Digital ElevationModel(DEM)refers to a digital map of the surface of the Earth that only shows the bare ground,without any buildings,plants,or other characteristics.However,obtaining unlimited access to DEM data at hi...Digital ElevationModel(DEM)refers to a digital map of the surface of the Earth that only shows the bare ground,without any buildings,plants,or other characteristics.However,obtaining unlimited access to DEM data at high and medium resolutions is very hard.Consequently,users often question the accuracy of freely available DEMs and their suitability for various applications.By comparing them to Global Positioning System(GPS)elevation data,this study aimed to identify themost reliable and widely available DEM for various terrains.The objectives of this study were to generate DEMs fromdifferent open sources and validate the accuracy of these DEMs using GPS elevation data.Various DEM types including Sentinel-1,ALOS PALSAR,SRTM,AW3D30,and ASTER were compared.Root Mean Square Error(RMSE)andMean Error(ME)were used to measure the difference between the DEM-derived elevations and the GPS-measured elevations.The results showed that even though Sentinel-1 has higher resolutions,the accuracy of the DEM from Sentinel-1 depends on issues including coherence and interferometry,surface features,and temporal stability.On the other hand,ALOS PALSAR could accurately represent surfaces in some situations.Additionally,DEMs with lower resolutions,such as SRTM and AW3D30,demonstrated greater consistency across various types of terrain.In contrast,the ASTER DEM showed more variability in complex terrains.While freely available DEMs are easy to use and accessible,their accuracy varies depending on the source and terrain features.Future improvements could include adding more ground control points and using advanced filtering methods to enhance precision.展开更多
基金funded by the Ministry of Higher Education Malaysia(MOHE)through the Fundamental Research Grant Scheme(FRGS/1/2021/WAB07/UiTM/02/1).
文摘Digital ElevationModel(DEM)refers to a digital map of the surface of the Earth that only shows the bare ground,without any buildings,plants,or other characteristics.However,obtaining unlimited access to DEM data at high and medium resolutions is very hard.Consequently,users often question the accuracy of freely available DEMs and their suitability for various applications.By comparing them to Global Positioning System(GPS)elevation data,this study aimed to identify themost reliable and widely available DEM for various terrains.The objectives of this study were to generate DEMs fromdifferent open sources and validate the accuracy of these DEMs using GPS elevation data.Various DEM types including Sentinel-1,ALOS PALSAR,SRTM,AW3D30,and ASTER were compared.Root Mean Square Error(RMSE)andMean Error(ME)were used to measure the difference between the DEM-derived elevations and the GPS-measured elevations.The results showed that even though Sentinel-1 has higher resolutions,the accuracy of the DEM from Sentinel-1 depends on issues including coherence and interferometry,surface features,and temporal stability.On the other hand,ALOS PALSAR could accurately represent surfaces in some situations.Additionally,DEMs with lower resolutions,such as SRTM and AW3D30,demonstrated greater consistency across various types of terrain.In contrast,the ASTER DEM showed more variability in complex terrains.While freely available DEMs are easy to use and accessible,their accuracy varies depending on the source and terrain features.Future improvements could include adding more ground control points and using advanced filtering methods to enhance precision.
