摘要
Rising sea levels due to global warming and climate change impact may prove a disaster for small islands. Accurate DEM （digital elevation model） can help to understand SLR （sea level rise） impact, coastal zones flooding risks assessment and hydrological attributes modeling and extraction. Currently, DEMs are available from several different sources using active and passive remote sensing systems. This research compares absolute surface heights accuracies retrieved from three independent DEMs datasets. The Shuttle Radar Topographic Mission （SRTM-V4.1） and the Advanced Space borne Thermal Emission and Reflection Radiometer （ASTER-V2.1） with 30-m pixel size, and a DEM-5 of 5-m spatial resolution generated from high topographic contour lines map at scale of 1：5,000 using simple Kriging interpolation method. Moreover, topographic attributes （slope and aspect） have been retrieved and compared. For the elevations validation purposes, a dataset of 400 GCPs uniformly distributed over the study site were used. These were measured using a DGPS assuring ± 1 and ± 2 cm accuracies, respectively, for planimerry and altimetry. The obtained results show that globally the landscape scale plays an important role in the selection of the DEM pixel size, which must reflect the real topographic attributes. Indeed, the derived DEM-5 from high topographic contours map （1：5,000） using simple Kriging exhibit the best accuracy of ±0.65 m which is less than the tolerance or the total error （±0.78 m） calculated based on errors sources propagation. Then, the results show an accuracy of ± 3.00 m for SRTM-V4.1 which is less than the absolute vertical height accuracy （±5.6 m） advocated by NASA for African continent and Middle-East regions. As well, the achieved ASTER accuracy was ± 8.40 m compared to the estimated error （±17.01 m） by USGS and JAXA. Obviously, high spatial resolution and accurate DEM-5 is a crucial requirement to simulate and evaluate costal zones inundation under different SLR and storm flow scenarios for small islands. Decidedly, the elevation of small islands with topographic features not higher than 134 m can be estimated using SRTM-V4.1 with relatively acceptable accuracy. Whereas, this DEM is not significantly consistent for accurate SLR scenarios simulations. Without doubt, ASTER-V2.1 DEM was an excellent alternative compared to SRTM with 90-m pixel size, but actually with SRTM-V4.1 full resolution （30-m） ASTER-V2.1 will likely see its limited uses in geosciences applications. Indeed, ASTER is not providing accurate information to simulate the impact of SLR scenarios on small islands.
