Abstract

Development of DEMs from LiDAR Data for Watershed Delineation
Track: Water Resources
Authors: Francisco Olivera, Celso Ferreira, Dean Djokic

The use of Light Detection and Ranging (LiDAR) data to develop high-resolution Digital Elevation Models (DEM) is becoming standard in hydrologic studies. However, the advantages of increasing the resolution are partially offset by the increased file size, difficulty to handle them, slow screen rendering and increase computational effort. Terrain Dataset - developed by ESRI -- is an alternative format for storing topographic data that intelligently decimates data points and creates simplified, but equally accurate, DEMs. In this study the impact that the thinning method (window size or z-value), pyramid level and the interpolation technique (linear or natural neighbor) used to create the DEMs have on the watersheds delineated from them is evaluated. Case studies based on LiDAR datasets from Florida, Texas, Colorado and Maryland are presented. Based on standardized performance metrics (watershed accuracy and computational effort) and topographic characteristics, guidelines for developing DEMs from LiDAR data are proposed.