WINTER FORECAST PERFORMANCE OF AN OPERATIONAL MESCOSCALE MODELLING SYSTEM IN THE NORTHEAST U.S. – WINTER 2002-2003

Anthony P. Praino IBM Thomas J Watson Research Center, Yorktown Heights, NY

Lloyd A. Treinish IBM Thomas J Watson Research Center, Yorktown Heights, NY

We examine the forecast performance of an operational mesoscale modeling system dubbed ‘Deep Thunder’ for the 2002 – 2003 winter season over the northeast United States. Current operations utilize nested domains at 16, 4 and 1 km resolution with explicit cloud microphysics centered on the New York City metropolitan area, but also includes surrounding coastal Atlantic waters. This provides the opportunity to evaluate detailed snow accumulation predictions for a number of interesting events that affected the greater New York City area. Model skill is compared with significant snow events during the 2002-2003 season as well as considering the operational availability of such results. Performance is examined in forecasting total snowfall amounts using two snow algorithms as well as spatial and temporal distributions of snowfall. Results indicate model skill in predicting synoptic scale winter storm events with a focus on mesoscale impacts.