المجلة الدولية للعلوم والتقنية

The accuracy of generated terrain model depends on the interpolation mechanism adopted. A point-source data to provide a precise description of the morphology needs different data grid interpolation to make up the vacancy between the data sources. This paper presents a comparison of different kinds of data grid interpolation methods-based on Surfer Software11, and artificial neural network (ANN) for estimating spatial continuous data. A local geoid model was first generated according to interpolation methods from the geoid leveling obtained by using GPS/Levelling data for a study area. Subsequently, a (ANN) model was used to generated the local geoid model of the study area with the same data. The selected interpolation methods and ANN were evaluated, in terms of efficiency factor E, mean absolute percentage error MAPE% also the minimum and maximum value. A direct comparison between contouring, 3D surface mapping , volumes and areas for the surfaces, was discussed and evaluated. Concluded that ANN can be used for local leveling modelling as an alternative interpolation method with high accuracy, in terms of MAPE%=0.09 and E =0.97. ANN, has the potential to provide accurate predictions and site profiling contour and 3D surface maps.