Exploring the capability of some GIS surface interpolators for DEM gap fill

Scene occlusions in aerial or satellite imagery are at times an unavoidable occurrence resulting from natural or mechanical effects. At their most sev ere, occlusions can render terrain modeling and feature extraction tasks in complete or impossible in affected regions of an image, creating gaps of mi ssing information in the extracted data sets. In lieu of alternative source s, a simple solution for gap fill of digital elevation data may be to perfo rm interpolation, provided the gap is not impracticably large. The expedien cy that interpolation offers must be appropriately justified vis-g-vis the potential for substantial uncertainty in the estimates. This study evaluate s the gap fill capabilities of four ESRI Arc/Info offered surface interpola tor functions: (1) kriging, (2) splining, (3) inverse distance weighting, a nd (4) surface trend analysis. Using a simple error assessment approach wit h different terrain textures, the evaluation tends to favor the use of the splining function when interpolating for the kind of gaps and sample config urations considered.