Factor analysis for aerosol optical depth and its prediction from the perspective of land-use change. (October 2018)

• Record Type: Journal Article
• Authors: Zhang, Wenting; He, Qingqing; Wang, Haijun; Cao, Kai; He, Sanwei
• Is Part Of: Ecological indicators. Volume 93(2018)
• Page Start: 458
• Page End: 469
• Publication Date: 2018-10
• Subjects: Aerosol optical depth -- Wuhan agglomeration -- Geographically weighted regression -- Land use; Environmental monitoring -- Periodicals; Environmental management -- Periodicals; Environmental impact analysis -- Periodicals; Environmental risk assessment -- Periodicals; Sustainable development -- Periodicals; 333.71405
• DOI: 10.1016/j.ecolind.2018.05.026
• Languages: English
• ISSNs: 1470-160X

Abstract

Highlights: The spatial heterogeneity of aerosol can be handled by geographical weight regression. Under urban sprawl the aerosol would increase spatially in 2030. Different urban sprawl scenarios differed simulated aerosol in central urban areas. Abstract: This paper presents the non-stationarity and autocorrelation (with a Moran's I index score of 0.75) of the MODIS-retrieved aerosol optical depth (AOD) of the Wuhan agglomeration (WHA) in Central China, using geographically weighted regression (GWR) to identify the spatial relationships between AOD and its impact factors. In addition to the socio-economic factors, i.e., GDP and population, vegetation cover, elevation, land-use density and landscape metrics are also considered. Faced with the rapid process of urbanization and the impact of land-use change on AOD, which has been confirmed in previous studies, we propose an AOD prediction method, combining a land-use change simulation model, a cellular automata and Markov chain (CA-Markov) model, and spatial relationships built by GWR to represent the spatial distribution of AOD in 2030. The results suggest that the GWR model is able to address the spatially varying relationships, with an R-squared value, corrected Akaike's information criterion (AICc), and standard residual better than those of the ordinary least squares (OLS) model. Land-use simulation, with an accuracy of 89.76%, indicates that an increase in the built-up area and a decrease in the forest area will be the major trends of land-use change and will lead to increased AOD. The AOD simulation results indicate that the most developed areas, i.e., the cities of Wuhan and Huangshi, will be the AOD increase hot spots in the WHA. This study provides an alternative method to identify the varying spatial relationships between AOD and its impact factors. A spatial prediction method for AOD is developed from the perspective of land-use change, which will help land-use planners in decision making.