GIS surface interpolations are pure linear calculations, but many nonlinear geographic phenomena exist in the real world. A compass, with 360 degrees, describes a nonlinear situation. The same final direction may be reached by moving in either a clockwise or counterclockwise direction. How can nonlinear conditions be modeled in linear-restricted programs? Translation functions may be used successfully, but care must be taken when designing the model. Potential conflicts and sign problems await the unwary GIS analyst. This paper uses the specific example of calculating two-dimensional wind directions over flat terrain to illustrate the pitfalls and pinnacles of nonlinear interpolations.