Simulating urban growth processes incorporating a potential model with spatial metrics

► Built-up density, defined according to the PLAND value obtained through the moving window method, was used to quantify and categorize complex urban built-up landscapes into simple, quantifiable and identifiable patterns. ► Cost-distance was used instead of the straight-line distance to optimize the urban growth potential model and increase the accuracy of the simulation results. ► Integration of urban growth potential model with spatial metrics will extend the application value of both landscape metrics and the Potential Model.