Extended Abstract (100K)13A.2
Utilizing Short Range Ensemble Point Forecast Soundings for Severe Storms Forecasting
Jason J. Levit, NOAA/NWS/NCEP/SPC, Norman, OK ; and J. Hart, R. S. Schneider, D. R. Bright, and R. L. Thompson
Visualizing Short Range Ensemble Forecast (SREF) data has traditionally been accomplished through generating 2D horizontal plots of both standard and specialized data fields, such as the ensemble mean and median, spaghetti plots of a scalar value, or exceedance probability contours of single or combined parameters. Attempting to view and intelligently use point forecast soundings from SREF data is more problematic. For example, creating a single spaghetti-type SKEW-T diagram that contains sounding data from many ensemble members often is confusing, as multiple soundings on a single diagram presents challenges in the overall interpretation. Also, creating a vertical profile of various post-processed central tendency statistics from multiple soundings (such as the ensemble mean) will generally create a plot where useful and potentially significant features such as temperature inversions and PBL depth are smoothed out or are difficult to discern. Questions remain, therefore, on how to properly view SREF soundings to obtain the most useful information for weather forecasting, and in our case, the severe storm environment. As a first test, we create “postage stamp” style SKEW-T plots from a few test cases from the Spring 2008 severe weather season in an attempt to visually examine the range of sounding forecasts from the SREF. Using these plots as a starting point, we discuss their potential uses, and examine emerging methods of visualizing and extracting information from SREF soundings, both in a visual and statistical sense. Finally, the multi-model, multi-physics composition of the SREF, an overall strength of the system, is known to produce pre- and post-convective environmental sounding characteristics that are strongly influenced by specific model and physics configurations (e.g., choice of convective parameterization). This suggests that cluster analysis techniques to identify and classify sounding solutions by model may need to be employed in any statistical analysis of SREF soundings
Session 13A, Ensemble Forecasts of Severe Weather
Wednesday, 29 October 2008, 1:30 PM-3:00 PM, North & Center Ballroom
Previous paper Next paper
Recorded presentation