Impacts of varying the integration depth on performance of updraft helicity as numerical guidance for severe thunderstorms forecasting

Convective-allowing numerical weather prediction (NWP) models provide opportunities for forecast guidance for severe thunderstorms unavailable with coarser-resolution models. In particular, the models begin to produce features on the scale of real thunderstorms and allow users to draw on the history of numerical thunderstorm simulations in order to estimate the relationship between model weather and actual weather. This new generation of forecast guidance is fundamental in bringing Warn-on-Forecast to fruition. However, how to construct reliable probabilistic information regarding severe convective phenomena when many of these phenomena will not be explicitly resolvable in larger domain model configurations for many years to come (e.g., explicit prediction of tornadoes will require grid spacing on the order of a few tens of meters) remains a major obstacle to fulfilling the goals of WoF.

Interpretation of numerical guidance derived from identification of model-based convective phenomena has been a subject of much study in recent years (e.g., Kain et al. 2008, Sobash et al. 2008, Schwartz et al. 2009, Kain et al. 2010). Kain et al. (2008) introduced the concept of updraft helicity (UH), calculated over the 2 to 5-km layer, as a descriptor of storm behavior. Conceptually, UH identifies rotating updrafts in model-simulated thunderstorms that can then be used to predict actual supercell thunderstorms. Sobash et al. (2008) showed that UH evaluated over this depth showed skill as a forecast aid. Here, we look at the sensitivity of the results to the depth over which the updraft helicity is calculated. We use output from the 4-km grid spacing WRF model run at NSSL in 2008 as forecasts for severe weather reports collected by the National Weather Service. In addition to the impact of integration depth, we will look at the impact of considering different neighborhoods in the evaluation of the forecasts. It is likely that considering a particular grid point to be a forecast for a neighboring area is a more valuable way to interpret the forecasts than just looking at a single location, as shown by Clark et al. (2010) for precipitation. We will estimate the impacts of neighborhood verification on severe thunderstorm forecasts.

References:

Clark, A. J., W. A. Gallus, Jr., and M. S. Weisman, 2010: Neighborhood-based verification of precipitation forecasts from convection-allowing NCAR WRF model simulations and the operational NAM. Wea. Forecasting, 25, in press.

Kain, J. S., and Coauthors, 2008: Some practical considerations regarding horizontal resolution in the first generation of operational convection-allowing NWP. Wea. Forecasting, 23, 931–952.

Kain, J. S., S. R. Dembek, S. J. Weiss, J. L. Case, J. J. Levit, and R. A. Sobash, 2010: Extracting unique information from high resolution forecast models: Monitoring selected fields and phenomena every time step. Wea. Forecasting, in press.

Schwartz, C. S., and Coauthors, 2009: Next-day convection-allowing WRF model guidance: A second look at 2-km versus 4-km grid spacing. Wea. Forecasting, 24, 3351-3372.

Sobash, R., D. R. Bright, A. R. Dean, J. S. Kain, M. Coniglio, S. J. Weiss, and J. J. Levit, 2008: Severe storm forecast guidance based on explicit identification of convective phenomena in WRF-model forecasts. Preprints, 24th Conf. on Severe Local Storms, Savannah, GA, Amer. Meteor. Soc., 11.3. [Available online at http://ams.confex.com/ams/pdfpapers/142187.pdf.]