Effect of Compactness of Urban Growth on Regional Landscape Ecological Security

Abstract

With rapid urbanization destroying the ecological environment, scholars have focused on ways to coordinate harmonious development using urban spatial layouts and landscape ecological security. To explore landscape ecological security (the landscape elements, spatial positions and connections that are of key significance to the health and safety of ecological processes) from the perspective of urban form evolution pattern will help to open a new perspective of urban management research, and become the basic work of urban space policy and the implementation of the beautiful China strategy. Based on urban growth and land use data from 356 cities in China, this study applied a geographically weighted regression (GWR) model to quantify the impact of China’s urban growth pattern on landscape ecological security at the spatial level. The research results show that: (1) To some extent, the infilling growth pattern has a certain effect on the enhancement of regional landscape ecological security; (2) In the three control variables (DEM, Population density and GDP), the following conclusions are drawn: regional landscape planning should reasonably allocate landscape resources according to the local topographic features to obtain a higher landscape ecological security; The increase of population density leads to the fragmentation and diversity of the landscape in some regions, which makes the landscape ecological security weak; more economically developed areas have stronger landscape ecological security. This paper highlights the importance of urban growth patterns to landscape ecological security. In addition, considering the different urban evolution trajectories in developed and developing countries, this study proposes targeted development recommendations, providing a reference for urban managers to formulate reasonable development policies and to realize sustainable development with the goal of landscape safety management and control.