The Effect of DEM Quality on Sea Level Rise Exposure Analysis

Abstract:

Sea level rise (SLR) caused by climate change could cause significant disruptions in coastal communities across the world. Current projections estimate that we may see in the vicinity of 1 meter of SLR by the end of the century, and due to collapsing ice sheets in West Antarctica, more than 3 meters of global SLR appear very likely in the long run. It is therefore crucial that we begin to accurately understand both the short- and long-term effects this level of flooding could have on each country's land area and population. However, while we have high-resolution digital elevation models (DEMs) publicly available for some parts of the world, such as the coastal lidar datasets distributed by NOAA for the US, most of the rest of the world is only covered by much poorer-quality data, such as data from SRTM (3 arcsec, or roughly 90m, horizontal resolution). In this work, we perform SLR analysis using both NOAA lidar- and SRTM-derived DEMs in the United States, in order to understand how using low-quality DEMs affect the final analysis results. We find that in many states, the computed population exposure at 1 meter SLR is over 2x higher when using the Lidar DEM, compared to the results computed from SRTM. In addition to the clear differences in horizontal resolution, this very large difference in computed exposure could likely be explained by the fact that SRTM is based on surface elevation, while the Lidar DEM is based on bare earth elevation. We therefore conclude that any worldwide SLR analysis using SRTM would produce exposure estimates that are far too low, and higher-quality global DEMs are necessary in order to generate exposure analysis of reasonable accuracy.