Interpolation methods for geographical data: Housing and commercial establishment markets

The estimation of commercial property prices in a touristic city can be explored through spatial interpolation methods, but in the presence of small sample sizes, auxiliary stochastic processes that are correlated with the prices of commercial establishments are needed. The aim of this paper is to compare the various estimates of commercial establishment prices in Toledo (Spain) provided by methods based on inverse distance weighting, 2-D shape functions for triangles, kriging and cokriging (the housing prices being the auxiliary stochastic process). The results indicate that kriging improves the classical interpolation methods and that cokriging has a clear advantage over kriging.