Subsampling Large-Scale Digital Elevation Models to Expedite Geospatial Analyses in Coastal Regions

Abstract

Murphy, K.A.; Zawada, D.G., and Yates, K.K., 2022. Subsampling large-scale digital elevation models to expedite geospatial analyses in coastal regions. Journal of Coastal Research, 38(6), 1236–1245. Coconut Creek (Florida), ISSN 0749-0208. Large-area, high-resolution digital elevation models (DEMs) created from light detection and ranging (LIDAR) and/or multibeam echosounder data sets are commonly used in many scientific disciplines. These DEMs can span thousands of square kilometers, typically with a spatial resolution of 1 m or finer, and can be difficult to process and analyze without specialized computers and software. Such DEMs often can be subsampled to expedite analysis with negligible impact on results for large-scale geospatial analyses. Subsampling can be achieved by creating a grid of points that specify the locations from which to extract elevation values from the DEM. This paper presents a method that can be used to accurately perform subsampling of large-scale, high-resolution DEMs using GIS software. This subsampling method was applied to two LIDAR-derived DEMs encompassing 242 km2 of the northern Florida Reef Tract as an example application and to test subsampling accuracy. Results indicate that subsampling 1-m-resolution DEMs using a 2-m-spaced grid results in no significant difference in mean elevation or other basic statistics for analyses performed over multiple spatial scales ranging from 1 km2 to 242 km2.