Wind and Diffusion Modeling for Complex Terrain

Abstract

Atmospheric transport and dispersion over complex terrain were investigated. Meteorological and sulfur hexafluoride (SF6) concentration data were collected and used to evaluate the performance of a transport and diffusion model coupled with a mass consistency wind field model. Meteorological data were collected throughout April 1995. Both meteorological and plume location and concentration data were measured in December 1995. The meteorological data included measurements taken at 11?15 surface stations, one to three upper-air stations, and one mobile profiler. A range of conditions was encountered, including inversion and postinversion breakup, light to strong winds, and a broad distribution of wind directions. The models used were the MINERVE mass consistency wind model and the SCIPUFF (Second-Order Closure Integrated Puff) transport and diffusion model. These models were expected to provide and use high-resolution three-dimensional wind fields. An objective of the experiment was to determine if these models could provide emergency personnel with high-resolution hazardous plume information for quick response operations. Evaluation of the models focused primarily on their effectiveness as a short-term (1?4 h) predictive tool. These studies showed how they could be used to help direct emergency response following a hazardous material release. For purposes of the experiments, the models were used to direct the deployment of mobile sensors intended to intercept and measure tracer clouds. The April test was conducted to evaluate the performance of the MINERVE wind field generation model. It was evaluated during the early morning radiation inversion, inversion dissipation, and afternoon mixed atmosphere. The average deviations in wind speed and wind direction as compared to observations were within 0.4 m s?1 and less than 10° for up to 2 h after data time. These deviations increased as time from data time increased. It was also found that deviations were greatest during inversion dissipation. The December test included the release and tracking of atmospheric tracers. The MINERVE?SCIPUFF modeling system was used to direct remote sensing equipment. Posttest analyses were performed to determine model reliability. It was found that plume centroid position as determined by the models was within 10% of the observed plume centroid.