交通驱动下微观地块尺度的城市土地利用变化模拟——以深圳市为例

Simulation of Urban Land-use Change at Micro Land Parcel Scale Driven by Traffic: A Case Study of Shenzhen

城市交通作为土地利用空间格局变化的重要驱动因素,在城市发展模拟研究中值得重视。如何有效挖掘城市交通因素并引入地块尺度城市土地利用模拟成为重要议题。本文提出一套基于矢量元胞自动机考虑交通因素的城市土地利用变化模拟框架(T-VCA)。该框架综合交通流、信息流、经济流等数据量化城市连通性因子,基于路网数据有效量化交通可达性因子,将其引入矢量元胞自动机模型,能够有效模拟微观地块尺度下的城市土地利用变化。以深圳市为研究区,本研究所提出的T-VCA模型模拟精度最高(FoM=0.266),相较于最新的基于随机森林的矢量元胞自动机(RF-VCA)模型模拟精度提高了11.05%,景观指数相似度高达96.00%,表明考虑交通因素可以有效提高城市发展模拟精度。T-VCA模型较仅考虑城市连通性的矢量元胞自动机(C-VCA)模型和仅考虑交通可达性的矢量元胞自动机(A-VCA)模型精度分别提高2.67%和3.75%,表明城市连通性因素适用于挖掘新兴城区和远郊区的土地利用转化规则,而城市交通可达性因素适用于挖掘发展成熟的中心城区的土地利用转化规则。本研究可为城市规划人员在城市交通和土地利用管理方面提供参考。

As an essential driving factor of land use spatial pattern change, urban traffic deserves attention in urban development simulation studies. An important issue is how to effectively mine the urban traffic factors and introduce them into the fine-scale urban land use simulation. This paper proposes an urban land-use change simulation framework(T-VCA) based on vector-based cellular automata and considering traffic factors. The framework integrates traffic, information and economic flows to quantify urban connectivity. Based on the road network data, the traffic accessibility factor is effectively quantified. The framework introduces them into the vector-based cellular automata model and can effectively simulate urban land-use changes at the micro-plot scale. Taking Shenzhen as the study area, the T-VCA model achieves the highest accuracy(FoM=0.266), whose accuracy is 11.05% higher than the latest RF-VCA model. And the similarity of the landscape index is up to 96.00%. These indicate that considering traffic factors can effectively improve the accuracy of urban development simulation. Compared with C-VCA which is only considering urban connectivity and A-VCA only considering traffic accessibility, the accuracy of the T-VCA model increases by 2.67%and 3.75%, respectively. The study shows that the urban connectivity factor is suitable for mining the land use conversion rules in emerging urban areas and distant suburban areas. In contrast, the urban traffic accessibility factor is suitable in mature central urban areas. This study can provide references for urban planners in urban transportation and land use management.