We propose a 4-year comprehensive modeling program to 1) evaluate and improve the representation of topography in mesoscale models and its effect on simulations of vertical exchange processes in urban basins or valleys, and 2) evaluate existing subgrid-scale turbulence parameterizations and develop improved treatments for use in urban basins or valleys. Our hypothesis, which forms the basis of our proposed research, is that the inability of current mesoscale models to accurately predict many types of vertical exchange processes in basins and valleys is due primarily to inadequate representation of topography and subgrid-scale turbulent motions. Although a number of factors may contribute to forecast errors and new conceptual or numerical approaches may be required to significantly improve the prediction of vertical mixing processes, our approach is to focus on these two major aspects of the problem with the goal of improving the ability of mesoscale models to simulate processes affecting vertical transport and mixing in urban basins and valley.
The approach we will use to achieve these objectives consists of a systematic set of model simulations and comparisons with data. Two mesoscale models will be used for the proposed work: the RAMS model and the NCEP's Eta model. Data sets to be used to evaluate the model simulations include those from several previous boundary-layer field campaigns in urban basins, e.g., the IMADA-AVER experiment in Mexico City, the field campaign in Phoenix, and the Southern California Ozone Study (SCOS-97) in Los Angeles, as well as future field campaigns conducted by the VTMX program. We also plan to collaborate with CASES-99 investigators to evaluate the performance of turbulence parameterizations over flat terrain . In this way, we hope to determine how well various parameterizations, developed for the stable boundary layer over flat terrain, simulate turbulent quantities in areas of complex terrain.
The first objective involves a critical investigation of numerical treatments of topography that focuses on the vertical coordinate system, terrain resolution, and development of a subgrid-scale terrain parameterization. Our previous experience with mesoscale modeling in areas of complex terrain has led us to suspect that the use of a traditional 'terrain-following' vertical coordinate as in many mesoscale models to accommodate sloping terrain may be a fundamental reason for the failure of mesoscale models to adequately predict terrain-induced flows and boundary-layer development in intermountain basins and valleys. The newer approach using the 'step-mountain' coordinate has shown promising results for simulations at synoptic scales, but its advantages at the mesoscale in basins and valleys have yet to be demonstrated. To determine how much of the model forecast error can be attributed to the use of a terrain-following coordinate and whether the eta coordinate will lead to consistently superior forecasts of these circulations, we will perform identical simulations using both the 'eta', and the 'sigma' coordinate and evaluate the results using field campaign observations. The NCEP's Eta model will be used for this purpose because of its easy conversion from the eta to the sigma coordinate, so that the differences in the model results can be attributed solely to the use of different coordinate systems.
In addition to the use of terrain-following coordinates in models, insufficient horizontal and vertical resolution may be another reason for the large forecast errors in intermountain basins and valleys. Mesoscale models have typically used 1 to 10 km horizontal grid spacings and 20 to 50 m vertical grid spacings near the surface, which may not be enough to resolve sharply sloped sidewalls of basins, narrow valleys and canyons, or other subgrid-scale terrain irregularities. This raises two questions: 1) Are the effects due to subgrid-scale terrain variations large enough to significantly affect mesoscale model simulations? and 2) What is the best way to parameterize the these effects to improve simulations of flows associated with vertical transport and mixing? The first question will be addressed by comparing results from simulations with coarse resolution with those using very fine resolution that explicitly resolve most of the terrain variances and with observations. The second question will involve evaluations of existing approaches and development of new approaches to handle the effects of subgrid-scale terrain variations. High resolution simulations will be compared to simulations that incorporate a subgrid-scale terrain parameterization to assess the dynamic and thermodynamic effects of small-scale terrain variations on the development of slope winds and valley wind systems.
The second objective concerns the physical parameterizations that affect the prediction of mean and turbulent motions associated with vertical mixing processes. Several existing turbulence parameterizations will be evaluated by comparing their prediction to observations and to results from large-eddy simulations. The sensitivity of the simulated mean and turbulence fields to the choice of turbulence parameterization will be quantified to determine which parameterization schemes are most suitable for describing vertical exchange processes that occur during stable conditions in basins and valleys. Turbulence measurements and large-eddy simulations will be employed to develop new conceptual and numerical approaches in the treatment of turbulence in mesoscale models for use in basins and valleys. Some of the questions related to this objective are: How sensitive are model forecasts to the choice of turbulence parameterization schemes? To what degree are the mean quantities affected? What is the impact on turbulence quantities that affect vertical mixing of pollutants? Are there existing turbulence parameterizations that are more suitable for urban basins or valleys? If so, what processes do these schemes include that make them more suitable?
We would particularly like to invite other investigators to participate in our proposed research. We are especially interested in collaborations that would add instrumentation for turbulence measurements or provide turbulence quantities using technique of large-eddy simulations.
 |
Proposals Approved for Funding Home Page |