Robust Principal Component Analysis and Geographically Weighted Regression: Urbanization in the Twin Cities Metropolitan Area (TCMA)

In this paper, we present a hybrid approach, robust principal component geographically weighted regression (RPCGWR), in examining the land change as a function of both extant urban land use and the effect of social and environmental factors in the Twin Cities Metropolitan Area (TCMA) of Minnesota. We used remotely sensed data to treat urban land use via the proxy of impervious surfaces. We then integrated two different methods, Robust Principal Component Analysis (RPCA) and Geographically Weighted Regression (GWR) to create an innovative approach to model urban land use. The RPCGWR results show significant spatial heterogeneity in the relationships between proportion of impervious surface and the explanatory factors in TCMA. We link this heterogeneity to the ‘sprawling’ nature of land change that has moved outward from the core Twin Cities through to their suburbs and exurbs.