Cellular automata for simulating land-use change with a constrained irregular space representation: A case study in Nanjing city, China

Abstract

Traditional cell-based cellular automata (CA) models use a regular cellular grid to represent geographic space, and new approaches to CA models have explored the use of a vector representation of space instead of a regular grid to characterize urban space more realistically. However, less attention has been paid to modeling the interaction between the geospatial information and the irregular cells. To date, the majority of spatial boundaries have been created by individual agencies in an uncoordinated manner. As a consequence, the potential uses of the data collected for land-use change models are limited. In this paper, we propose a new vector-based CA model based on a new constrained irregular space representation using the theory of hierarchical spatial reasoning. For dividing the geographic space considering different items, first land patches are considered as the minimum division unit; then aggregation rules, including attribute, geometric and boundary barrier constraints, are defined; and finally different levels of spatial units are formed through land patches based on aggregation rules. The proposed model is used to simulate the land-use changes in Nanjing, Jiangsu Province, China. The performance validation and comparison illustrate the feasibility of the proposed space representation in a CA model. By using this model, it is expected that the use of the real spatial boundaries that are employed in urban planning could help provide a flexible paradigm to consider various drivers or constraints for realistically simulating land-use changes.