Geographically Weighted Regression (GWR) is a statistical technique that allows variations in relationships over space to be measured within a single modeling framework. The output from GWR is a set of surfaces, each surface depicting the spatial variation of a relationship.