Many users of ensemble weather forecasts benefit from objective probabilistic guidance, which requires the ensemble to be formally calibrated. Since 2007 the Dugway Proving Ground (DPG) in Utah has been using an operational mesoscale ensemble NWP system for mission-critical weather forecasting. Xcel Energy has been utilizing a similar system for forecasting wind speed and, critically, ramp events, of high interest for wind energy applications. In both cases, the ensemble comprises 30 members, each a different configuration of the Four-Dimensional Weather system jointly developed by the National Center for Atmospheric Research and the U. S. Army Test and Evaluation Command.

To calibrate the respective ensemble weather forecasting systems, we have tested a variety of post-processing techniques: quantile regression, a Kalman filter bias correction algorithm that is run in analog space rather than in time, another method based purely on the analog concept, logistic regression, and finally, linear variance calibration. In this work, we compare and contrast the performance of these five post-processing techniques for calibrating both temperature and wind speed forecasts at different lead-times and for different amounts of hindcast (training) data, both individually, and as a blended product. Generally we see large improvements in skill through post-processing over the raw uncalibrated ensemble, with additional further improvements by using a blending of approaches.

The presenter will describe the mesoscale ensemble, review the steps used to calibrate the ensemble forecast, and present verification statistics and figures from operational ensemble forecasts comparing and contrasting the post-processing approaches' improvements relative to the raw uncalibrated ensemble.