Spatial interpolation of elevation data with variable density: A new methodology to derive quality DEMs

A new methodology for spatial interpolation of elevation data with variable density is proposed. The method is based on two-step interpolation and data processing to minimize interpolation artifacts caused by variable data density. In the first step, the parameterization of the spline interpolation method is focused on areas with sparse data that need smoother interpolation. Then, the resulting surface in these areas is randomly sampled to densify the original data set. In the second step, the parameterization of the interpolation method is focused on areas with a required high level of detail. The final resulting surface contains the properties of surfaces optimized for different data densities and levels of detail.