850 Exploration of Global Model Predictions of a High Impact Winter Weather Event Using the THORPEX Interactive Grand Global Ensemble (TIGGE)

Wednesday, 9 January 2013
Exhibit Hall 3 (Austin Convention Center)
Daniel S. Russell, Current Affiliation, UCLA/ Department of Atmospheric and Oceanic Sciences, Norman, OK; and S. M. Hitchcock, D. Parsons, and N. New

Handout (1.8 MB)

On January 28-29, 2010, a winter storm event blanketed parts of the southern Great Plains and Tennessee Valley regions with severe ice and snow that caused schools and businesses to be shut down for several days. In Oklahoma alone, this event caused over 200,000 homes and businesses to lose power and cost electric companies in excess of 43 million U.S. dollars. Early warning of such events could improve society's response and thus mitigate some of the detrimental impacts of the storm. Investigations by Legg and Mylne (2004, 2007) suggest that ensemble systems can provide early warnings of severe events, including winter storms, up to four days in advance, while skill diminishes slowly between days five and six. A subset of the THORPEX Grand Global Interactive Ensemble (TIGGE) archive, including the European Center for Medium Range Forecasts (ECMWF), the National Center for Environmental Prediction (NCEP), and the United Kingdom Meteorological Office (UKMO), is used in this study to provide an analysis of the model predictions of this January 28-30, 2010 event. The ensemble perturbation members from each of the three forecast centers are used here to calculate a new ‘combination forecast'. This ‘combination forecast' is defined as the percentages of perturbation members from all three forecast centers that meet temperature and precipitation thresholds. It is shown here that the ‘combination forecast' of precipitation, surface temperature, and 850 millibar temperature provide predictive capabilities up to four days before the event, particularly for forecasts of the locations of freezing rain. These results, when combined with the Legg and Mylne study, suggest that early warning of major freezing rain events may indeed be possible. The ‘combination forecasts' allowed for accurate forecasts despite the relatively poor quality of one of the ensemble members. The result has significant implications as freezing rain is particularly challenging to forecast, and the use of multiple ensembles is often difficult. The combination of ensembles from different forecast centers had been impractical due to inconsistencies in grid spacing between different forecast centers. The TIGGE allows users to seamlessly interpolate model forecasts from different forecast centers to a user-defined grid. TIGGE is a research archive that is not available in real-time. The availability of a subset of the ensemble data in real-time could, with appropriate tools, improve the prediction of high impact weather.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner