Interpolation & Surfaces

Interpolation produces a value for every cell, including the cells a kilometre from the nearest observation, and the output gives no hint which is which. These guides cover the methods — IDW, kriging, TIN and areal interpolation — but spend as much time on what they cannot tell you: why cross-validation is optimistic where the samples are dense, why kriging variance is a map of sample geometry rather than of error, why a finer cell size buys nothing, and how to ship a distance-to-nearest-sample raster so the reader can see where the surface is measurement and where it is arithmetic.