NWC REU 2013
May 22 - July 30



Photo of author

Evaluation of the National Severe Storms Laboratory Mesoscale Ensemble

Brianna Lund, Dustan Wheatley, and Kent Knopfmeier


What is already known:

  • Ensemble-based weather forecasts are increasingly used as guidance in the prediction of severe storms.
  • During the 2013 Spring Forecast Experiment, the NSSL Mesoscale Ensemble (NME) was run daily in a simulated forecasting environment.

What this study adds:

  • In regards to reproducing realistic mesoscale environments in which storms were observed to form, the NME performed comparably to a commonly used operational mesoscale model, the Rapid Refresh (RAP) model.
  • Both modeling systems were characterized by relatively small errors in their placement of the dryline and the positioning and strength of storm-induced cold pools, although the NME was run at much lower computational expense.


Accurate short-term forecasts are critical for forecasters when anticipating severe weather events and improving such forecasts has long been a focus for meteorologists. The recent emergence of ensemble based data-assimilation systems has proven to be a promising step toward the improvement of these vital forecasts. The National Severe Storms Laboratory Mesoscale Ensemble (NME) is a 36-member ensemble that provides hourly forecasts and analyses of a variety of products used for severe weather forecasting, such as soundings and 2-m temperature fields. This project seeks to quantitatively evaluate said products through comparison to observations from a number of sources (surface stations, rawinsondes, etc.), including the Oklahoma and Texas mesonets.

Full Paper [PDF]