Assessing the Utility of Three Calibration Methods for Digital Elevation Models (DEMs)

Permafrost degradation in the Arctic is negatively effecting the environment, structures, and livelihoods of residents in Arctic areas. Melting of permafrost from increased global temperatures causes inconsistent land subsidence and compromises infrastructure, including buildings, roads, and energy pipelines. ArcticDEM data, created by the Polar Geospatial Center from DigitalGlobe, Inc. imagery, is being used to study permafrost degradation, including long-term change, in the North Slope Borough, Alaska. To utilize Arctic DEM for this purpose, it must be calibrated. This study examines the utility of three methods of calibration of ArcticDEM data from each year of 2010, 2011, 2012, 2015, 2016, and 2017. The first method is to calibrate the data using internal referencing, assuming the majority of the area has not changed in elevation and adjusting subsequent years to the first data strip collected in 2010. The second method identifies rock outcrops within the study area and assumes that these outcrops do not change in elevation. The DEMs are all adjusted so the rock outcrops are consistently the same elevation throughout the time series. The third method utilizes an extraneous data source, e.g. LiDAR data from the 2010 USGS North Slope of Alaska LiDAR project, obtained through the NOAA Data Access Viewer. The LiDAR data is compared with the ArcticDEM data strips, and the ArcticDEM are shifted to match the elevation of the LiDAR data. The results of all three calibration methods are compared for precision and accuracy.