Notas sobre comparações entre distintos modelos digitais de elevação

Abstract

The growing interest and use of Cartographic Digital Products (CDPs) in many activity areas has fomented a real revolution in Geographical Information Systems (GIS) technologies and methodologies. To adapt to these changes, some national and international norms and standards has been proposed as way to regulate and generate CDPs as well as to evaluate their accuracy. In the present work, Digital Elevation Models (DEMs) of distinct acquisition sources (VANT, ASTER, SRTM and TOPODATA) are compared for their reliability to control points (collected with total station and geo-referenced with GNSS RTK). The Brazilians 1984’s Cartographic Accuracy Standard (PEC) and 2016’s Cartographic Accuracy Standard to CDPs (PEC-PCD) was also compared for their feasibility and adequacy. It was observed here that commonly neglected cares, as the complete equalization of the entry data and of the distinct analysis tools (the ArcGis software tools, in this case) are indispensable measures for distinct DEMs shows the best fidelity to the original data and, consequently, the lower systematic error. In this sense, it was observed that the ArcGis’s clipping tool “Mask” and the point extraction tools “3D” and “Sample” generate significantly different values of altitude than the topographic ones. The VANT Digital Terrain Model (VANT DTM) showed the best accuracy between all CDPs followed by the ASTER, SRTM, VANT DMS and TOPODATA Digital Surface Models (DSMs). Finally, the PEC-PCD was most adequate than PEC to evaluate DEMs.