SRH SSD 2006-03
6/2006
Technical Attachment
Widespread Wind Damage from 2 June 2004 Derecho
in the ArkLaTex
Bill Murrell, Forecaster
WFO Shreveport, LA
I. Introduction
Northwest flow is notorious for producing severe weather
episodes in late spring into early summer in southwest
Arkansas, northern Louisiana, and northeast Texas, commonly
referred to as the ArkLaTex. Complexes of thunderstorms
develop in the Plains with the aid of a nocturnal low
level jet and interaction with a mid-level disturbance
or frontal boundaries. These thunderstorms then usually
propagate southeast not only toward the Shreveport (SHV)
Weather Forecast Office (WFO) forecast area, but into
a very moist and unstable environment. However, these
complexes of storms usually arrive during the morning
hours, which is typically when these mesoscale convective
complexes (MCC) begin to weaken or dissipate all together.
This weakening of the MCC is due to the dissipation of
the nocturnal low-level jet and stabilization of the atmosphere.
Residual boundaries left over from these complexes have
the potential to help initiate thunderstorms later in
the afternoon.
Around sunrise on June 2nd, the second of three complexes
of storms that affected the area moved southeast out of
the ArkLaTex. Subsidence behind this intense complex of
storms allowed for plenty of insolation across the region
with temperatures warming to near 90 degrees. These temperatures,
combined with high dew points in the upper 60s and lower
70s, resulted in very unstable conditions south of an
approaching cold front, which was located across northern
Oklahoma and northwest Arkansas. In the afternoon of June
2nd, convection formed on the boundary across northern
Oklahoma, and propagated southeast during the afternoon
and early evening into the SHV WFO forecast area when
the atmosphere was very unstable.
The derecho event was initiated by a weak shortwave trough
in the northwest flow aloft and mesoscale forcing associated
with the front. These storms surged southeast and moved
through the ArkLaTex during the late afternoon into the
evening producing a plethora of damage from straight-line
winds as high as 85 mph.
II. Derecho Conceptual
Model
By definition a “derecho” is a convective
system that produces wind damage from gusts greater than
26 m s-1 (50 kt) for at least a length of 250 nautical
miles. The damage reports from these high winds must also
occur in a progressive pattern (nonrandom) with no more
than a 3-hour gap between reports (Weisman 1993). Johns
and Hirt (1987) describe two different synoptic patterns
for derecho events. One synoptic pattern that leads to
the formation of derecho events occurs typically during
the winter and spring months associated with mid-latitude
cyclones and associated strong cold fronts moving across
the country. The other pattern is common in the late spring
and summer with weak synoptic systems in westerly to northwesterly
500 mb flow. Over 75% of derecho events begin close to
peak heating of the day along and to the north of a west
to east oriented boundary (Johns 1987). In addition, “pooling”
of low-level moisture near a quasi-stationary front is
a common characteristic of derecho events (Johns 1987).
Extreme instability and moderate wind shear are necessary
along the track of the convective system in order to maintain
itself long enough to be classified as a derecho event.
Model simulations suggest the thermodynamic instability
and vertical wind shear profile conducive for derecho
events were convective available potential energy (CAPE)
of 2000 J/kg or greater, with a wind shear of at least
20 m s-1 over the lowest 2.5 - 5 km AGL (Weisman 1993).
|
Figure 1. The four stages of the
evolution of an idealized bow echo (Weisman 1993). |
Once the environmental parameters come together to initiate
a derecho event, there are storm scale processes that
greatly impact the evolution of such an event. Weisman
(1993) developed a schematic diagram showing four stages
for an idealized bow echo (Figure 1). This scenario can
also be applied to a squall line or derecho event. Initially,
convection developing in a flow pattern with sufficient
shear will tilt downshear from the ambient wind shear
(Figure 1a). A cold pool will eventually develop beneath
the storms generating a horizontal circulation opposite
of the circulation generated by the ambient wind shear.
This results in a more vertical updraft and stronger convection
(Figure 1b). Eventually the cold pool circulation becomes
stronger than the ambient wind shear circulation, and
the updraft begins to tilt rearward (upshear) over the
“cold pool” at the surface (Figure 1c). This
is the third stage in an idealized bow echo evolution
usually beginning after 2 hours. At this stage, an elevated
rear-inflow jet (RIJ) begins to develop. The final stage
(Figure 1d) is when the RIJ in combination with the ambient
wind shear balances the cold pool circulation resulting
in deeper lifting (more vertical updraft). However, enough
instability and low-level vertical wind shear must be
present to produce a RIJ strong enough, when combined
with the ambient wind shear, to balance the circulation
of the cold pool. Otherwise, the convection begins to
decay as lifting along the gust front becomes weaker due
to the cold pool circulation being the dominant circulation
(Weisman 1993).
III. Analysis of the Environment
On the morning of June 2nd, surface analyses indicated
a cold front from northern Oklahoma into northern Arkansas,
and a decaying squall line across south central Mississippi
and southeast Louisiana (Figure 2). Local 12 UTC soundings
indicated that with daytime heating the atmosphere would
become unstable by afternoon with Lifted Index values
near -5 and CAPE near 2000 J/kg. Low level flow through
850 mb was generally weak, but moisture was already established
south of the front with surface dew points in the lower
to mid 60s. The pooling of moisture near the front was
evident by the surface equivalent potential temperature
(theta-e) gradient within one hundred miles of the front,
decreasing from 352K in central OK to 324K near the Kansas
border. In addition, there was an 850 mb theta-e ridge
max just south of the frontal boundary of 332K to 340K.
At 500 mb, weak disturbances were embedded in the west
northwest flow across the region. The 500 mb flow near
the frontal boundary in northern Oklahoma during the day
on June 2nd was a bit stronger than usual for early summer
between 40 and 50 kt, with 0 – 6 km shear between
20 and 25 m s-1. Farther south into the ArkLaTex, the
0 – 6 km shear was only around 12.5 m s-1. In addition,
the Southern Plains was in the right rear quadrant of
an 80 kt jet streak.
During the morning, ongoing convection was north of the
boundary in northern Oklahoma and southern Kansas with
abundant solar insolation south of the boundary. In the
wake of the previous night’s severe convection across
east Texas and north Louisiana, strong subsidence only
allowed a few cumulus clouds to form despite temperatures
near 90 and dew points in the mid 60s to near 70. A weak
disturbance in the northwest flow, in combination with
some upper level divergence and temperatures rising into
the mid to upper 80s, helped break a small capping inversion
(CAP) south of the frontal boundary across east central
Oklahoma, allowing scattered to numerous thunderstorms
to develop by midday. These thunderstorms quickly became
a squall line and headed south southeast toward the ArkLaTex
in the northwest flow aloft.
These thunderstorms were moving into an increasingly
unstable environment. By mid afternoon, CAPE values were
near 3500 J/kg, with lifted index values near -10 ahead
of the approaching squall line. LAPS surface helicity
was on the low side, with values between 50 and 75 m2/s2.
Ahead of the destructive squall line, Shreveport’s
00 UTC 3 June 2004 sounding indicated a very unstable
atmosphere with a LI of -9, CAPE of 3000 J/kg, SWEAT of
526, and Total Totals of 57. Additionally, there was high
moisture content with 1.72 inches of precipitable water.
The 0 to 3 km helicity was only 64 m2/s2.
IV. Model Data
The 00 UTC 2 June 2006 GFS, NAM, and NGM models did indicate
the initial squall line development late in the evening
on 1 June 2004. This squall line formed around 04 UTC
2 June 2004 over northeast Texas, moving through east
Texas and much of Louisiana during the early morning of
2 June 2004. Therefore, instead of scattered thunderstorms
as predicted by all three models for the daytime on 2
June 2004, intense insolation with no convection occurred
across the region. This occurred due to subsidence in
the wake of the severe line of thunderstorms in combination
with a weak CAP in place. The convection that formed around
midday on 2 June 2004 was closer to the front in Oklahoma
and Arkansas where the CAP was more easily broken. Models
did indicate if convection formed, it would move southeast
in the northwest flow aloft. LAPS and MSAS data in the
afternoon revealed the squall line was moving into an
extremely unstable atmosphere which could sustain the
squall line through the area despite outrunning the front
and upper level dynamics.
V. Radar Data
The squall line entered McCurtain County in southeast
Oklahoma around 2220 UTC 2 June 2004 and exited the southeast
part of the SHV CWA by around 05 UTC 3 June 2004. This
line of storms, when it was moving through southwest Arkansas
into extreme northeast Texas, was 350 miles long and about
30 miles wide (Figure 3). It extended from Mount Ida,
Arkansas, through Texarkana to Wichita Falls, Texas. A
few Mid-Altitude Radial Convergence (MARC) signatures
were present as these storms moved through southeast Oklahoma
and southwest Arkansas (Figures 4 and 5). These signatures
were quite strong with 110 knots of convergence between
10 and 15 thousand feet. Base velocities began to show
strong inbound winds as the line of storms moved to within
65 miles of the Shreveport WSR-88D. The Shreveport WSR-88D
estimated 75 knots of wind toward the radar around 7000
feet. An observed wind of 58 kts (67 mph) was recorded
at Texarkana Regional Webb Field at 0007 UTC 3 June 2004.
The base velocity image around the same time indicated
80 kts of inbound velocities just west into Bowie County
between Texarkana and New Boston with around 55 kts of
inbound velocities near Texarkana.
Reflectivity was not as intense across east Texas, but
an intense gust front pushed well ahead of the line of
storms. This was most likely the result of the cold pool
circulation of the derecho dominating the ambient wind
shear across east Texas. The gust front probably did as
much or more damage than the actual line of storms across
east Texas. Gregg County Airport (GGG) recorded a peak
wind speed of 42 knots at 0117 UTC 3 June 2004, while
the convection was still 15 miles to the north on the
Gregg and Upshur County line (Figure 6). Tyler Pounds
Field ASOS equipment malfunctioned as the gust front pushed
through. However, FAA personnel estimated a gust to 45
knots well ahead of the storms. Radar and ASOS data also
shows clearly that Downtown Shreveport (DTN) had its highest
wind speed before the line of storms. Although, the line
of storms and gust front were much closer together across
northwest Louisiana where more ambient wind shear was
present. This is not to say wind damage did not occur
within the storms across east Texas, but the data suggests
most of the strongest winds across east Texas occurred
along the gust front.
VI. Damage Reports
Widespread wind damage was reported with this derecho
event. An estimated 2.2 million dollars of damage occurred
in the WFO Shreveport’s area of responsibility.
Wind was by far the primary cause of damage with only
a few hail reports. The ratio was roughly seven to one.
The worst damage occurred across southwest Arkansas and
parts of northeast Texas and north Louisiana. One fatality
occurred in Franklin County in northeast Texas when a
tree fell on top of a mobile home. Another 6 people were
known to be injured during this event. Interstate 30 was
closed from Hope to Prescott Arkansas due to downed trees
and overturned trucks. Other state highways in Arkansas
were also closed for the same reason. The Union Parish
Office of Emergency Preparedness declared a state of emergency
for the entire parish due to widespread damage caused
by downed trees. Wind damage occurred in 47 out of 48
counties and parishes in WFO Shreveport’s CWA.
VII. Conclusion
Although the environment for intense organized storms
was not favorable (lack of wind shear) across Shreveport’s
forecast area June 2nd, the atmosphere just to the north
had the necessary ingredients for explosive development
once convective temperatures were met. As the thunderstorms
across north Oklahoma developed, they quickly evolved
into a line and moved southeast in the northwest flow
into an area of strong instability. The gust front outrunning
the line of convection across east Texas suggested that
the cold pool circulation dominated the environmental
ambient shear.
This event further highlights the need for meteorologists
to have and maintain situational awareness during severe
weather events. In this case, warnings were needed well
ahead of the derecho in east Texas along the distinct
gust front. A best practice is to include specific wording
about the gust front in the warnings. For example, WFO
Shreveport uses the canned statement “Wind damage
with this line of storms will occur well ahead of any
rain or lightning. Do not wait for the sound of thunder
before taking cover”. If a severe gust front or
outflow boundary moves too far from the parent thunderstorm
complex to justify a severe thunderstorm warning, a high
wind warning might be appropriate.
VIII. References
Johns, R. H., and W. D.
Hirt, 1987: Derechos: widespread convectively induced
windstorms.
Weather and Forecasting, 2, 32-49.
Weisman, Dr. Morris L., 1993: The Genesis of Severe, Long-Lived
Bow Echoes
Journal of the Atmospheric Sciences, 50, 645-670.
|
Figure 2. The surface weather analysis
for 12 UTC 2 June 2004. It shows an exiting squall
line across Mississippi and southeast Louisiana and
a quasi-stationary front across northern Oklahoma. |
|
Figure 3. A radar mosaic of the
350 mile long severe squall line during the evening
of 2 June 2004. |
|
Figure 4. A mid-altitude radial
convergence (MARC) signature present in the storm-relative
motion velocity product taken at 2346 UTC on 2 June
2004. |
|
Figure 5. A reflectivity image
at the same time the MARC signature is present in
Figure 4. |
|
Figure 6. The reflectivity image
at 0118 UTC 3 June 2004 showing the gust front 15
miles ahead of the line of thunderstorms. At the
same time (0117 UTC) Gregg County Airport (KGGG)
recorded a wind gust of 42 knots. That was the highest
gust for KGGG during this event. |
|