A Review of Derechos and the Role of Upper-Level Wind Shear on the Maintenance of the Associated Convective Systems

1
A Review of Derechos and the Role of Upper-Level
Wind Shear on the Maintenance of the Associated
Convective Systems

• Michael C. Coniglio
• National Research Council/NOAA/NSSL

2
Outline

• Derecho Overview
• Definition, climatology, environments,
  forecasting issues
• Theories for strength and longevity of strong,
  linear MCSs
• Some issues with RKW Theory (Rotunno et al.
  1988 2004 JAS)
• Alternative Idea
• Numerical simulations
• connection to observations

3
What is a Derecho?

• In general, a long swath (gt 400 km) of widespread
  severe wind gusts (gt 26 m s-1) produced by an
  extratropical Mesoscale Convective System (MCS)
• Johns and Hirt (1987 WAF)
• Need occasional stronger wind gusts
• Bentley and Mote (1998 BAMS)
• Include other convective structures
• Coniglio and Stensrud (2004 WAF)
• Compromise between above two definitions

4
Derecho Characteristics

• Derecho-producing MCSs often contain bow echoes
• Occur on wide range of scales (10-400 km and 20
  minutes to several hours)
• Generally reflect forward advancement of
  convectively cooled outflow (larger scale outflow
  termed cold pool)

Fujita (1978)
5
Derecho Characteristics

• Some derecho-producing MCSs contain embedded
  supercells and/or cyclonic circulations
• Most severe wind damage often observed with these
  circulations NORTH of the bow-echo apex

Atkins et al. (2004) AMS 22nd Conference on
Severe Local Storms
6
Example

• Progressive Derecho 27-28 May 2001

see Miller et al. (2002) AMS 21st Conference on
Severe Local Storms
7
Example

• Serial Derecho 13 May 2002

8
Why do we care?

• Propensity for extreme winds

Bob Johns, SPC Norman, OK
9
Why do we care?

• Boundary-Waters Derecho July 4, 1999
• 477,000 acres blown down
• Sustained winds gt 58 mph for 30-40 minutes
• Peak gusts gt 100 mph

100 km
USDA Forest Service, Superior National Forest
10
Why do we care?

• Boundary-Waters Derecho July 4, 1999

Minnesota Department of Natural Resources
11
Why do we care?

• 1986-2003 fatalities due to
• Derechos 154
• Hurricanes 254
• F0 F1 Tornadoes 71
• F0, F1 F2 Tornadoes 229
• Insured Losses
• July 16, 1980 event - 1.3 billion
• (Isabel 1.6 billion)
• May 31, 1998 event - 431 million
• (Bonnie 394 million)

Ashley and Mote (2004) AMS 22nd Conference on
Severe Local Storms
12
When do derechos occur?

• 270 events from 1980-2001

70 occur in May-Aug (Warm Season)
Coniglio and Stensrud (2004) WAF
13
Where do derechos occur?

• 171 events, all months, 1986-2001

Coniglio and Stensrud (2004) WAF
14
Where do derechos occur?

• 113 events, May-Aug, 1986-2001

Coniglio and Stensrud (2004) WAF
15
Where do derechos occur?

• 58 events, Sep-Apr, 1986-2001

Coniglio and Stensrud (2004) WAF
16
Can we forecast them?

• Short term (0-4 h) forecasts
• Aided greatly by WSR-88D Radar Network
• Bow echoes and persistent embedded circulations
• Longer term forecasts (6-24 h)?
• Not there yet
• Continued poor performance of operational models
  in predicting warm-season precip (see June 2004
  BAMS)
• Timing and location of convective initiation
• ETA model with parameterized convection has
  difficulty with realistic MCS propagation

Bukovsky et al. (2004) AMS 22nd Conference on
Severe Local Storms
17
Can we forecast them?

• Cloud-resolving forecast models?
• Develops convective motions on the grid scale
• Weather Research and Forecasting (WRF) model
  holds promise
• 4-km resolution 36-h WRF simulations initialized
  at 00Z run during BAMEX project
• often develops bowed-MCS structures and does
  reasonably well at 2-6 h lead times
  (www.joss.ucar.edu/bamex/meetings/weisman_wrf/),
  but
• Predictability and reliability at longer lead
  times (gt 6 h) is poor
• Ensemble approach?

18
Can we forecast them?

• Pattern recognition of synoptic-scale
  environments Warm season vs. dynamic
• Zonal Pattern and many hybrids exist near
  continuum of flow-pattern types

Warm Season Pattern
Dynamic Pattern
Johns (1993) WAF
Coniglio et al. (2004) WAF
19
Can we forecast them?

• Ingredients-based techniques
• Used for predicting convective mode and intensity
• Broad similarity between derecho and supercell
  environments
• Morphology of initiating process and the
  distribution of initial convective cells may be
  important
• Discrete vs. linear propagation
• Growth of the convectively generated cold pool
• Renders pre-convective forecasts of derechos
  difficult

Doswell and Evans (2003) Atms. Res.
20
Can we forecast them?

• Ingredients-based techniques lead to parameter
  clues for MCS strength and longevity
• Persistence of external forcing
• Depth and stability of low-level inflow
• Strength and distribution of vertical wind shear
• Strength and distribution of cold thunderstorm
  outflow
• What are the physical connections between the
  environments and MCS strength and longevity
• Idealized numerical simulations useful

Evans and Doswell (2001) WAF
Coniglio et al. (2004) WAF
21
Model Environments
X

• Organized bow echoes like moderate to strong
  low-level shear in idealized numerical
  simulations
• Likelihood of bow echoes as depth of
  shear-layer
• Organized bow echoes do not occur if the shear
  layer extends above 5 km (Weisman and Rotunno
  2004 JAS)

X
Modified from Weisman (1993) JAS
22
Physical Connection?

• Based on vorticity balance between cold pool and
  low-level vertical wind shear (RKW Theory)
• Upright convection (deepest lifting) favored when
  circulations balance
• RIJ forms between sloping updraft and cold pool
• Vorticity associated with elevated RIJs help to
  prop-up the updrafts to promote further
  vigorous convection and mesoconvective
  organization
• Process is greatly favored with strong low-level
  shear and weak shear above the cold pool

Weisman (1993) JAS
Weisman and Rotunno (2004) JAS
23
What about Observations?
91 derechos sampled during initial or early
mature stages
Coniglio et al. (2004) WAF
24
What about Observations?
91 derechos sampled during initial or early
mature stages
X mean
Coniglio et al. (2004) WAF
25
What about Observations?
0000 UTC 28 MAY 2001 OUN
CAPE 3800-4800 J/kg CIN 0 J/kg LI
-10 0-2.5-km shear 11 m/s 5-10-km shear 30 m/s
see Miller et al. (2002) AMS 21st Conference on
Severe Local Storms
26
Models vs. Reality

• Cooler-season events often do have strong
  low-level wind shear and weaker shear aloft, but
• Conundrum Idealized numerical models probably
  most relevant to warm-season environments
• Many warm-season cases have relatively weak
  low-level shear and significant shear above the
  cold pool
• RKW mechanism (cold pool/low-level shear balance)
  may not be the primary controlling mechanism
• Alternative Ideas?
• Is there some basic dynamical importance to the
  existence of deep unidirectional vertical wind
  shear above the cold pool?

Coniglio et al. (2004) WAF
27
Alternative Ideas?

• Guided by the observations, what is the role of
  the upper-level shear in idealized simulations on
• the basic 2D lifting of environmental air above a
  cold pool
• the 3D evolution of simulated convective systems?

28
2D Simulations

• Use NCOMMAS (Wicker and Wilhelmson 1995) to
  produce a set of dry 2D density current runs in
  neutral stability
• Use DxDz250 m, x240 km, z16 km
• Introduce a cold pool through a cooling
  function
• Run out to 1.5 h
• Use trajectories to calculate vertical parcel
  displacements after 0.5 h
• 7 runs 0 to 30 m s-1 of 5-10 km shear in 5 m s-1
  increments

20 m s-1 over 0-5 km
12 m s-1 over 0-2.5 km
Cold pool motions
29
Results Maximum vertical velocity
Strongest upward motion occurs for case with
no upper-level shear
30
Results Maximum low-level (0-2 km) parcel
displacements
Lifting is enhanced for weak to
moderate upper-level shear
31
Results 2D, no upper-level shear
32
Results 2D, 10 m s-1 upper-level shear
33
Conclusion 2D simulations

• Upper-level shear (above moderate low-level
  shear) increases vertical parcel displacements
  through an overturning updraft (Moncrieff 1981),
  despite lower w along interface (Shapiro 1992,
  Moncrieff and Liu 1999)
• Shear entirely above cold pool can accomplish
  this
• What happens in 3-D?

34
Results 3D simulations, 2 - 4 hSurface
precipitation mixing ratio, gust front, winds
0 m s-1 5-10 km shear
15 m s-1 5-10 km shear
35
Results 3D simulationsstorm-relative wind
profiles at 3.5 h
36
Results 3-D simulationsTrajectories 3- 4 h
0 m s-1 shear
15 m s-1 shear
30 m s-1 shear
37
Observations108 derechos sampled in
weak/moderate forcing
29
35
44
Large differences in 5-10 km shear between
beg./mature decay
Note absence of critical Level for decay
soundings
Coniglio et al. (2004) WAF
38
Applications

• Can provide nowcast for demise of strong, linear
  MCSs lifting of critical layer
• Requires forecast of cold pool motion (Corfidi
  2003 WAF) storm-relative wind profile
• Probabilistic forecasts of linear MCSs
  structure/longevity at longer lead times?
• Forecast experiment to take place in Summer
  2005 at Storm Prediction Center
• May help explain how strong MCSs can persist
  after dark in situations not forced by a LLJ
• Overturning of residual-layer parcels may not
  care about nocturnal inversion (as long as cold
  pool is replenished)

39
Final Thoughts

• While low-level shear is important, upper-level
  shear also is important - existence and location
  of critical layer with upper-level shear
  important component of MCS structure and
  maintenance
• Systems are maintained for longer periods in 3D
  and are substantially larger (not shown)
• Observations suggests importance of upper-level
  shear
• Role of 3D convective structures on maintenance
  unresolved

40
THANKS FOR COMING!