COMPARATIVE PERFORMANCE ANALYSIS OF INTERPOLATION TECHNIQUES AND DGPS MEASURED HEIGHTS

Abstract

The main objective is to evaluate heights interpolation methods, and evaluate accuracy measures from observed and predicted data to find the best fit model. This research focuses on the application of four interpolation methods (IDW, Kriging, Natural Neighbor and Spline) to estimate the unknown value from the digital elevation models (DEM). Results from this study show the interpolation using GIS techniques is effective and has a higher level of accuracy compared to conventional methods, especially in the areas with undulating terrain. The IDW interpolation method has a SD of 0.230, SE as 0.0048, MSE of 0.0542, RMSE as 0.232882 and MAE of 0.110. This was closely followed by the KG interpolation method the the NN model and finally the SP model respectively and the range of error across the different interpolation methods over the control points ranges between -0.5 to +5.7m. Station CP10 has +0.5783 while MB03 was -0.585 and CP4 has no or neglegeable error of 0.000000m. This study concluded that the best interpolation method was IDW while KG and NN followed. The SP model performed poorly based on this research