mcgill freezing rain

1
Physical Processes Synoptic Patterns
Accompanying Cold-Air Damming Erosion
Wendy Stanton Gary Lackmann Mike Brennan North
Carolina State University
NWS CSTAR Co-Lab, 15 October 2002
RUC analysis sfc obs for 12 UTC 10 October 2002
GFS 48-h Fcst
Eta 48-h Fcst
1
2
Outline

• I. Briefly review earlier research on CAD types
• A. Maximum temperature impacts
• B. Precipitating and dry CAD
• II. Recent CAD erosion research results
• A. Review of CAD erosion mechanisms
• B. Synoptic erosion scenarios
• C. Case study Model CAD erosion versus
  observations

2
3
I. Earlier CSTAR CAD Research

• One of original C challenges Forecasting
  issues related to Cold-Air Damming (CAD)
• Earlier CAD Research (C. Bailey)
• CAD detection algorithm (captured full CAD
  spectrum)
• Developed CAD sub-type composites
• see http//www4.ncsu.edu/nwsfo/
• Stratification of CAD by sensible weather impacts
  
• Maximum temperature in damming region
• Precipitating vs. dry

3
4
A. High-Impact Vs. Low-Impact CAD

• Stratified classical CAD cases by impact on Tmax
• One composite contains cases with Tmax 15F or
  more below climatology at GSO (high impact)
• The other cases where Tmax was 2F or less below
  climo, or above climo at GSO (low impact)
• Can you tell which is which?

5
5
CAD is Not a Monolithic Phenomenon!
Sea level pressure, anomaly
HIGH-IMPACT CAD
LOW-IMPACT CAD
6
6
B. Precipitating Vs. Dry CAD

• Original CAD sub-types identified in 10-yr
  climatological sample
• Major distinction dry onset versus
  precipitating CAD

4
7
CAD Sub-Type Summary

• Sea-level pressure does not tell the whole CAD
  story!
• Key upper-air features
• Upper jet dominates precipitating high-impact
  cases
• Ridge at 500 mb west of CAD region in dry onset
  composite
• 850-mb ridge axis over CAD region in dry onset
  composite
• E-W high elongation in precipitating, high-impact
  cases
• N-S elongation in dry onset weak impact cases
• For details, see Bailey et al. 2002 paper
• (submitted to WAF over the summer, in review)

7
8
II. CAD Erosion Research

• Specific NWS CSTAR challenge CAD erosion
• Model guidance Usually pre-mature erosion of
  CAD cold dome. Why?
• Much work needed to understand CAD erosion
• What physical processes are active?
• Do different physical processes act in different
  synoptic settings?
• Which processes are most problematic for NWP
  models?
• How can model representations be improved?
• Which situations can forecasters expect to be
  most problematic?

8
9
II. A. Candidate CAD Erosion Processes

• 1.) Heating from bottom up (surface warming)
• Cold dome mixes out due to surface heating
• Solar radiation
• Upward soil heat flux (cold air over warm ground)
• Clear-air, warm season CAD often ends in this
  manner

Red before Blue after
9
10
Candidate CAD Erosion Processes

• 2.) Turbulent mixing at cold dome top (inversion)
• Strength of mixing depends on strength of shear,
  inversion (Ri)
• Strong shear/weak inversion (small Ri) mixing
  favored
• Result Top-down CAD erosion

Fort Meade, MD
10
11
Candidate CAD Erosion Processes

• 3.) Divergence near surface
• Pressure falls to north, cyclone to NW of CAD
  region
• Stronger PGF, winds in northern CAD region
• Depth of cold dome decreases via continuity
  (Lackmann Overland 1989)

Cross-sectional view
Inversion layer
UP
11
12
Candidate CAD Erosion Processes

• 4.) Inland Coastal Front Movement
• Boundary of cold dome (CF) moves inland
• May accompany pressure falls to N, surface
  divergence
• Usually only affects eastern areas of damming
  region

12
13
Candidate CAD Erosion Processes

• 5.) Downslope Scouring Two Scenarios
• a.) Cold-Frontal Passage b.) Coastal Cyclone
• (as in Bell Bosart 1988)

L
L
13
14
B. CAD Erosion Composites

• CAD Demise Composites Surrogate for dominant
  physical processes in CAD erosion
• Compositing Strategy
• Used classical CAD cases, 1984-1995 (from Bailey
  algorithm)
• 90 total cases
• Examined patterns in MSLP and surface ?e
• Noted locations of hi/low centers fronts at
  erosion times
• Classified each case according to synoptic
  pattern
• 1.) Cold Front Passage (14 cases)
• 2.) Residual Cold Pool (23 cases)
• 3.) Coastal Low (25 cases)
• 4.) NW Low (23 cases)

14
15
(No Transcript)
16
1.) Cold-Frontal Passage

• Dominant Erosion Processes
• Downslope northwesterly flow, subsidence,
  adiabatic warming
• Model Performance
• Adequate, provided accurate depiction of cold
  front passage
• Eta often too fast with cold FROPA in North
  Carolina
• Features
• Cold front arrives in damming region within 6 hrs
  of CAD demise

16
17
(No Transcript)
18
2.) Residual Cold Pool

• Dominant Erosion Processes
• Solar heating within cold dome
• CAD airmass doesnt always go away when Baker
  ridge does
• Model Performance
• Hinges on model ability to properly depict clouds
  (problematic)
• Largest error near surface
• Features
• Surface high ill-defined and baggy
• No synoptic systems influencing erosion

18
19
(No Transcript)
20
3.) Coastal Low

• Dominant Erosion Processes
• Downslope flow, adiabatic warming
• Divergence, cold dome air advected toward coast
• Similar to case documented by Bell and Bosart
  1988
• Model Performance
• Adequate, provided accurate forecast of coastal
  cyclone!
• Features
• Damming region remains in cold air during erosion

20
21
(No Transcript)
22
4.) Northwestern Low

• Dominant Erosion Processes
• Solar heating surface-based mixing within cold
  dome
• Mixing and entrainment at inversion
• Divergence, Inland coastal front movement
• Features
• Flow in damming region becomes southerly,
  pressure falls to N
• CAD cold dome does not disappear just because
  wind shifts!!!
• Model Performance
• POOR, difficulty representing multiple physical
  processes
• Radiation, cloud, mixing problems

22
23
Summary of Erosion Scenario Composites

• Erosion scenario independent of CAD type
• Synoptic patterns Link to dominant erosion
  process
• Multiple processes may contribute to demise
• NW Low scenario most difficulty for NWP due to
  nature of physical processes

23
24
Summary of Erosion Scenario Composites

• NW Low and Residual Cold Pool slowest erosion
  scenarios

39 h
30 h
26 h
20 h
24
25
C. Case Study Model Experiments

• Elucidate physical process(es) of CAD erosion in
  representative case study(ies)
• Select two recent CAD cases in which the Eta
  model exhibited large errors
• Perform detailed observational analyses
• Pinpoint cause of operational model failure (with
  representation of CAD erosion)

25
26
L

• 10-14 December 2001
• Onset 12/10 00Z,
• Erosion began 12/13 03Z,
• Demise 12/14 15Z
• Unusual, multi-phase event,
• NW Low erosion scenario, erosion prior to cold
  frontal passage
• Poorly handled erosion by Eta, esp. in SW damming
  region

L
27
Case Study Eta SLP Forecast
Analysis
Eta 24-h Fcst
Eta SLP Analysis, 00Z 13 December 2001
Eta 24-h SLP forecast, valid 00Z 13 December 2001
27
28

• Bias Forecast - Analysis
• Forecast 330 K
• Analysis 318 K
• Eta ?e 12 K too high!
• Forecast stronger gradient on eastern edge of
  cold wedge
• Damming region can be seen in error pattern

27
29
Eta 12 hr Forecast Vs. Observed
Greensboro, NC (GSO)
Observed Red
Model Blue
12/13 at 00Z
30
Case Study Conclusions

• Observations cold dome eroded from top-down
• Model cold dome eroded from bottom-up
• Model surface heating too strong
• Too much sun gets through Eta clouds? (NCEP)
• Betts et al., 1997 and Yucel et al. 1998
• Upward heat flux in soil?
• Model sensitivity experiments underway
• Examine cloud/radiation issue (transmissivity,
  path length)
• Upper-air moisture profiles, cloud fraction
  versus satellite

31
Thanks to

• NOAA/NWS CSTAR program (HQ, ER)
• NWS CSTAR offices (esp. Kermit Keeter, Scott
  Sharp, Gail Hartfield, Neil Stuart, Larry Lee)
  and many others
• Jeff Waldstreicher (Profiler and RASS images,
  encouragement)
• NCSU Al Riordan, Chris Bailey (now at HPC), Lian
  Xie
• NCEP Brad Ferrier
• All of you for your comments and constructive
  criticism!

31
32
RASS data
30