Assessment and comparison of Digital Elevation Model (DEM) products in varying topographic, land cover regions and its attribute: a case study in Shikoku Island Japan

Abstract

Digital Elevation Model (DEM) is the model of Earth’s surface and is significantly variable for science application. The DEMs based on free provided are the 10 m DEM produced by the Geographical Survey Institute of Japan (GSI-DEM), Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) Global DEM, Shuttle Radar Topography Mission (SRTM), Global Multi-resolution Terrain Elevation Data 2010 (GMTED2010), Hydrological data and maps based on Shuttle Elevation Derivatives at multiple Scales (HydroSHEDS), and Global 30 Arc-Second Elevation (GTOPO30) that are normally used in scientific researches. These terrain data have been generated by different methods, such as Stereoscopic Photogrammetry, RADAR-SAR interferometry, LIDAR, and GPS. The generated data contain an error from collecting and processing methods that have related to morphology relief and land cover type. In this study, the six open-source DEMs and their attributes are evaluated by the referent elevation points observed by GPS, in Shikoku Island Japan. It was revealed that the error of DEMs is an effect on the terrain characteristic. Overall, the accuracy values of fine-resolution DEMs based on the RMSE value are 5.9 m, 9.9 m, and 10.1 m for GSI-DEM, ASTER, and SRTM, respectively. The coarse-resolution DEMs revealed the accuracy of RMSE as 18.2 m, 69.4 m, and 61.8 m for GMTED2010, HydroSHEDS, and GTOPO30, respectively. The accuracy of the slope and river network is also evaluated. This approach might be used to recommend for a new generation of DEM and revealed for the accuracy of the Earth’s surface.