Abstract: After studying on the shape change in a group of spatio temporal data for land use, the authors concludes that there are only four basic types of shape change and the time slices for each case of shape change can be interpolated by curve interpolation and shape interpolation. Thus, a powerful interpolation algorithm for shape interpolation is needed in order to realize the visualization of shape change. The algorithm is able to interpolate for both curve and polygon. The commonly used interpolation algorithm can not achieve this, because it is a local interpolation algorithm. After investigating into the interpolation algorithm deeply, the authors concludes that a global interpolation algorithm based on potential field theory can be used for shape interpolation in the visualization of spatio-temporal data for land use.