Sensitivity of Ensemble Forecasts of Supercells to Initial Condition Uncertainty

The sensitivity of full-physics ensemble forecasts of supercells to initial condition (IC) uncertainty is investigated. The motivation for the study largely stems from the NOAA Warn-on-Forecast program, where the primary objective is to develop a storm-scale ensemble prediction system that will generate probabilistic guidance for severe weather forecasts and warnings. The initial/boundary conditions for our simulations were generated from the real-time NSSL Experimental Warn-on-Forecast System for Ensembles (NEWS-e) during the 2016 NOAA HWT Spring Forecasting Experiment.  Our primary goal here is not to replicate observed supercell evolution, but rather to isolate the effect of IC uncertainty using a perfect- model assumption with horizontal grid spacing that can resolve the storm’s updraft and mesocyclone reasonably well (1 km).

After downscaling the 3-km NEWS-e analyses to the 1-km grid, the forecast sensitivity to IC uncertainty is assessed in two ways. First, we successively reduce the initial 1-km ensemble perturbations to 50%, 25%, and 10% of the original perturbations.  Second, ensembles are initialized at successively later times to examine the impact of IC uncertainty as a function of storm maturity.  This work will help to illuminate the impact of IC errors in contemporary storm-scale NWP.