Impacts of Stochastic Ice Microphysical Parameters on Mesoscale Convective System Ensemble Simulations

 
Poster PDF

Authors

McKenna Stanford — Columbia University
Adam Varble — Pacific Northwest National Laboratory
Hugh Clifton Morrison — University Corporation for Atmospheric Research
Wojciech Grabowski — National Center for Atmospheric Research (NCAR)
Greg McFarquhar — University of Oklahoma
Wei Wu — NOAA National Ocean Service

Category

Microphysics (cloud, aerosol and/or precipitation)

Description

The parameterization of ice microphysical properties can have significant impacts on important simulated mesoscale convective system (MCS) characteristics, including cloud and precipitation structure, cloud radiative properties, and system evolution. Most microphysics schemes in cloud-resolving models employ power law relationships with constant coefficients to represent particle mass, area, and terminal fallspeed as a function of maximum dimension. However, observations suggest that these parameters vary as a function of time, space, and environmental conditions such as temperature. To address this natural variability, a stochastic ice microphysics framework was implemented into the Predicted Particle Properties (P3) microphysics scheme in the Weather Research and Forecasting (WRF) model. The framework was first applied to the unrimed mass-size relationship (m=aDb) with covarying “a” and “b” values constrained by distributions retrieved from aircraft observations and tested over a range of spatiotemporal autocorrelation scales. Next, the stochastic framework was applied to the riming collection efficiency of cloud water onto ice (Eci) for which observational constraints are limited. 3D ensemble simulations are performed for a well-studied squall line event that occurred on 20 May 2011 during the Midlatitude Continental Convective Clouds Experiment (MC3E). The ensemble spread for domain-wide median outgoing longwave radiation brightness temperature, ice water path, and optical depth ranges from 1-8 °C, 10-80 g m-2, and 0.25-1.5, respectively, depending on the time analyzed. The impacts of the stochastic framework are further assessed using large ensembles of 2D idealized convective system simulations and smaller ensembles of 3D “real” simulations with the “real” simulations evaluated using surface observations, observed radar reflectivity and radial velocity at multiple wavelengths, and satellite-measured radiative fluxes. The 3D “real” simulations are also analyzed for the 23-24 May 2011 MC3E supercell event and compared to results from the 20 May squall line simulations. Lastly, stochastic ensemble spread is shown to be comparable to initial condition ensemble spread for some forecast times and variables, while multi-physics ensemble spread is greater.