GEOGRAPHICAL RESEARCH ›› 2017, Vol. 36 ›› Issue (3): 553-572.doi: 10.11821/dlyj201703013

• Articles • Previous Articles     Next Articles

Landscape pattern optimization in peri-urban areas based on the particle swarm optimization method: A case study in Longquanyi District of Chengdu

Dinghua OU(), Jianguo XIA()   

  1. College of Resources, Sichuan Agricultural University, Chengdu 611130, China
  • Received:2016-07-14 Revised:2016-12-05 Online:2017-03-20 Published:2017-03-22

Abstract:

Landscape pattern optimization is one of the important ways in achieving regional ecological security. In order to optimize landscape spatial layout of economic development, ecological protection and overall consideration scenario, a case study was carried out in Longquanyi District of Chengdu based on landscape suitability assessment and landscape quantity optimization. The method of particle swarm optimization (PSO) was used to establish landscape pattern spatial optimization model and algorithm. The results showed that, the landscape pattern spatial optimization model and algorithm based on PSO was the effective method of landscape pattern optimization. The model and algorithm could efficiently simulate landscape distribution by using particle space positions, conduct spatial pattern optimization, and realize united coupling of quantity structure and spatial distribution optimization. The dominant landscapes in economic development scenario were orchard, urban-rural residential and industrial-mining area in a target year. The landscape pattern showed that farmland, urban-rural residential and industrial-mining area dominated the western flatland region, while eastern mountainous area was mainly dominated by orchard. The dominant landscapes in ecological protection scenario were forest, urban-rural residential and industrial-mining area. The landscape pattern showed that the western flatland region was mainly dominated by farmland, orchard, urban-rural residential and industrial-mining area, while the eastern mountainous area was dominated by forest. The dominant landscapes in overall consideration scenario were forest, orchard, urban-rural residential and industrial-mining area. The landscape pattern showed that farmland, urban-rural residential and industrial-mining area dominated the western flatland region, while eastern mountainous area was dominated by forest and orchard. Compared with other scenarios, the overall consideration scenario could be the largest potential possibility in the future, since the economic, ecological and comprehensive benefits here would be the most optimized and promoted, implying the best spatial layout scheme for landscape pattern in the study area in the target year.

Key words: peri-urban areas, landscape pattern, spatial optimization, particle swarm optimization, constrained optimization method, Logistic regression model