The Role of Boundary Layer Circulations in Convection Initiation

1
The Role of Boundary Layer Circulations in
Convection Initiation
Michael S. Buban
11/05/09

• Cooperative Institute for Mesoscale
  Meteorological Studies (CIMMS)
• School of Meteorology, University of Oklahoma,
  Norman OK
• NOAA/National Severe Storms Laboratory, Norman OK

2
Importance of CI forecasting

• Thunderstorms produce a variety of high-impact
  weather (e.g. tornadoes, hail, high wind,
  flooding rain, lightning).

• A great deal of time and money has been
  dedicated to improving forecasts of thunderstorms
  and their attendant severe weather threats.

• A primary forecasting challenge is to determine
  where, when, or whether or not convection will
  develop.

• During the late spring and early summer of 2002
  the International H20 project (IHOP) was
  conducted to document the 3-D evolution of water
  vapor in the atmosphere with the hopes of
  improving CI and quantitative precipitation
  forecasts.

3
Background Parcel Theory

• Air is viewed as a parcel which moves
  according to several rules

1) No air is exchanged through the parcel
boundaries (no mixing).
2) The potential temperature and water vapor
mixing ratio remain constant during dry adiabatic
motions.
3) The temperature decreases at the dry
adiabatic lapse rate (9.8 K/km) during
unsaturated ascent.
4) The temperature decreases at the moist
adiabatic lapse rate (as a function of T, p, qv)
during saturated ascent.
4
Background Soundings

• The initially unsaturated air rises dry
  adiabatically (red) from near the surface to the
  LCL.

• During dry adiabatic ascent the temperature
  decreases and RH increases.

• Above the LCL air rises moist adiabatically
  (blue).

Adapted from Weiss and Bluestein (2002)

• To reach the LFC air must rise above the
  capping inversion (CIN)

• Above the LFC the air parcels, now warmer than
  their environment, will continue to rise driven
  by buoyancy forces, realizing CAPE.

5
Parcel theory - shortcomings

• Mixing due to both coherent turbulence and
  viscous effects is neglected between the parcel
  and the environment during transport.

• Mixing can substantially alter the scalar (I.e.
  T,qv,qc,etc.) and momentum properties of a
  parcel, especially in regions where strong
  gradients of these fields exist.

• Sounding analysis of CAPE, CIN, etc. is
  1-Dimensional, the real atmosphere is
  4-Dimensional.

• In the vertical, air within an updraft is a
  mixture from many different levels

• The atmosphere is a continuous fluid that obeys
  highly non-linear equations - initial
  uncertainties amplify in time (chaos) and changes
  in the state of the system further change the
  state of the system (feedbacks).

6
CI Getting air to the LFC

• To initiate deep moist convection a quantity of
  air must reach its LCL and LFC. How does this
  occur?

• Near surface air can be heated
  (insolation/conduction) becoming less dense than
  its surroundings and (assuming minimal cap)
  rising to its LCL/LFC due to buoyancy alone.
  Does this ever happen?

• Air can be forced to its LCL/LFC by a vertical
  pressure gradient.

• Vertical pressure forces may develop due to a
  variety of boundary layer features and exist
  across many scales.

7
CI Fronts

• Fronts are defined as the transition zone
  between two air masses with different densities.
  Typically the density difference is primarily due
  to temperature.

• Frontal zones may extend hundreds of kilometers
  in the along-line direction and contain density
  gradients only several km wide across the front.

• Density differences create solenoidally forced
  circulations within the frontal zone, with
  enhanced low-level convergence and vertical
  motion.

• The strength of the frontal circulation is
  modulated by the strength of the density gradient
  and the local LFC relies on the thermodynamic
  state of the air within the frontal zone.

8
CI Drylines

• The dryline is a moisture gradient separating
  warm, moist air flowing off the Gulf of Mexico
  from the hot, dry air originating over the SW US.

• Extending as much as several hundred km in the
  N-S direction, across-dryline gradients may be
  only 1 km or less.

From Ziegler and Rasmussen (1998)

• Like fronts, drylines frequently develop
  solenoidally forced circulations, with low-level
  convergence and enhanced vertical motion.

• Density gradients are typically weaker than for
  fronts and are due to a combination of small
  temperature differences and larger moisture
  differences between air masses.

9
CI Non-frontal boundaries

• Other non-frontal near-surface boundaries can
  also force air to the LFC. Some examples

• Sea/lake breeze boundaries owing to
  differential heating over land and water.

• Outflow boundaries from previous precipitation.
  Density gradients occur due to the temperature
  difference between the rain cooled and ambient
  air.

• Boundaries due to topographical effects,
  differences in vegetation, differences in soil
  moisture, etc.

• Although caused by different effects,
  boundaries are dynamically similar, featuring
  low-level convergence and enhanced vertical
  motion - to varying degrees.

10
CI or no CI

• The strength of the density gradient across the
  boundary controls the intensity of the updraft.

• The local thermodynamic state and source region
  of air within the updraft will affect the parcel
  LFC height.

• Parcels must reach the LCL/LFC prior to leaving
  the mesoscale updraft (Ziegler and Rasmussen
  1998).

From Weckwerth and Parsons (2006)

• Boundary orientation and interaction with the
  ambient vertical shear can produce updrafts more
  or less favorable for CI.

• CI is dependent on the amount of CIN to
  overcome, the effects of updraft entrainment, and
  other along-line variability.

11
CI BL convection

• Buoyancy driven circulations may arise in the
  CBL allowing parcels to reach their LCL/LFC

• The linearized equations of motion under certain
  conditions admit solutions in several forms (e.g.
  rectangular or hexagonal open convective cells,
  horizontal convective rolls, etc.)

From Brown (1980)

• Most frequently the CBL exhibits combinations of
  several modes of convection.

• Although updrafts with this dry convection may
  reach values similar to those along strong
  boundaries, they generally do not penetrate far
  above the CBL top and generally initiate
  convection only in weakly capped environments.

12
CI boundary interactions

• The intersection of two boundaries (triple
  point) has been shown to be a preferential
  region for CI.

• The circulation from one boundary may be lifted
  over the intersecting boundary.

From Weiss and Bluestein (2002)

• Enhanced vertical motion may result when the
  updrafts of the two boundaries are in phase.

• Intersections may be between any two
  boundaries. For example cold front/dryline,
  dryline/outflow boundary, front/HCR, etc.

13
CI Misocyclones

• In recent years some attention has been focused
  on the presence of small-scale (1-3 km) vortices
  or misocyclones which have been observed to
  propagate along boundaries.

• Originally their importance seemed to be mainly
  in relation to non-supercell tornadogenesis, but
  recent studies have shown them to play a
  potentially important role in CI.

• Especially when boundary secondary circulations
  are weaker, interactions of misocyclones with the
  boundary may provide enhanced regions of
  frontogenesis and vertical motion.

• Misocyclones are coherent structures, tend to
  move with the mean BL flow, and can last for
  substantial periods of time (30 min. or more),
  increasing their effect on local air parcels.

14
Misocyclone observations
Marquis et al. (2007) - 3 June 2002
Pietrycha and Rasmussen (2004) - 10 June 1999

• Misocyclones were observed along the dryline in
  west Texas by analyzing mobile mesonet transects.

• Misocyclones were observed in multiple-Doppler
  radar analyses along a cold front in the eastern
  Oklahoma panhandle.

15
Misocyclone observations - 2
Marquis et al. (2007) - 19 June 2002
Arnott et al. (2006) - 10 June 2002

• Misocyclones were observed along a cold front
  and a dryline in western Kansas using
  multiple-Doppler radar data.

• Misocyclones in different cases behave
  similaraly as they track along the boundaries.

16
Dryline misocyclones - airflow
W (ms-1)

• A wave moves northward along the DL with the
  misocyclone remaining located near its inflection
  point.

• The misocyclone location also corresponds to
  the intersection of the DL with a longitudinal
  HCR extending upstream into the dry air.

• Both the DL and the vortex show surprisingly
  little evolution over 9 minutes.

• The misocyclone has a maximum vertical
  vorticity on the order of 9 x 10-3 s-1

Buban et al. (2007) - 22 May 2002
17
Simulations - Motivation

• Detailed analyses of boundary layer features
  were made possible using high-resolution
  observations obtained during the IHOP field
  project.

• These analyses however, are still lacking the
  spatial and temporal resolution to fully
  understand the dynamics governing the initiation
  and evolution of small-scale circulations such as
  misocyclones.

• Using a numerical model allows for the
  simulation of BL features on very fine scales in
  space and time.

• Simulations give a complete and internally
  consistent data set across many scales from which
  features of interest can be analyzed.

18
Model Description

• The nonhydrostatic, cloud resolving
  Collaborative Model for Multiscale Atmospheric
  Simulation (COMMAS) was used to simulate the
  dryline and surrounding environment on 22 May
  2002.

• The model employs 3rd order advection and a 1.5
  order subgrid-scale turbulence parameterization.
  Surface fluxes were calculated with a modified
  version of the Deardorff (1978) force-restore
  land surface/atmosphere exchange model.

• Inflow boundary conditions were obtained by
  spatially and temporally interpolating the
  9-minute spaced Lagrangian analyses (for
  thermodynamic variables) and 3-minute spaced
  multiple-Doppler analyses (for u and v momentum).

• The simulation was run on a 30 km x 30 km x 6km
  grid with a 200 m horizontal grid spacing and
  stretched vertical grid spacing using a 15 m deep
  lowest layer.

19
Simulated horizontal qv
2300
2309

• Overall larger scale structure such as dryline
  orientation and moisture gradient similar to the
  observational analyses.

• BL west of the dryline develops structures such
  as HCRs and open convective cells.

2318
2327

• Simulation develops smaller scale structures
  not resolvable in the observational analyses.

• Misocyclones develop and propagate northward
  along the dryline.

20
Simulated qv cross-sections
2300
2320

• Simulation produces strong secondary
  circulation along the dryline and roll
  circulations west of the dryline, as seen in the
  radar analysis wind fields.

• Smaller-scale plumes of moisture both east and
  west of the dryline develop as inclusion of a
  radiation and surface flux scheme allow
  instability near the surface to arise.

• Magnitudes of the sensible and latent heat
  fluxes in the model are consistent with those
  measured by flux sensors in the domain.

21
Simulated qv - qc surfaces

• Deeper moisture plumes produce clouds along and
  east of the dryline.

• West of the dryline higher based cumulus
  develop along convergence bands.

• Moisture fields are modified due to the
  presence of misocyclones along the dryline.

• Forcing the model with radar analysis winds at
  the inflow boundaries produces similar momentum
  structures in the interior as seen in the
  individual radar analyses.

22
Simulated DL misocyclones
2303
2306

• Perturbations in the radar wind analyses
  introduced at the lateral inflow boundaries
  amplify downstream and produce misocyclones.

• The misocyclones produced in the model are
  similar kinematically to those in the radar
  analyses but with smaller-scale structure.

2309
2312

• Misocyclones acting on the moisture fields
  produces structures not resolvable in the
  observational analyses and maximum vertical
  vorticity values are 2-3 times greater than in
  the radar analyses.

23
Simulated vortices - surfaces

• Misocyclones act on the moisture field
  producing local perturbations.

• North of the misocyclone convergence deepens
  the moist layer resulting in enhanced probability
  of air reaching the LCL.

2321
2324

• South of the misocyclone moisture is scoured
  from the west, locally shifting the dryline
  eastward.

2327
2330
24
Misocyclone - conceptual model
From Buban et al. (2007)

• Friction induced horizontal vorticity at very
  low levels may also be tilted and stretched.

• The dryline secondary circulation may also
  provide a source of vertical vorticity if tilted
  into the vertical by the updraft.

• Horizontal vorticity associated with an HCR
  crossing the DL is tilted and stretched.

25
Summary/Conclusions

• For CI to occur, air must rise to the LCL and
  LFC

• Air generally doesnt get to the LCL/LFC
  through buoyancy alone, so some other forcing
  mechanism must be present for CI to occur

• Many studies have found near-surface boundaries
  provided enhanced convergence and vertical motion
  to allow parcels to reach their LFC.

• Misocyclones have become a subject of interest
  recently for their potential role in CI.

• Observations and numerical simulations have
  shown a tendency for increased convergence and a
  deepening moist layer north of misocyclones,
  resulting in a greater likelihood of air reaching
  the LCL and possibly greater probability for CI.