Three representatives from SPoRT participated in the 2010 Spring Experiment in Norman, OK. The annual Spring Experiment is hosted by the Hazardous Weather Testbed and contains two separate components: the Experimental Forecast Program (EFP) and Experimental Warning Program (EWP). Jonathan Case and Bill McCaul attended the EFP during the weeks of 17 May and 1 June, respectively, while Geoffrey Stano participated the EWP during the week of 7 June. This blog post highlights the EFP and the suite of high-resolution models and products used.
This year’s EFP had three focus areas:
- Severe weather experimental outlooks with probabilities of severe hail, wind, and/or tornadoes for two time windows at 2000-0000 UTC and 0000-0400 UTC,
- Quantitative Precipitation Forecast (QPF) experimental outlooks focusing on probabilities of exceeding 0.5” and 1.0” in six-hour windows ranging from 1800-0000 UTC and 0000-0600 UTC, and
- Aviation experimental convective outlooks forecasting coverage of 25,000 ft echo tops (18+ dBZ) and reflectivity exceeding 40 dBZ at three separate times of 2100, 2300, and 0100 UTC. They also issued day-2 outlooks in the afternoon update.
The EFP activities had a heavy reliance on high-resolution numerical weather prediction guidance from both ensemble and deterministic models run by the Centers for Analysis and Prediction of Storms (CAPS), NSSL, and NCAR. Each day, participants were divided into one of the three forecast groups to take on the day’s challenge centered on a region of meteorological interest. The Aviation group typically had a larger area of responsibility due to planning associated with flight tracks and traffic in a real-world scenario.
CAPS provided a set of 26 high-resolution modeling members that used a variety of initial conditions from the NCEP Short-Range Ensemble Forecasts, 3DVAR data assimilation using radar data, and a wide range of different physics schemes within the models. Most model runs were done at 4-km grid mesh using the Weather Research and Forecasting (WRF) model and the Advanced Research WRF dynamical core. The NSSL WRF model also provided unique deterministic output fields to aid in severe weather forecasting, including a WRF forecast total lightning threat and hourly maximum quantities of reflectivity, updraft helicity, 10-m wind speeds, etc. The hourly maximum quantities help to depict the evolution of convective features in between the regular hourly model output by keeping track of the maximum values at every grid point during the model integration. These unique products in the NSSL WRF represent SPoRT contributions for the EFP.
SPoRT participants noted that due to the quantity of data to examine combined with the time constraints of delivering the experimental products, most high-resolution guidance used during the Experiment consisted of the CAPS ensemble output fields. There was not much opportunity to examine the experimental WRF lightning threat forecasts because the fields were only available in a single deterministic model (i.e. NSSL WRF).
Figures 1 and 2 show an example set of products created by the severe and QPF teams, respectively, for 19 May on which an outbreak of severe thunderstorms and tornadoes occurred across much of Oklahoma. Figure 3 shows an experimental aviation team forecast later that week from 21 May.
Finally, the Meteorological Evaluation Tools model verification package was employed to verify some of the high-resolution model output. The verification page can be seen at: http://verif.rap.ucar.edu/eval/hwt/2010/.