On the Accuracy of Geoid Heights Derived from Discrete GNSS/Levelling Data Using Kriging Interpolation

Abstract Local geoid models presenting higher resolution than global ones are generally derived by a combination of different datasets, integrating individual pure astrogeodetic, gravimetric and GNSS/levelling solutions. To define local geoid, interpolators may be applied starting from dataset height values. It is well known that the accuracy resulting depends not only interpolation method, but also points numerosity distribution. This article aims to analyse performance Kriging approaches in dependence density dataset. The experiments carried out on heights extracted random way an already existing model: subsets organized containing increasing number same area each them submitted interpolations (Universal Ordinary Kriging). compared with original one residuals calculated evaluate point density. results demonstrate efficiency methods, highlighting possibility achieve (a few centimetres) using 1 point/100 sqkm, absence gravity anomalies. provides better Universal undulations between minimal millimetres) when high involved. Furthermore, highlight limit leave Cross validation since it supplies direct comparison for both Kriging, used.