Evaluating Stochastic Perturbations of Microphysical Parameters in Convection-Permitting Ensemble Forecasts of an Orographic Precipitation Event

Accurate precipitation forecasts in mountainous areas, such as the Olympic Mountains in Washington state, USA, are difficult due to many factors influencing precipitation amounts including the role of complex terrain and numerical weather prediction biases. In particular, uncertainties in the parameterization of microphysical processes can lead to large forecast errors. One way to account for microphysics scheme (MP) uncertainty is to design convection-permitting ensemble forecasts, using methods such as stochastic parameter perturbation (SPP) to vary MP parameters. The goal of this research is to evaluate and improve the utility of SPP for representing MP uncertainty in ensemble forecasts of cool-season orographic precipitation. We do so first, by identifying uncertain MP parameters that affect the distribution of orographic precipitation, and second, by running an ensemble using SPP to vary these uncertain parameters.

We focus on a case from the Olympic Mountains Experiment (OLYMPEX) that occurred during 12–13 November 2015, with observed precipitation amounts in excess of 370 mm on the western slopes of the Olympics. This case was associated with an atmospheric river-type storm and had several distinct precipitation periods (pre-frontal, warm sector, and post-frontal), useful for studying MP deficiencies across a range of thermodynamic environments. We run nested simulations of the event at convection-permitting horizontal grid spacing using the Weather Research and Forecasting (WRF) model configured similarly to the operational High-Resolution Rapid Refresh model. All simulations use Thompson-Eidhammer aerosol-aware MP. Sensitivity experiments are conducted by varying fixed parameters within MP that affect processes both above and below the melting level, such as snow fallspeed equation coefficients, snow capacitance, cloud water and snow particle size distribution shape parameters, and collection efficiencies of cloud water by snow and rain.

Fixed-value experiments are performed to determine appropriate perturbation magnitudes for individual parameters, then SPP ensemble simulations are run varying a set of these parameters. Results from the WRF simulations are compared to precipitation observations from OLYMPEX, including rain gauges and disdrometers. WRF forecasts are also evaluated against in situ and remote sensing OLYMPEX observations, such as the University of North Dakota Citation aircraft cloud probes and an array of ground-based scanning and profiling radars. In this manner, the amplitude of SPP performed on various MP parameters can be physically constrained by observations.