Relationships of Several Stability Indices to Convective Weather Events in Northeast Colorado

Paul Schultz NOAA, Environmental Research Laboratories, Forecast Systems Laboratory, Boulder, Colorado

Search for other papers by Paul Schultz in
Current site
Google Scholar
PubMed
Close
Restricted access

Abstract

Seven familiar stability indices were computed from sounding data for each of 83 days of a convection forecasting experiment conducted during the summer of 1985 in northeast Colorado. Observations of convectively driven weather events were collected; the values of the indices were compared against this dataset to examine their performance as predictors of severe weather (large hail, tornadoes, high wind) and significant weather (nonsevere but important from an economic or public safety standpoint). The results of the analysis are

  1. Benchmark values of the indices that give their typical magnitudes on active days versus quiescent days. These values, compared with those computed in other regions, illustrate the potential fallacy of interpreting the indices in the absence of analogous region-specific reference statistics.

  2. Rankings that determine which indices worked best in this experiment. The highest ranked indices were the SWEAT index for severe weather and buoyancy for significant weather. Interestingly, SWEAT was the worst of those tested for significant weather.

  3. Quantitative convection forecasting guidance. The observed relative frequencies of severe and significant convection as functions of the seven indices are presented in graphical form. When used in a forecasting context, these observed relative frequencies can be interpreted as probabilities of severe and/or significant weather. Some of the graphs are clearly bimodal; no explanation for this behavior is offered.

Some of the benefits that would be realized by collecting more data, in this and other regions, are suggested. For example, there is a good possibility that some indices show particular skill for certain types of events (e.g., hail vs high wind, etc.), but the present dataset is too small to clearly establish any such connections.

Abstract

Seven familiar stability indices were computed from sounding data for each of 83 days of a convection forecasting experiment conducted during the summer of 1985 in northeast Colorado. Observations of convectively driven weather events were collected; the values of the indices were compared against this dataset to examine their performance as predictors of severe weather (large hail, tornadoes, high wind) and significant weather (nonsevere but important from an economic or public safety standpoint). The results of the analysis are

  1. Benchmark values of the indices that give their typical magnitudes on active days versus quiescent days. These values, compared with those computed in other regions, illustrate the potential fallacy of interpreting the indices in the absence of analogous region-specific reference statistics.

  2. Rankings that determine which indices worked best in this experiment. The highest ranked indices were the SWEAT index for severe weather and buoyancy for significant weather. Interestingly, SWEAT was the worst of those tested for significant weather.

  3. Quantitative convection forecasting guidance. The observed relative frequencies of severe and significant convection as functions of the seven indices are presented in graphical form. When used in a forecasting context, these observed relative frequencies can be interpreted as probabilities of severe and/or significant weather. Some of the graphs are clearly bimodal; no explanation for this behavior is offered.

Some of the benefits that would be realized by collecting more data, in this and other regions, are suggested. For example, there is a good possibility that some indices show particular skill for certain types of events (e.g., hail vs high wind, etc.), but the present dataset is too small to clearly establish any such connections.

Save