WEST TEXAS MESONET OBSERVATIONS OF WAKE LOWSWEST TEXAS MESONET OBSERVATIONS OF WAKE LOWSAND HEAT BURSTS ACROSS NORTHWEST TEXASAND HEAT BURSTS ACROSS NORTHWEST TEXAS

Mark Conder, Steve Cobb, and Gary Skwira, Mark Conder, Steve Cobb, and Gary Skwira, National Weather Service, Lubbock TXNational Weather Service, Lubbock TX

SSUMMARY AND CONCLUSIONSUMMARY AND CONCLUSIONS

Wake lows (WLs) and heat bursts (HBs) are mesoscale phenomena associated with thunderstorms that can result in strong and sometimes severe (~26 m s-1), near-surface winds. The operational detection of severe winds with WLs/HBs is problematic because the associated convection can appear to be innocuous via WSR-88D data. The thermodynamic environment that supports WL and HB development is similar, and is often compared to the classic “onion” sounding structure as described by Zipser (1977). Observational studies such as Johnson et al. (1989) show that nearly dry-adiabatic lapse rates are prevalent in the lower to mid-troposphere, with a shallow stable layer near ground level. This type of environment is most commonly observed in the high plains during the warm season.

While documented heat bursts were infrequent in the past, the recent expansion of surface mesonetworks are increasing the likelihood that these events will be sampled. One such network is the West Texas Mesonet (WTXM), a collection of over 40 meteorological stations spread across northwest Texas. During the period 1 June 2004 to 30 August 2006, the authors have recorded 10 WL/HB events using the WTXM data. Some of the more extreme measurements include a 15 °C increase in temperature at the Pampa and McClean WTXM sites in one event and 35 m s-1 wind gusts at sites in Brownfield and Jayton on separate occasions. Additionally, several of the events have caused considerable property damage.

In this study, the mesoscale environments of 10 WLs/HBs were investigated through analysis of WTXM data, and augmented by radar, satellite and upper air data, in the hope of establishing some of the typical characteristics to aid forecasters in recognizing their onset.

Sequence of surface divergence plots along with wind barbs for 0530 UTC, 0600 UTC, and 0630 UTC, 7 Jul 2004. A wind gust of 27.0 m s-1 occurred at TAHO at 0555 UTC, with a maximum gust of 30.2 m s-1 at MACY at 0625Z. Divergence is in units of 10-4 s-1. ►

The six wake low cases associated with strong winds that were identified along with maximum pressure drop and wind gust observed for each event (within one hour of onset). ▼

SSTUDY DOMAIN AND THE WEST TEXAS MESONETTUDY DOMAIN AND THE WEST TEXAS MESONET

KLBB WSR-88D composite reflectivity with surface observations overlaid from 0556 UTC and 0626 UTC, 7July 2004. Note the wedge of lower reflectivity in the vicinity of Tahoka around the time of the peak wind gusts. ▼

1. The five-minute West Texas Mesonet observations allow near real-time tracking of wake lows and heatbursts through the examination of the pressure and temperature readings, along with calculated fields such as divergence and pressure tendency

2. However, the data has not revealed a reliable method using the surface data exclusively to predict the onset of severe wind speeds associated with these features

3. Examination of satellite and radar data may aid the determination of areas susceptible to HB/WL-induced high winds. Specifically, IR imagery may show areas of dry-air subsidence (wedges in the mid-level cloud cover), while radar data may be used to monitor the evolution of the trailing stratiform region with an emphasis on the north or northwest portion of the TSR and decreasing reflectivity trends 4. In some cases (about half the ones examined so far), detectable pressure falls in (the case of wake lows) and temperature rises/dewpoint falls (in the case of heatburst) do occur with in advance of the strong winds and careful analysis of the 5-minute WTXM data can provide a 5 to 15 minute lead time

5. A simple detection algorithm will be developed and tested by the NWS Lubbock office in 2007 to determine if it can increase the warning lead- time for these events

1. The five-minute West Texas Mesonet observations allow near real-time tracking of wake lows and heatbursts through the examination of the pressure and temperature readings, along with calculated fields such as divergence and pressure tendency

2. However, the data has not revealed a reliable method using the surface data exclusively to predict the onset of severe wind speeds associated with these features

3. Examination of satellite and radar data may aid the determination of areas susceptible to HB/WL-induced high winds. Specifically, IR imagery may show areas of dry-air subsidence (wedges in the mid-level cloud cover), while radar data may be used to monitor the evolution of the trailing stratiform region with an emphasis on the north or northwest portion of the TSR and decreasing reflectivity trends 4. In some cases (about half the ones examined so far), detectable pressure falls in (the case of wake lows) and temperature rises/dewpoint falls (in the case of heatburst) do occur with in advance of the strong winds and careful analysis of the 5-minute WTXM data can provide a 5 to 15 minute lead time

5. A simple detection algorithm will be developed and tested by the NWS Lubbock office in 2007 to determine if it can increase the warning lead- time for these events

HHEATBURSTSEATBURSTS

The table below shows the heat burst cases in this study along with maximum change in temperature (Max ∆T) and dewpoint (Max ∆Td) and maximum wind gust for each event. ▼

AACKNOWLEDGMENTSCKNOWLEDGMENTSAACKNOWLEDGMENTSCKNOWLEDGMENTSThe authors wish to thank Wes Burgett of the West Texas Mesonet for helping with data collection, John Lipe of NWSFO-LUB for helping with data processing, and Todd Lindley of NWSFO-LUB for reviewing the paper.

TTHE HEATBURST OF 18 SEPTEMBER 2005HE HEATBURST OF 18 SEPTEMBER 2005

Date (UTC)Maximum

Pressure DropMaximumWind Gust

7 July 2004 -1.7 mb 29.3 m s-1

4 June 2005 -6.4 mb 28.8 m s-1

6 June 2005 -4.4 mb 35.1 m s-1

18 Sept 2005 -4.1 mb 25.6 m s-1

27 May 2006 -3.7 mb 34.2 m s-1

21 June 2006 -5.8 mb 28.4 m s-1

Date (UTC)MaxΔT

MaxΔTd

MaxWind Gust

23 May 2005 7.8 °C -11.0 °C 19.9 m s-1

4 June 2005 3.3 °C -9.4 °C 28.5 m s-1

6 June 2005 5.8 °C -7.2 °C 35.1 m s-1

18 Sept 2005 6.1 °C -9.1 °C 25.6 m s-1

27 May 2006 7.2 °C -15.0 °C 34.20 m s-1

21 June 2006 6.7 °C -13.3 °C 28.35 m s-1

12 July 2006 5.6 °C -1.7 °C 28.80 m s-1

WWAKEAKE LOWSLOWS

TTHE WAKE LOW OF 7 JULY 2004HE WAKE LOW OF 7 JULY 2004

Station

Time of HB Onset (UTC)

10 minute maximum values 60 minute maximum values

ΔT(°C)

ΔTd

(°C)ΔWSm s-1

Gustm s-1

ΔT(°C)

ΔTd

(°C)ΔWSm s-1

Gustm s-1

LEVE 03:40 3.2 -3.2 5.0 12.9 3.4 -3.2 11.8 19.8

REES 03:55 1.4 -3.4 1.9 10.7 4.3 -8.5 6.8 14.0

LBBW 04:20 1.8 -4.0 -1.0 7.5 5.4 -8.8 11.7 23.6

ANTO 04:30 0.8 -1.0 2.4 8.0 6.1 -7.7 12.7 18.9

ABER 04:30 2.4 -3.9 3.9 9.4 3.0 -5.2 3.9 9.4

SLAT 04:30 5.0 -8.3 10.5 17.0 5.2 -9.1 15.6 25.0

RALL 05:40 1.9 -3.9 6.3 18.5 4.4 -7.1 14.5 25.6

WHIT 05:45 0.7 -1.3 2.4 12.2 3.2 -7.4 5.4 16.7

Damage sustained to a sheet metal roof at Memorial Baptist Church in Lubbock, TX, on 18 Sept 2006. The sheet metal roof, which served as a large hail and rain roof about 6 inches above a permanent roof, was pealed off and became airborne during the heatburst event. The roof landed on trees and a residence to the north side of the church. ►

Comparison of the 0000 and 1200 UTC, 18 Sep 2005 skew-T plots for Amarillo (KAMA) and Midland (KMAF), TX. Note the deep dry adiabatic layer to near 500 mb, except for a shallow inversion near the surface by morning. ▼The table below is a summary of West Texas Mesonet Station data for 17 Sep 2005, listed in order from west to east. ∆T and ∆Td

refer to the difference between the initial temperature and the highest (in the case of ∆T) or lowest (in the case of ∆Td) values in the subsequent 10 or 60-minute period. ∆WS refers to the difference between the 5-minute average wind speed at the HB onset time and the peak 5-minute average wind speed recorded in the 10 or 60-minute period. Gusts refers to the peak 3-second gust recorded during the 10 or 60-minute period.

For the purposes of this study, HEAT BURST ONSET is defined as time of the simultaneous rise in temperature and decrease in dewpoint of any magnitude that preceded a pronounced heat burst event.

◄(Left) Time series of 5-minute observations from the West Texas Mesonet LBBW Station (located in northwest Lubbock) for 0300 to 0700 UTC, 18 Sep 2005 (10 pm to 2 am local time). All variables are 5-minute averages except for “GUM”, which is the peak 3-second gust in each 5-minute period.

(Right) Objective analysis of temperature (°C, solid yellow contours) and dewpoint (°C, dashed blue contours) for 0400, 0500, and 0600 UTC, 18 Sep 2005 – showing the progression of the heatburst near the center of the domain ►

Live and archived data from the West Texas Mesonet is available at:

www.mesonet.ttu.eduwww.mesonet.ttu.edu

AABSTRACT / INTRODUCTIONBSTRACT / INTRODUCTION

0400 0500 0600

The graph below shows the same station data from the graph normalized with respect to the time of the heatburst onset and the temperature/dewpoint. Notice the temperature increase and dewpoint decrease plateaus approximately 20 to 30 minutes after onset. ▼

The graph below shows the average (solid marker) and standard deviation (smaller, open marker) for all the stations (n=32) which showed a heatburst signature in the seven events. The red line depicts the temperature rise and the blue line depicts the dewpoint fall. ▼

The schematic below shows two possible heatburst mechanisms associated with the wake low region of a MCS. Isolated thunderstorm cells in the right thermodynamic environment can also trigger heatbursts (figure from Johnson, 2001). ▼

The schematic below (from Johnson and Hamilton, 1988) shows the positions of the mesohigh and wake low relative to the trailing stratiform region of a symmetric MCS (squall line-type). ▼

Comparison of the 0000 and 1200 UTC, 7 Jul 2004 skew-T plots for Amarillo (KAMA) and Midland (KMAF), TX. Note the deep dry adiabatic layers from 700 to 500 mb. ▼

0556 UTC 0626 UTC

0530 UTC 0600 UTC 0630 UTC

This wake low resulted in severe wind gusts reports of 30.2 m s-1 at MACY and 27.0 m s-1 at TAHO between 0500 and 0700 UTC. Scattered thunderstorms that developed during the evening hours across the western and central portions of the domain evolved into a squall line exhibiting a typical symmetric trailing stratiform MCS structure by 0400 UTC (not shown). By 0556 UTC, composite reflectivity (see images below) revealed that the MCS had entered the dissipating stage. One to 2 mb hr-1 pressure falls began to develop in the vicinity of TAHO on the back edge of the TSR, behind an area of pressure rises associated with the mesohigh. The image to the right shows the pressure tendency (mb hr-1 * 10) near the time of the peak gust at TAHO. Notice that TAHO is located at the interface between the mesohigh and wake low.

0555 UTC

Previous studies have shown that wake lows typically form during the decaying stage of a mesoscale convective system (MCS) near the back edge of the trailing stratiform precipitation region (TSR). In this area, parcel descent is aided by a dry subcloud layer where evaporative cooling is inadequate to maintain saturation or offset adiabatic warming; and the parcel continues to the surface despite being warmer than its environment. Additionally, rear-to-front flow aloft within the TSR, due to a descending rear-inflow jet, may enhance subsidence warming. WLs are commonly associated with pressure falls of 4 mb hr-1 and can be tracked through the pattern of pressure falls or surface divergence. The pressure difference between the WL and the mesohigh can also lead to a strong pressure gradient between the two regions.

Heatbursts are exemplified by rapid temperature increases and dewpoint decreases noted in surface observations – usually accompanied by gusty winds. They are primarily a nocturnal phenomenon, resulting from descending, adiabatically-warmed air penetrating a shallow nocturnal inversion near the ground. Previous studies have hypothesized that the drying is in part due to an insufficient rate of evaporation of the water droplets in the downdraft to maintain saturation during their descent to the surface. This hypothesis was supported by Fujita (1985) -who also noted that a drop in dewpoint temperature could be caused/enhanced by the entrainment of drier environmental air into the downdraft. Heatbursts may be rare because downdrafts usually do not possess sufficient momentum to reach the surface upon encountering a deep stable layer normally present in the Zipser-type thermodynamic environment. Thus, the speed of the downdraft and the depth of the stable layer will determine whether or not the downdraft is sufficient to breech the stable layer.

Archived data from various stations of the West Texas Mesonet was used in this study. Each station reports five-minute averages of temperature, dewpoint temperature, wind speed and direction, station pressure and altimeter setting, along with the three-second peak wind gust.