ELEVATION ERROR COMPARISION OF COMMONLY USED DIGITAL EVATION MODELS (DEM) FOR NATURAL RESOURCE MANAGEMENT WITH GROUND CONTROL POINTS

Land surveying has importance in monitoring, planning and detailed mapping for natural resources in lands. Consider the success in management plans, especially in area of engineering, more precious data employed to promote high-efficiency and cost optimization of projects. Engineering projects require more precision and attention in forest lands having huge part of natural resources. Otherwise engineers may cause irreversible damage to nature in addition to wasting money and time. Field oriented engineering works, especially, digital terrain models (DTM) are one of the most important data. Thus, this study examines vertical accuracies of open-source DEMs. Thus, this study primarily evaluates the DEMs mentioned in many research papers. Within the scope of this study, Advanced Spaceborne Thermal Emission Reflectometer Global DEM (ASTER GDEM 2), Shuttle Radar Topography Mission (SRTM (1) V3) at 30 m resolution, and DEMs derived from orthophotos, which are commonly used by researchers, were compared. The validation of DEMs investigated by using standard deviation (STD), mean error (ME) and the root mean square error (RMSE) to investigate errors in elevation data. Surveying for 116 reference points in the test-site, which has barren and poorly vegetated (Grass/Shrub) land surfaces, was carried out by using total station. The precision of the orthophoto-derived DEM’s error, is better and more reliable than that of other DEMs in bare lands, such as forest openings. The investigated RMSE values were 1.62 m for the orthophoto-based DEM, 2.68 m for SRTM 1 V3 and 11.79 m for ASTER GDEM 2. The orthophoto-based DEM showed higher accuracy than satellite-based models.