The article discusses a method to accurately classify elevation changes using a combination of surface characteristics. By merging these characteristics, a more detailed representation of elevation changes can be created, helping to select the best elevation model and modeling parameters. The study found that errors in elevation models depend on the geomorphologic features of the area, and different methods were used to evaluate these errors. By adjusting modeling parameters, it was possible to reduce the standard deviations of the elevation model, improving its accuracy.