1. Introduction
A wake low may be described simply as a mesoscale area of low pressure at the rear of a mesoscale convective system (MCS). Wake lows were initially examined by Fujita (1955) and have since been studied by many authors (e.g., Handel and Santos 2005; Gaffin 1999; Johnson 2001). Yet, a comprehensive understanding of their thermodynamics remains “elusive” (Johnson 2001). Loehrer and Johnson (1995) noted their strong winds. The consensus is that wake lows are associated with subsidence at the rear of an MCS; cooling due to sublimation, melting, and evaporation in the descending air is more than offset by adiabatic warming, producing an “overshooting” bottom associated with strong positive buoyancy (e.g., Gallus and Johnson 1995; Smull and Jorgensen 1990; Smull et al. 1991; Johnson and Hamilton 1988). The warm perturbation hydrostatically produces negative pressure perturbations at the surface. This overshooting is thermodynamically similar to a heat burst (e.g., Bernstein and Johnson 1994). Stumpf et al. (1991) found that significant pressure falls in a wake low could be directly attributed to the downward depression of a surface-based cold pool (Johnson 2001; see Fig. 1), and suggested that such a process may explain the very large pressure falls in wake lows where the parent MCS moves over a surface inversion or stable layer (e.g., Bosart and Seimon 1988).
Schematic cross section of an MCS with a mesohigh at the front and a wake low at the back edge of the stratiform precipitation area (adapted from Johnson 2001). The gray circled area illustrates the area where the surface-based cool air (cold pool or stable layer) is depressed downward, causing a fall in surface pressure. Vectors are not necessarily to scale; the strongest winds in the wake low are located at the point of lowest pressure.
Citation: Weather and Forecasting 26, 5; 10.1175/WAF-D-11-00021.1
Johnson and Hamilton (1988) proposed that the subsidence is associated with the rear-inflow jet (RIJ) in the MCS. Many studies have shown that wake lows are most intense when the RIJ is “blocked”; that is, it does not flow through the entire stratiform precipitation area (e.g., Stumpf et al. 1991; Johnson and Bartels 1992). This blocking is associated with large convergence aloft and a rapid descent of the RIJ. Microphysical processes help produce the downward vertical motion. Schmidt and Cotton (1990) suggest that gravity waves play a role in MCS-related surface pressure perturbations.
The intent of this paper is to examine three significant wake lows over Alabama and Georgia. We will examine their environments, typically including a low-level stable layer and midlevel dry air behind the MCS. Radar and surface observations will be examined. The magnitudes of the surface winds in the wake lows, and the damage they produced, will be observed, including factors contributing to the presence of these winds. It should be noted that the strong winds associated with most severe squall lines are from a westerly direction, while those in wake lows are from an easterly direction. This may allow for more efficient downing of trees in wake lows, given that a different subset of trees is exposed/susceptible to the highest winds in wake lows than those exposed/susceptible to the more frequent strong westerly winds accompanying squall lines.
2. A review of the events
a. 13 April 2009
An intense wake low propagated across much of northern Alabama and northern Georgia during the morning hours of 13 April 2009. This wake low was a widespread, significant weather event, downing thousands of trees (many onto homes and automobiles), causing an estimated $4 million in property damage, several injuries, and even one fatality in the Atlanta metropolitan area (due to a tree falling on the person’s vehicle) (NWS 2009; NCDC 2011). At least 150 000 Alabama Power Company customers lost power, and damage was reported in more than 50 counties in Alabama and Georgia (NWS 2009; NCDC 2011). Pressure falls of 8–10 hPa in 2 h were common, as were wind gusts greater than 20 m s−1, with the highest recorded wind gust (26 m s−1) occurring at an elevated site in downtown Birmingham, just south of the most intense damage swath (for surface observations in all three wake lows, see Fig. 2). The wake low was propagating toward the southeast (from 310°) at 16 m s−1. Since the perturbation winds in a wake low are in the opposite direction of the wake low propagation, ambient easterly low-level winds (also opposite of the propagation direction) enhanced the net winds in the wake low (e.g., Coleman and Knupp 2009).
Observed surface pressure (hPa, solid dark line) and wind gusts (m s−1, gray dots) from the three wake lows examined herein: (a) Atlanta, GA (ATL), 13 Apr 2009; (b) BHM, 20 Dec 2007; and (c) BHM, 22 Feb 1998. (d) The analog wind trace from BHM, showing a peak wind gust of 49 kt (25 m s−1) on 22 Feb 1998 [(c) and (d) are adapted from Bradshaw et al. (1999)].
Citation: Weather and Forecasting 26, 5; 10.1175/WAF-D-11-00021.1
Vertical profiles of the Brunt–Väisälä frequency (N, s−1), representing static stability. Data are derived from soundings at (a) 1200 UTC 13 Apr 2009, Peachtree City, GA (FFC); (b) 0000 UTC 21 Dec 2007, Calera, AL (EET); and (c) 1200 UTC 22 Feb 1998 at EET.
Citation: Weather and Forecasting 26, 5; 10.1175/WAF-D-11-00021.1
Radar observations from 13 April 2009 (Figs. 4a and 5a) show the classical features of a wake low. The lowest-elevation plan-position indicator (PPI) reflectivity scan indicates an asymmetric MCS (e.g., Skamarock et al. 1994; Loehrer and Johnson 1995, especially their Fig. 22). There was a long, fairly sharp gradient in radar reflectivity at the back edge of the MCS (Fig. 4a) that extended almost 3 km vertically (Fig. 5a), underneath anvil precipitation. These sharp reflectivity gradients are a common feature in wake lows (e.g., Haertel and Johnson 2000). The horizontal gradient in reflectivity at 2 km MSL was 4 dB km−1, indicating significant subsidence and evaporation of precipitation. This was likely associated with the deep, descending RIJ (Fig. 5b) that did not penetrate the entire stratiform precipitation area, meaning it was “blocked,” as is often the case in intense wake low events (see section 1). The evaporation of precipitation was likely aided by dry air in the RIJ. Figures 6a and 6b shows an 8-km-deep and 400-km-wide region of dry air (relative humidity < 30%) to the west of the MCS, near the main source region of RIJ air. The 0.5° PPI velocity image (Fig. 4b) shows the strong low-level winds at the back edge of the stratiform rain area; the zoomed-in area indicates winds near 30 m s−1.
The 0.5°-elevation PPI radar scans (100-km range rings). (a) Reflectivity (dBZ; see legend at bottom) from the BMX Weather Surveillance Radar-1988 Doppler (WSR-88D) at 0637 UTC 13 Apr 2009. (b) As in (a), but for base radial velocity (m s−1; see legend at bottom). (c) The reflectivity from BMX at 2227 UTC 20 Dec 2007. (d) As in (c) but for velocity. (e) Reflectivity from the FFC WSR-88D at 1808 UTC 22 Feb 1998. (f) As in (e), but for velocity. The two white squares in (b) and (d) indicate the area of the zoomed-in base velocities in the insets (m s−1, legend at right in each panel). In these insets, deep purple indicates radial velocities > 25 m s−1, and peach colors indicate velocities > 30 m s−1.
Citation: Weather and Forecasting 26, 5; 10.1175/WAF-D-11-00021.1
Cross sections (as described in the text) of radar data (a) along the azimuth from 310° (negative range values) to 130° (positive range values) of reflectivity (dBZ) from the BMX radar at 0637 UTC 13 Apr 2009. (b) As in (a), but for radial velocity (m s−1), with positive values representing winds toward positive ranges on both sides of the radar. (c),(d) Similar to (a),(b), but at 2227 UTC 20 Dec 2007, along the azimuth from 300° to 120°. (e),(f) As in (a),(b), but at 1808 UTC 22 Feb 1998 from the FFC radar, along the azimuth from 225° to 45°. See text for an explanation of the lack of COS.
Citation: Weather and Forecasting 26, 5; 10.1175/WAF-D-11-00021.1
(a) National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis of 500-hPa RH (%) at 0600 UTC 13 Apr 2009. (b) North American Mesoscale Model (NAM) east–west cross section of RH along the line in (a), where arrows indicate the approximate location of RIJ, based on radar data. (c) NCEP–NCAR reanalysis of 500-hPa RH (%) at 1800 UTC 20 Dec 2007. (d) As in (a), but at 1800 UTC 22 Feb 1998. [Legends for (a),(c), and (d) are below (c) (Kalnay et al. 1996).]
Citation: Weather and Forecasting 26, 5; 10.1175/WAF-D-11-00021.1
To eliminate the “cone of silence” (COS) in the radar cross sections, a method was employed that assumed the wake lows were two-dimensional and quasi–steady state over a short period of time, 10–20 min (valid assumptions in each of these cases). A time-to-space conversion (thus requiring the steady-state assumption) was performed. The data from an earlier volume scan, when mesoscale features obscured by the COS at the main volume scan time were visible, were advected at the speed of movement of the wake low, and substituted only in the COS region. The data were then smoothed and fitted onto a Cartesian grid, with δx = 1 km and δz = 250 m.
Another inference may be made from Eq. (1). The p′ at a given height in the atmosphere is determined by the integral of all vertical displacements above that point. Also, the vertical displacements at lower levels (with higher density ρ0) have the largest effect on the pressure perturbation below. Therefore, the maximum p′, and therefore the maximum wind perturbations, should occur at or near the surface. This was clearly the case in this event. Doppler velocity images (Fig. 5b) show winds decreasing from −25 m s−1 at 750 m AGL (the sign is relative to the wake low propagation direction) to about 15 m s−1 in the RIJ at 3 km AGL, producing a vertical wind shear of O(10−2 s−1). This wind shear in wake lows may present a hazard to aviation (e.g., Johnson 2001; Meuse et al. 1996).
b. 20 December 2007
Another significant wake low affected much of northern Alabama on 20 December 2007, including the Birmingham (BHM) metropolitan area during the evening rush hour. Numerous trees and power lines were blown down, and a roof was blown off of a business in Decatur, Alabama (NWS 2007). The surface pressure dropped 8 hPa at BHM in about 1 h, with measured wind gusts at BHM reaching 16.5 m s−1. Higher gusts of 20–24 m s−1 were recorded at slightly elevated stations (e.g., Elliott 2010). The wake low was propagating toward the southeast at 13 m s−1. In this case, velocity–azimuth display (VAD) analysis indicates ambient 5 m s−1 winds at 150 m AGL from the east-southeast; therefore, these winds also added to the wake low perturbation winds. The atmosphere over northern Alabama was again characterized by conditional instability aloft, with a layer of more stable air near the surface (the average N below 1800 m MSL was approximately 0.02 s−1; the average N from 1800 to 5000 m was less than 0.01 s−1; see Fig. 3b).
The parent MCS in this case was similar in morphology to that in section 2a, with an asymmetrical shape. Radar cross sections (Figs. 5c and 5d) show a strong, descending RIJ, not penetrating the stratiform rain area, associated with rapid descent and a sharp horizontal reflectivity gradient (up to 5 dB km−1) at the back edge of the rain. Once again, dry air was present to the northwest of the MCS, in the source region of RIJ air (Fig. 6b). The RIJ and wake low combined to produce large vertical wind shear of around 0.016 s−1.
c. 22 February 1998
A very large-amplitude wake low moved across almost the entire state of Alabama, and parts of Georgia, on 22 February 1998, producing widespread structural damage, downed trees, and power outages (e.g., Bradshaw et al. 1999). Pressure falls were not only large, but occurred very quickly (more than 8 hPa in less than 30 min at Birmingham and Anniston, Alabama; see Fig. 2c). Peak wind gusts at most Automated Surface Observing System (ASOS) stations were 15–20 m s−1, but wind damage at higher elevations around Birmingham indicated wind gusts in excess of 30 m s−1, and Birmingham–Shuttlesworth International Airport’s analog wind recorder indicated a peak wind of 25 m s−1 (49 kt; Fig. 2d) (Bradshaw et al. 1999).
One possible reason for the rapid pressure falls was likely the depth of the low-level stable layer. In this case, the sounding from BMX (station located south of Birmingham) indicates an average N of 0.014 s−1 up to 4000 m MSL, with much less stable air (average N = 0.009 s−1) above that. The parent MCS was structurally different from than those discussed in sections 2a and 2b, in that the wake low itself was only about 100 km directly behind the most intense convection. This MCS and trailing wake low moved toward the NE (from 225°) at a rather fast 25 m s−1. Bradshaw et al. (1999) point out that the synoptic environment was favorable for gravity wave generation (e.g., Uccellini and Koch 1987). The wind event discussed herein was a wake low, but it is possible that the entire MCS could have been associated with a long-wavelength internal gravity wave, in a wave-conditional instability of the second kind (CISK; e.g., Lindzen 1974) process. The fairly rapid movement of this wake low, and the ambient low-level winds being nearly parallel to the wake low’s movement, likely reduced the winds somewhat (e.g., Coleman and Knupp 2009). However, the very large magnitude of the pressure falls still indicates that wind perturbations should be around 20 m s−1, according to the nonlinear impedance relation (Coleman and Knupp 2010), and this is consistent with surface observations.
Radar data (Figs. 4c and 5c) show a tight reflectivity gradient extending over 300 km horizontally. This wake low was associated with a very intense, descending RIJ, with wind speeds in the RIJ of 50 m s−1. However, as in the other two cases presented herein, the RIJ is “blocked” (does not make it through the stratiform rain area). As a matter of fact, horizontal convergence values around 10−3 s−1 are observed over a fairly large horizontal extent between 2 and 4 km MSL, indicating the likelihood of a strong downdraft. Dry air was present at 500 hPa, upstream from the MCS (Fig. 6c), but farther way than in the previous two cases. However, the much stronger RIJ (near 50 m s−1) likely advected some of the dry air southwest of the MCS into the stratiform precipitation area. The RIJ and wake low–associated winds below produced extreme vertical shear, around 0.02 s−1.
3. Summary and conclusions
The cases examined herein demonstrate that intense wake lows may produce large areas of strong winds and wind damage. The large pressure falls produced numerous observations of winds of 15–25 m s−1. These wind speeds, combined with the large-scale damage, qualify these three wake low events as severe local storms. The extreme wind shear and associated turbulence in wake lows may also be an unrecognized and potentially significant hazard to aviation.
These cases show several observations common to all three wake lows, including 1) a relatively stable layer near the surface, with conditionally unstable air aloft, conducive to elevated thunderstorms and large pressure falls at the surface; 2) a strong, descending RIJ “blocked” in the stratiform precipitation region (shown in Doppler radar velocity cross sections), producing strong downdrafts there; 3) dry air at upper levels behind the parent MCS; and 4) a sharp horizontal reflectivity gradient (4–5 dB km−1) at the back edge of the stratiform precipitation region due to subsidence. Two of the three wake lows were also accompanied by ambient surface winds in the opposite direction of the wake low propagation. Operational forecasters may apply these observations to specific MCSs to ascertain the possibility of a wake low event.
Acknowledgments
The authors wish to thank Declan Cannon (NWS) for thoughtful discussions. The authors also wish to thank the reviewers. This research was funded by a grant from the National Oceanic and Atmospheric Administration (NOAA Grant NA07OAR4600493).
REFERENCES
Bernstein, B. C., and Johnson R. H. , 1994: A dual-Doppler radar study of an OK PRE-STORM heat burst event. Mon. Wea. Rev., 122, 259–273.
Bosart, L. F., and Seimon A. , 1988: Case study of an unusually intense atmospheric gravity wave. Mon. Wea. Rev., 116, 1857–1886.
Bradshaw, J. T., Murphy R. A. , and Pence K. J. , 1999: The Alabama gravity wave event of February 22, 1998: A classic mesoscale forecasting challenge. Preprints, Eighth Conf. on Mesoscale Processes, Boulder, CO, Amer. Meteor. Soc., 130–135.
Coleman, T. A., 2008: The interactions of gravity waves with mesocyclones and tornadoes. Ph.D. dissertation, University of Alabama in Huntsville, 220 pp.
Coleman, T. A., and Knupp K. R. , 2009: Factors affecting surface wind speeds in gravity waves and wake lows. Wea. Forecasting, 24, 1664–1679.
Coleman, T. A., and Knupp K. R. , 2010: A nonlinear impedance relation for the surface winds in pressure disturbances. J. Atmos. Sci., 67, 3409–3422.
Elliott, J. B., cited 2010: Additional reports. [Available online at http://www.alabamawx.com/?p=4607.]
Fujita, T. T., 1955: Results of detailed synoptic studies of squall lines. Tellus, 7, 405–436.
Gaffin, D. M., 1999: Wake low severe wind events in the Mississippi River valley: A case study of two contrasting events. Wea. Forecasting, 14, 581–603.
Gallus, W. A., Jr., and Johnson R. H. , 1995: The dynamics of circulations within the trailing stratiform regions of squall lines. Part I: The 10–11 June PRE-STORM system. J. Atmos. Sci., 52, 2161–2187.
Haertel, P. T., and Johnson R. H. , 2000: The linear dynamics of squall line mesohighs and wake lows. J. Atmos. Sci., 57, 93–107.
Handel, R. R., and Santos P. , 2005: Observations from the April 13 2004 wake low damaging wind event in south Florida. NWA Elec. J. Oper. Meteor., 2005-EJ7. [Available online at http://www.nwas.org/ej/pdf/2005-EJ7.pdf.]
Johnson, R. H., 2001: Surface mesohighs and mesolows. Bull. Amer. Meteor. Soc., 82, 13–31.
Johnson, R. H., and Hamilton P. J. , 1988: The relationship of surface pressure features to the precipitation and airflow structure of an intense midlatitude squall line. Mon. Wea. Rev., 116, 1444–1473.
Johnson, R. H., and Bartels D. L. , 1992: Circulations associated with a mature-to-decaying midlatitude mesoscale convective system. Part II: Upper-level features. Mon. Wea. Rev., 120, 1301–1321.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437–471.
Koch, S. E., and O’Handley C. , 1997: Operational forecasting and detection of mesoscale gravity waves. Wea. Forecasting, 12, 253–281.
Lindzen, R. S., 1974: Wave-CISK in the tropics. J. Atmos. Sci., 31, 156–179.
Lindzen, R. S., and Tung K.-K. , 1976: Banded convective activity and ducted gravity waves. Mon. Wea. Rev., 104, 1602–1617.
Loehrer, S. M., and Johnson R. H. , 1995: Surface pressure and precipitation life cycle characteristics of PRE-STORM mesoscale convective systems. Mon. Wea. Rev., 123, 600–621.
Meuse, C., Galusha L. , Isaminger M. , Moore M. , Rhoda D. , Robasky F. , and Wolfson M. , 1996: Analysis of the 12 April 1996 wind shear incident at DFW airport. Preprints, Workshop on Wind Shear and Wind Shear Alert Systems, Oklahoma City, OK, Amer. Meteor. Soc., 23–33.
NCDC, cited 2011: Georgia. NCDC storm events. [Available online at http://www4.ncdc.noaa.gov/cgi-win/wwcgi.dll?wwEvent~Storms.]
NWS, 2007: Alabama. Storm data and unusual weather phenomena, December 2007. National Weather Service Rep., 3 pp. [Available online at http://www.srh.noaa.gov/images/bmx/stormdat/2007/bmxdec2007.pdf.]
NWS, 2009: Alabama. Storm data and unusual weather phenomena, April 2009. National Weather Service Rep., 41 pp. [Available online at http://www.srh.noaa.gov/images/bmx/stormdat/2009/bmxapr2009.pdf.]
Schmidt, J. M., and Cotton W. R. , 1990: Interactions between upper and lower tropospheric gravity waves on squall line structure and maintenance. J. Atmos. Sci., 47, 1205–1222.
Skamarock, W. C., Weisman M. L. , and Klemp J. B. , 1994: Three-dimensional evolution of simulated long-lived squall lines. J. Atmos. Sci., 51, 2563–2584.
Smull, B. F., and Jorgensen D. P. , 1990: Pressure and buoyancy perturbations near an intense wake low in a midlatitude mesoscale convective system. Preprints, Fourth Conf. on Mesoscale Processes, Boulder, CO, Amer. Meteor. Soc., 214–215.
Smull, B. F., Jorgensen D. P. , and Hane C. E. , 1991: Comparison of retrieved pressure and buoyancy perturbations with in situ observations of an intense wake low in a midlatitude mesoscale convective system. Preprints, 25th Int. Conf. on Radar Meteorology, Paris, France, Amer. Meteor. Soc., 135–138.
Stumpf, G. J., Johnson R. H. , and Smull B. F. , 1991: The wake low in a midlatitude mesoscale convective system having complex convective organization. Mon. Wea. Rev., 119, 134–158.
Uccellini, L. W., and Koch S. E. , 1987: The synoptic setting and possible energy sources for mesoscale wave disturbances. Mon. Wea. Rev., 115, 721–729.
