Toward Improvements to the COAMPS Urban Canopy Parameterization Scheme

Coastal urban areas represent a significant challenge to mesoscale numerical weather prediction (NWP) models due to the extremely heterogeneous surface properties which can result in localized areas of heating and cooling, wind circulations, and moisture variability. All of these effects are likely to be subgrid scale for most operational mesoscale NWP models. For this reason, urban effects must be parameterized. A number of urban canopy parameter-ization (UCP) schemes have been developed and used in case studies with a reasonable success. Nonetheless, when profiles of turbulence quantities (turbulence length scale, turbulent kinetic energy, eddy coefficients, etc.) are compared with large eddy simulation (LES) models in urban environments, substantial deficiencies are detected. A necessary preliminary step in seeking improvements in UCP schemes is careful documentation of these deficiencies.

In the present study, we use the nonhydrostatic mesoscale model COAMPS® to perform simulations with and without the multi-level UCP incorporated into the model. We utilize both three-dimensional, multiple nest and single column model implementations of COAMPS. We seek to understand the impact of urban structure parameters, including grid volume mean building height, building fraction, roof top fraction, anthropogenic heating, urban drag, and building morphology on the behavior of the urban boundary layer. We will perform detailed comparisons with LES models and identify deficiencies. We will also devise and test potential improvements to the UCP and boundary layer parameterization to address these deficiencies.