Concurrent sensitivities of an isolated deep convective storm to parameterization of microphysics, horizontal resolution, and environmental sounding

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Thursday, 6 February 2014: 1:45 PM
Room C202 (The Georgia World Congress Center )
Annareli Morales, Colorado State University, Fort Collins, CO; and C. M. Villanueva-Birriel and H. Morrison

Bulk microphysical (MP) parameterizations describe the processes by which water and ice particles grow and precipitate within a cloud. Uncertainties in climate simulations and operational forecasts remain due to the choice of MP scheme. To analyze the effect different parameterizations have on cloud and precipitation processes, this study simulated a single, isolated deep convective storm occurring in a low wind shear environment using the Weather Research and Forecasting (WRF) model in idealized mode at “convection-permitting” scale (i.e., without a convection parameterization). Three MP parameterizations available in WRF (WSM6, Thompson, and Morrison) were compared at four different horizontal resolutions (horizontal grid spacing Δx = 2, 1, 0.25, and 0.125 km). Overall, there was considerable sensitivity of precipitation, storm structure, and dynamics to both microphysics parameterization and Δx. Differences between schemes were generally consistent at the various resolutions tested. Out of the three schemes, WSM6 produced the highest rain evaporation rate and hence lowest surface precipitation rate and efficiency. This occurred partly due to setting of the constant intercept parameter, N0, to a relatively large value in WSM6 (i.e. more numerous, smaller particles that evaporate faster). Mean surface precipitation rate and convective updraft mass flux both showed a non-monotonic sensitivity to Δx, with peak values of convective updraft mass flux tending to occur at intermediate Δx (250 m to 1km). This sensitivity to Δx was also found to be highly dependent on the storm environment (atmospheric sounding).