Delimiting the urban growth boundary for sustainable development with a pareto front degradation searching strategy based optimization model

Abstract

Optimal allocation of newly-added urban land is significant to delimit the urban growth boundary (UGB) for territorial spatial planning, and it is also the basic guarantee to promote the sustainable development of cities. Exploring a model to generate optimal allocation pattern has become one of the key technologies in UGB delineation. In this study, a fast land-use assignment model (FLAM) was proposed adopting Pareto front degradation searching strategy. In this model, urban lands to be allocated were defined as agents, and a heuristic landscape indicator was creatively induced for assigning agents to optimal positions. They first selected positions with the highest urban growth suitability, then some of them with lower suitability gradually moved from original positions to those around urban patches with maximum compactness, which was measured with the area of urban patch and the number of grids allowed for urban growth within the corresponding minimal outer rectangle. Optimal solution could be obtained until all the agents were assigned under the condition of maximum utility both in suitability and compactness. This searching and updating strategy actually can make optimal solutions always along the Pareto front. Taking Guangzhou metropolitan area in China as an example, FLAM was performed and validated with the following aspects: the sensitivity of model's parameters, response to planning demands, and optimization efficiency. Compared with cellular automata (CA) and ant colony optimization (ACO), FLAM has better performance on urban land allocation. Results show that FLAM can obtain reasonable scenarios with the advantages of few model parameters, fasting evolution speed and strong scalability, which can be well applied to support UGB delimitation.