Abstract:Urban spatial patterns are usually affected by different factors in urbanization processes. Scientific interpretations of the effects of underlying determinants of spatial patterns are important for a better understanding of urban developments. However, only few studies have quantitatively examined the spatiotemporal relationships between spatial patterns and driving factors. This study explores the use of remote sensing (RS), spatial metrics and geographically weighted regression (GWR), with a case study in Xuzhou city, China, to analyse the spatial patterns of urban growth as well as their spatiotemporally varying relationships with three driving factors (1) slope, (2) distance to major urban centres, and (3) distance to major roads. The historical urban growth from 1990 to 2010 was derived from multi-temporal remote sensing images. Spatial metrics were used to quantify the urban growth patterns for different periods. The effects of the factors on urban growth patterns were further investigated using GWR. The results indicate that the spatial patterns of Xuzhou have significantly changed along the urbanization process. GWR performs better than ordinary least squares (OLS) in interpreting the relationships indicated by higher adjusted R-2, lower corrected Akaike information criterion (AICc) values and reduced spatial autocorrelations of residuals. The parameters of the driving factors obtained from GWR indicate that their effects on spatial patterns are spatiotemporally varying. The findings help in better understanding the effects of the considered factors on spatial patterns, as well as to provide support for urban planning and management.

Spatiotemporally Varying Relationships Between Urban Growth Patterns and Driving Factors in Xuzhou City, China