems, such as invading cultivated land,occupying ecological land and so on. Stimulating the urban expansion can help predicta city’s spatial pattern, providing suggestions for solving the above problems. In this research, based on opening source data, like Weibo check-in data, night-light data and Gaode electronic map data, we tried to build the indicators system of urban expansion.
Then, with logistic regression model, we calculated the expanding potentiality results of every spatial grid, and corrected the results. Then, we calculate the ecological importance to constrain urban expansion. Finally, using cellular automata (CA) model, we stimulated the expansion process in different time intervals, and checked the stimulation results’ accuracy through Kappa coefficient. The findings including: (1) From 2013 to 2017, using opening source data can achieve better results, and the Kappa coefficient of stimulation results and standard data is 0.9146. (2) There exists difference of results’ accuracy of different districts, and the districts in central area have a higher accuracy than the outside districts in stimulation results. (3) From 2018 to 2035, we found that the increment of central districts is limited, and the expansion pattern is the filled type. By contrast, the outside districts possess much more amount, mainly distributed in Huadu district, Baiyun Airport Economic Zone, and Nansha district, and their expansion pattern is the sprawling type instead. This research introduced opening source data into urban planning. The Experiments results indicate that using opening source data is able to make urban planning more scientific and reasonable. Furthermore, our results can assist in anticipating Guangzhou’s spatial pattern in the future, providing suggestions for land-use planning, delineating strategic reserve area and arranging major projects.
Keywords：multi opening-source data, urban expansion simulation, cellular automata, Guangzhou
文章来源：《城乡规划》2020 NO.1
