Comparison Of Digital Surfacemodelling Techniques For Slopingrnhill Terrain Using Gps Data

摘要

High resolution (cm level accuracy) digital terrain models (DTM) are needed for slope evaluation studies and landslide monitoring applications. The present paper investigates the efficacy of some commonly used surface interpolation techniques to generate the DTM of a sloping hill terrain influenced by an active landslide. The point co-ordinates acquired using kinematic Global Positioning System (GPS) survey are used as input data and co-ordinates of some identified control points obtained using static GPS survey are used as validation points. The main focus is on identifying the best-fit technique for terrain modelling in terms of accuracy of generated surface, which can possibly be used for quantifying temporal changes in slope parameters of steep mountainous terrains. Nowadays, for geodesy applications, one of the most productive methods is use of precise point positioning obtained through GPS. Interpolation techniques like kriging, inverse distance weightage (IDW), spline, natural neighbour and radial basis function (RBF) are tested for accuracy of the generated surface. These methods give different levels of accuracy depending upon the precision of survey, terrain complexity, sampling density and the specific parameters applied to the method. Statistical analysis reveals that the surface reconstruction accuracy varies from 0.01 to 87 cm. The RMSE varies from 20 to 35 cm. Variogram analysis and surface accuracy plots reveal that RBF and kriging methods best represent the surface complexities at cm level accuracies.