To date, a lack of climatological studies establishing spatiotemporal domains, frequency of occurrence and intensity characteristics has limited broader understanding and public recognition of these windstorm types, with corresponding underrepresentation in operational warning frameworks. Here we present results of a comprehensive climatological study covering 23 years over the upper Midwest USA that offers the basis for establishing such methodologies. We applied detection algorithms to high-temporal resolution (1-min) conventional weather station observations, sourced from the Iowa State University online archive, covering a 540 x 540 km domain centered on Wisconsin, where both wake lows (WL) and mesoscale gravity waves (MGW) have been documented with some regularity. Case files were generated for all events meeting combined pressure and windspeed change threshold values, including plotted time series, tables of meteorological variables and WSR-88D radar reflectivity loops. The aims of this initial climatological analysis are to identify storm typologies associated with storm-strength mesoscale pressure fall phenomena and their frequency and intensity characteristics.
Results demonstrate the ubiquity of both storm types and are rather startling. An average of 11 events each year (6.8 WL, 4.1 MGW) generate winds exceeding the 17 m s-1 gale force threshold, while 0.57 events per year (0.33 WL, 0.24 MGW) have winds that reach 25 m s-1 storm force; for some sense of comparison, perhaps 25 MGW events in total have been documented throughout the United States in scientific literature to date. There are strong seasonality preferences too, with MGWs peaking in April and WLs in June. Multiple events produced wind damage and prompted the issuance of high wind warnings while in progress (though almost invariably without forewarning). In addition to summary findings from the climatological study, examples of different sub-classes of windstorms will be demonstrated in the presentation, and include some previously undescribed typologies.
We will furthermore demonstrate the results of experimental predictions of mesoscale windstorm occurrence following the model of NOAA Storm Prediction Center categorical outlooks. These forecasts apply subjective pattern recognition to operational numerical model outputs, which frequently demonstrate success in characterizing antecedent conditions, and in some cases, explicit forecasts of propagating linked wind-pressure change anomalies that subsequently verify in both space and time.

