Convection During SEAC4RS: Comparing Aircraft Observations to WRF Large-Eddy Simulations

Abstract:

Deep convection remains a challenge to accurately parameterize in global and climate models. Increases in computer power recently have allowed large-eddy simulations (LES; grid spacing of O(100 m)) of deep convection, which are beginning to increase our understanding of this unresolved issue. Our research examined the Weather Research and Forecasting model in LES mode (WRF-LES) as a potential tool to further our understanding of deep convective cloud dynamics and microphysics. Idealized and nested WRF-LESs were made for 02 September 2013, a day on which 3 aircraft from the recent NASA SEAC4RS campaign extensively sampled deep convection during all phases of its lifecycle. When modeling deep convection at the LES scale, one of the greatest uncertainties is the choice of cloud microphysical parameterization. Thus, we tested the sensitivity of the WRF-LESs to several microphysical schemes. Simulated flight tracks were used to evaluate the WRF-LESs against the dynamical and microphysical data gathered during the SEAC4RS aircraft cloud penetrations. Results indicated the importance of cloud microphysical parameterizations when making deep convective LESs, especially if they are used to develop cumulus parameterizations. Results from the idealized WRF-LESs then were used to “tune” a real-data run in which the WRF-LES domain was nested within a mesoscale domain. This multi-scale nesting of an LES provides a framework for making detailed simulations of case studies when high-resolution observed data are available for evaluation. This nesting approach also might provide a new method, which uses more realistic atmospheric forcing for the LES, to develop cumulus parameterizations.