ABSTRACT
Chapter 3 covers two generic tasks in GIS-based spatial analysis: spatial smoothing and spatial interpolation. The two are closely related and useful for visualizing spatial patterns and highlight spatial trends. Some methods (e.g., kernel density estimation or KDE) can be used for either spatial smoothing or interpolation. Spatial smoothing computes the averages using a larger spatial window to smooth the spatial variability, whereas spatial interpolation uses known values at some locations to estimate unknown values at other or any locations. A case study of mapping place-names in southern China illustrates some basic spatial smoothing and interpolation methods, such as the floating catchment area (FCA), KDE, and Inverse Distance Weighting (IDW) methods. A separate section focuses on area-based spatial interpolation, which transforms data in one areal unit to data in other areal units. This chapter also extends the popular KDE analysis to spatiotemporal KDE (STKDE) by adding a temporal dimension. A set of spatiotemporal visualization tools in ArcGIS Pro enables us to visualize temporal trends, spatial patterns, and the intersection between them with a space–time cube in 3D.
