Estimation of vertical accuracy of digital terrain model FABDEM on the territory of the republic of Serbia

Abstract

In 2004, a digital terrain model (DTM) was created in Serbia. It was obtained by vectorizing isohypses on topographic maps with a scale of 1: 25000. The resolution of the digital model of Serbia is 30 m, and the vertical accuracy, as the authors claim, is 10 m. From December 2021, the global model FABDEM (Forest And Buildings removed Copernicus DEM) is available to the public. It was created based on the data of the Copernicus GLO30 model so that data on forests and buildings were removed from it. Important features of the global model are its horizontal resolution of 30 m and vertical accuracy of less than 5 m. This paper aims to assess the vertical accuracy of the global FABDEM model and analyze the differences between the global FABDEM and DTM in the territory of the Republic of Serbia. Vertical accuracy and analysis of the two models were assessed by comparing the heights derived from FABDEM and DTM with the heights of the independent set of points of the gravimetric survey of Serbia (GSS). The heights of the GSS points were determined using classical geodetic methods, and they are relatively uniformly distributed on the territory of the Republic of Serbia. A short description of mentioned comparison is presented in this paper. Statistical analysis results that compare terrestrial heights with heights derived from DTMs show minor differences between the two data sets.