Scholarlink

<p>Most single-column, Gaussian dispersion-model algorithms exclusively consider the time-averaged Reynolds shear stress contribution to the friction velocity, neglecting any dispersive stress component inherently present from the lateral spatial averaging of the model. This work examines the impact of incorporating dispersive stresses when calculating the friction velocity used for dispersion parameterizations in idealized uniform and nonuniform urban areas. This is accomplished through wind tunnel-validated, neutrally-stratified Large Eddy Simulations (LES) at incoming wind flow angles of 0°, 10°, 30°, and 50°. Comparisons are made to the existing formulations in AERMOD, the EPA’s preferred Gaussian dispersion model. Given the issues with Gaussian models in complex environments, results are focused above the urban canopy, where the vertical shear stress from Reynolds and dispersive components are used to estimate a friction velocity to fit the log law velocity profile with surface roughness and displacement height. An urban model with rows of uniform buildings was found to feature negligible dispersive components to the friction velocity outside of any city blocks within a transition region where the dispersive shear stress was opposite in sign to the Reynolds stress. A second urban model containing nonuniform buildings featured meaningful dispersive stresses for nearly all city blocks and flow angles. The friction velocity and fitted surface roughness for all blocks of both models were compared to existing reduced order models used to predict urban friction velocity. The method used in AERMOD relies on a convective-like boundary layer assumption and has poor agreement with our results, whereas a simple power law expression reproduces all our data within 9.6% if the height at which the velocity matches from both the upwind to the urban wind profiles is known. Finally, the inclusion of dispersive shear stress in the friction velocity is reflected in the prediction of pollutant concentrations in the AERMOD formulations, leading to a + 20.6% and + 21.2% increase in predicted concentrations within the first row for the uniform and nonuniform cases, respectively, but otherwise a + 0.4% and − 15.6% change for all other rows. Future work could incorporate these effects in conjunction with the physical presence of buildings to improve the predictive power of Gaussian models in urban areas.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Including Dispersive Shear Stress in Urban Environments for Single Column Dispersion Models

  • 期刊论文

  • 发表日期：

    2025年8月21日

查看全文
  • Jonathan Retter,
  • David Heist,
  • Michael Pirhalla,
  • R. Chris Owen,
  • Wei Tang,
  • Terrance Odom,
  • Lydia Brouwer

摘要

Most single-column, Gaussian dispersion-model algorithms exclusively consider the time-averaged Reynolds shear stress contribution to the friction velocity, neglecting any dispersive stress component inherently present from the lateral spatial averaging of the model. This work examines the impact of incorporating dispersive stresses when calculating the friction velocity used for dispersion parameterizations in idealized uniform and nonuniform urban areas. This is accomplished through wind tunnel-validated, neutrally-stratified Large Eddy Simulations (LES) at incoming wind flow angles of 0°, 10°, 30°, and 50°. Comparisons are made to the existing formulations in AERMOD, the EPA’s preferred Gaussian dispersion model. Given the issues with Gaussian models in complex environments, results are focused above the urban canopy, where the vertical shear stress from Reynolds and dispersive components are used to estimate a friction velocity to fit the log law velocity profile with surface roughness and displacement height. An urban model with rows of uniform buildings was found to feature negligible dispersive components to the friction velocity outside of any city blocks within a transition region where the dispersive shear stress was opposite in sign to the Reynolds stress. A second urban model containing nonuniform buildings featured meaningful dispersive stresses for nearly all city blocks and flow angles. The friction velocity and fitted surface roughness for all blocks of both models were compared to existing reduced order models used to predict urban friction velocity. The method used in AERMOD relies on a convective-like boundary layer assumption and has poor agreement with our results, whereas a simple power law expression reproduces all our data within 9.6% if the height at which the velocity matches from both the upwind to the urban wind profiles is known. Finally, the inclusion of dispersive shear stress in the friction velocity is reflected in the prediction of pollutant concentrations in the AERMOD formulations, leading to a + 20.6% and + 21.2% increase in predicted concentrations within the first row for the uniform and nonuniform cases, respectively, but otherwise a + 0.4% and − 15.6% change for all other rows. Future work could incorporate these effects in conjunction with the physical presence of buildings to improve the predictive power of Gaussian models in urban areas.