Further Experiments in Mesoscale Ensemble Forecasting in the Western Arctic

Ensemble prediction techniques continue to be utilized among operational weather prediction centers and research groups. While approaches based on perturbing the initial model conditions remain popular for operational forecast applications as well as for longer-range prediction research, approaches based on perturbing the model physics have been seen to have some merit for specific applications on the mesoscale.

Previously, (Tilley et al., AMS 13th NWP Conference) we investigated the utility of ensemble techniques with the PSU/NCAR MM5 mesoscale modeling system for high latitude mesoscale prediction, focusing on the transition period from the warm season to the cold season and on perturbations to the model physics available in MM5 versions 1 and 2.

In this presentation, we will extend this work by considering an ensemble simulation of an extended heavy rain event during mid-August 2000, a time prior to the start of the transition from the warm to the cold season in the Western Arctic. As before, the bulk of the ensemble members will be generated by perturbing the model physics, but here we utilize version 3 of MM5 and generate a larger ensemble which includes members with varying land surface models and newly incorporated radiation and PBL schemes. We evaluate the skill of the ensemble, and its members, through a statistical verification approach utilizing standard measures of skill and will present those results, as well as an overall assessment of the utility of ensembles in high latitudes based upon these results and the results of our earlier work.