Assessing the predictability of band formation and evolution during three recent northeast U.S. snowstorms

Mesoscale precipitation bands are frequently observed in the comma-head portion of extratropical cyclones in the northeast United States. High-resolution models such as the MM5 and WRF models have shown the capability to predict mesoscale precipitation bands, suggesting an ensemble of high-resolution model forecasts may provide useful information concerning the predictability of band formation and evolution. This study utilizes a high-resolution multi-model, -initial condition, and -physics ensemble to assess the predictability of observed band formation and evolution during the 25 December 2002, 12 February 2006, and 14 February 2007 snowstorms.

A 12-km ensemble using the MM5 and WRF models and initialized using the operational National Centers for Environmental Prediction North American Mesoscale model, Global Forecast Systems model, and Short-Range Ensemble Forecast system was utilized. Microphysics, boundary layer, and convective parameterizations were varied amongst the ensemble members to further diversify the model solutions. The ensemble was initialized 12–24 h prior to band formation in each snowstorm.

Results from the ensemble forecasts show differences in the predictability of band formation between snowstorms, with nearly all members forecasting band formation in the 25 December 2002 and 14 February 2007 snowstorms, while just a few members forecasted band formation during the 12 February 2006 snowstorm. Although the 25 December 2002 and 14 February 2002 snowstorms exhibited high predictability in the occurrence of a band, ensemble member predictions of the timing and location of band formation and its subsequent evolution varied considerably. Possible reasons for these differences in the predictability of band occurrence between each snowstorm and band location and evolution amongst individual ensemble members will be presented at the conference. Discussion of the possible utility of high-resolution ensemble forecasts in real-time forecast operations will also be presented.