Comparing 3D Lidar Geometry to Vetted Urban Geometry Via CT-Analyst (registered trademark)
Abstract
Airborne lidar measurements of a 3D urban region provide a digital geometry database of 3D buildings and urban terrain that is within a few percent of a vetted geometry database painstakingly assembled from satellite or aircraft stereo imagery. The pacing item in preparing CT-Analyst (registered trademark) contaminant transport software for real time use in a new urban region has been the labor-intensive weeks to months needed to assemble the 1-meter resolution 3D geometry used to precompute the building influences on urban wind fields. The assembly of a good geometry database from lidar now can take only a day or two. NRL CT-Analyst force-protection software predicts the transport and dispersion of airborne contaminants over complicated-geometry landscapes. This fast (milliseconds) laptop system is also accurate because it is based on 18 wind field maps called NomografsTM that are precalculated for a built-up region. A set of nomografs typically covers a 12 km by 12 km urban region at 5-meter resolution. These nomografs are derived from 3D Computational Fluid Dynamics simulations taking a couple of days on a modest high-performance parallel cluster. This report compares the CT-Analyst contaminant density predictions, using 2-meter resolution nomografs, for a 4 km by 4 km area of downtown Denver based on an existing vetted 3D geometry and the corresponding geometry derived from airborne lidar measurements. The IDA Measure of Effectiveness (MoE), a cloud-to-cloud comparison metric computed at specific times, is used to compare predictions based on the two different geometry datasets. This report shows that lidar-based geometry allows good quality CT-Analyst predictions with much less investment in labor and preparation lead time.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 26, 2023
- Accession Number
- AD1194595
Entities
People
- Adam J. Moses
- Gopal Patnaik
- Jay Boris
- Keith S. Obenschain
- Kiran M. Donnelly
Organizations
- United States Naval Research Laboratory