Examination of An Intense Wake Low Event as a Severe Local Storm

On 13 April 2009, an intense wake low event occurred over northeastern Mississippi, northern Alabama, and northern and central Georgia, near the back edge of a mesoscale convective system (MCS). This was a significant weather event, as thousands of trees were toppled over this large region, some onto homes and vehicles. At least 150,000 Alabama Power customers lost power, with additional power outages in Georgia. Structural damage was widespread, occurring in over 50 counties in Alabama and Georgia. Property damage was estimated near $4 million. Injuries and one fatality were also reported. High wind warnings and/or wind advisories were issued by National Weather Service offices in Huntsville, Birmingham, and Atlanta. A wind gust of 26 m s-1 was recorded in downtown Birmingham, AL, and a wind gust to 23 m s-1 occurred at Atlanta airport. Gusts > 20 m s-1 were common over the entire region, and winds in excess of 15 m s-1 persisted at many locations for one hour or more. Pressure falls of 8-10 hPa in 1 to 2 hours were common.

The wake low began to develop over northeast Mississippi, in the northern, more benign part of the MCS (in terms of severe convective storms). However, wind profiler data from the NOAA 404 MHz site at Okolona, MS, and the UAH MIPS 915 MHz profiler, as well as NEXRAD radar data, indicate that the back edge of the MCS was associated with a strong, descending rear inflow jet (RIJ) with winds of 20 to 30 m s-1, and downward motion of 2 to 3 m s-1. Radar data also show a horizontally extensive area of virga falling from the back-sheared anvil, with returns of 20 dBZ or more extending 125 km behind the surface rainfall area. Synoptic data indicate that a significant dry slot wrapped northward around the parent upper low centered in eastern Kansas, with a large area of RH < 20% at 500 hPa over much of Arkansas and Louisiana as the wake low developed, and a large horizontal gradient in 500 hPa RH extending from Mississippi into Alabama. A NAM model sounding at 06 UTC at Birmingham shows a dry slot 5 and 6 km MSL, around the time the wake low was approaching BHM.

In this paper, trajectory analysis will be performed using wind profiler and Doppler radar radial velocity data, assuming the feature was two-dimensional and performing a time-to-space conversion (TSC). It appears that the extremely dry air at mid-levels may have been advected into the MCS by the RIJ, causing extensive evaporation of precipitation and associated cooling. This likely produced strong downward motion and warming in the stratiform region of the MCS. Due to the stable air below 700 hPa, the pressure signal at the surface due to the adiabatic warming quickly canceled, and then became much larger than, the signal due to evaporative cooling, producing the large surface pressure falls. The surface pressure signal is affected most by vertical displacements at lower levels, and by those where stability is greatest.

The magnitude of observed surface winds associated with any pressure disturbance are proportional to the amplitude of the disturbance and the background flow in the direction of the perturbation winds, and inversely proportional to the intrinsic speed of propagation of the disturbance. In this case, the extremely large amplitude (5 to 10 hPa) and small contribution from the component of background wind in the direction of the perturbation winds produced very high winds despite the relatively fast intrinsic speed of 20 m s-1. The large wind ambient shear and downward motion may have also produced downward transport of horizontal momentum, further enhancing the surface winds. Finally, radar and profiler data show that the largest magnitude southeasterly winds associated with the wake low were very near the surface, where the integrated vertical displacement above a point was the greatest. This, combined with the strong northwesterly RIJ, produced very large bulk wind shear of about 50 m s-1 from 0-5 km MSL. The vertical shear in some parts of the low-level flow were on the order of 10-1 s-1.

Other wake lows have caused widespread wind damage. Several have already been investigated preliminarily by the authors, including ones in Alabama, Mississippi, and Georgia on 20 December 2007, 12 April 2005, and 22 February 1998, and near St. Louis, Missouri on 28 April 1996. These wake lows often cause damage on the aerial scale and intensity of a bow echo or even a derecho event, and are in need of much more research on their formation, detection, and forecasting.