Application of Climate Statistics and Ensemble Forecasts in the Prediction of Severe Weather Episodes

Ensemble forecasts are widely known for their ability to reduce statistical errors through consensus or mean forecasts, as well as gauge forecast uncertainty by virtue of spread and probabilistic guidance. Ensemble forecasts can also be viewed relative to the long-term climate to assess potential departures from normal and the probability of exceeding climatic anomalies. The comparison of ensemble forecast spread relative to climatic variance provides additional information on predictability.

This paper examines the predictability of both weakly and strongly forced severe weather episodes by examining ensemble predictions of basic severe weather parameters. Parameters such as CAPE, deep-layer shear, and helicity relative to their climatic normals and variances are investigated. Joint parameter distributions are also analyzed. Results from the NCEP Short-Range Ensemble Forecast (SREF) system will show that climate information used in conjunction with basic and derived SREF severe weather parameters can help clarify the magnitude of the threat while quantifying appropriate levels of confidence.