15th Conference on Boundary Layer and Turbulence

Thursday, 18 July 2002: 5:14 PM
Observed and simulated turbulence kinetic energy and dissipation profiles in an urban valley during VTMX 2000
Jerome D. Fast, PNNL, Richland, WA; and W. J. Shaw
A month long meteorological field campaign sponsored by the Department of Energy’s Environmental Meteorology Program was conducted in the Salt Lake Valley during October 2000 to study vertical transport and mixing (VTMX) processes. Measurements made during the 2000 VTMX field campaign focused on nocturnal stable periods and morning and evening transition periods within the urban valley. A wide range of remote sensing and in situ measurements were made including those from surface meteorological stations, temperature data loggers, sonic anemometers, radiosondes, radar wind profilers, sodars, and lidars. Thermally-driven circulations, including nocturnal down-valley and canyon flows along the Wasatch Mountains, were regularly observed when the synoptic forcing was weak. In this study, a mesoscale model that employs the Mellor-Yamada 2.5 level turbulence closure and a 500-m horizontal grid spacing is used to simulate the evolution boundary layer in the Salt Lake Valley. The simulated mean conditions are evaluated using temperature, humidity, and wind profiles obtained from radiosondes and radar wind profilers. The model reproduced many of the observed mean conditions; however, to adequately describe the vertical mixing of pollutants the model must also produce realistic turbulence quantities. Therefore, the simulated profiles of turbulence kinetic energy and dissipation are compared with data derived from a radar wind profiler located in the center of the valley and sonic anemometers at several locations in the valley. The model produced the observed diurnal variation in turbulence kinetic energy, with higher values during the day and lower values at night. Nevertheless, there were periods where the simulated turbulence kinetic energy was significantly higher or lower than observed even though the simulated and observed mean conditions were similar. A discussion of modifications to the turbulence parameterization will be made.

Supplementary URL: