CAPS Storm-Scale Ensemble Forecasting System: Impact of IC and LBC perturbations

Since 2007, the Center for Analysis and Prediction of Storms at the University of Oklahoma produces realtime storm-scale ensemble forecast (SSEF) each year to support the NOAA Hazardous Weather Testbed (HWT) Spring Experiment. The 2013 CAPS SSEF ran from April 22 to June 7, 2013, consisting of 30 multi-model multi-physics ensemble members using three NWP model systems (WRF-ARW, COAMPS, and ARPS), with a domain covering the full continental United States with convection-allowing resolution at 4-km horizontal grid spacing. CAPS SSEF members were configured with a hybrid of initial/lateral boundary condition (IC/LBC) perturbations extracted from the operational Short-Range Ensemble Forecast (SREF) ensemble members (at 16 km grid spacing) and various combinations of physics options in microphysics, PBL and land-surface model, and radiation. Up to 140 Doppler weather radar data, with both radial wind and reflectivity, and other observation data were analyzed into the SSEF members in realtime using the ARPS 3DVAR and Complex Cloud Analysis system. This paper examines the impact of IC and LBC perturbations on QPF forecast ETS skill scores and ensemble spread through a 4-day small ensemble experiment that had been performed using different IC perturbation magnitudes and LBC perturbation strategies. While ensembles with both IC and LBC perturbations do confirm to have much larger spread than physics-only ensemble, LBC perturbations play key role in the spread increase. Smaller IC perturbation magnitude can result in higher QPF skill. Higher IC perturbation magnitude can have negative impact to QPF skill, even with larger ensemble spread. Verification of QPF for the entire 2013 Spring Experiment season will also be presented.