Detailed Case Study of a Carolina Coastal Front

1
Detailed Case Study of a Carolina Coastal Front

• Keith K. Contre

2
Goal of Case Studies

• understand physical processes
• coastal front movement
• CAD erosion

• through
• detailed data analysis
• performance evaluation of operational Eta
• MM5 model sensitivity tests

3
Case Study Selection

• onshore case
• good data
• sensible weather change w/ passage

• 1) onset synoptic pattern similar to composite
• typical onshore coastal front

• 2) onset synoptic pattern unlike composite
• not all fronts typical (comparison)

4
Case 1 15 Dec 2000, 12 UTC

• onset pattern different from composite
• poor real-time model performance
• NWS forecasters cited as problematic

5
MSLP and 500-mb Height at Onset
Onshore Composite
Eta Analysis Valid 15 Dec 2000, 12 UTC

• sfc trough in NW Gulf of Mexico
• southeasterly geostrophic winds along
  Carolina coast
• stronger 500-mb trough over eastern Great plains

• sfc low in NW Great Plains
• northeasterly geostrophic winds along Carolina
  coast
• weaker 500-mb trough over western Great plains

6
Inland Movement Too Soon in Model
Eta 30-h Forecast Valid 16 Dec 2000, 18 UTC
Manual Analysis Valid 16 Dec 2000, 18 UTC
Wilmington, NC (ILM)

• temp near 22 ºC
• wind southerly 10 m/s
• partly cloudy skies

• temp 12 ºC
• wind calm
• fog

7
Model Temps Behind Front Too Warm
Eta 30-h Forecast Valid 17 Dec 2000, 00 UTC
Manual Analysis Valid 17 Dec 2000, 00 UTC
Wilmington, NC (ILM)

• temp near 22 ºC
• wind southerly 12 m/s
• partly cloudy skies

• temp 14 ºC
• wind SE 2.5 m/s
• fog

8
Synoptic Overview
Eta Analysis Valid 15 Dec 2000, 18 UTC
Eta Analysis Valid 16 Dec 2000, 00 UTC
Eta Analysis Valid 16 Dec 2000, 06 UTC
Eta Analysis Valid 16 Dec 2000, 12 UTC
9
Synoptic Overview
Eta Analysis Valid 16 Dec 2000, 18 UTC
Eta Analysis Valid 17 Dec 2000, 00 UTC
Eta Analysis Valid 17 Dec 2000, 06 UTC
Eta Analysis Valid 17 Dec 2000, 12 UTC
10
Mesoscale Overview
Manual Analysis Valid 15 Dec 2000, 12 UTC
Manual Analysis Valid 15 Dec 2000, 18 UTC
Manual Analysis Valid 16 Dec 2000, 00 UTC
Manual Analysis Valid 16 Dec 2000, 06 UTC
11
Mesoscale Overview
Manual Analysis Valid 16 Dec 2000, 12 UTC
Manual Analysis Valid 16 Dec 2000, 18 UTC
Manual Analysis Valid 17 Dec 2000, 00 UTC
Manual Analysis Valid 17 Dec 2000, 06 UTC
12
Analysis Work In Progress
Pattern and Cause of CAD erosion?

• synoptic CAD erosion setting NW low
• CAD erosion process inland CF

What physical processes caused CAD to weaken and
CF move inland?

• CAD weakened by cold adv aloft no warm adv with
  CF
• sfc heating on cold side of CF not
  likely night
• Mixing (shear) across CF boundary aloft not
  likely Ri too large
• cold air div due to pressure falls to N most
  likely high moves NE

Why did Eta move front inland too quickly (day
vs. that night)?

• possibly too much solar radiation

13
MM5 SST Resolution Sensitivity Test

• Hypothesis
• Higher res SST
• better SST gradient at coast and Gulf Stream
• better sfc sensible heat fluxes
• better post-frontal air temps
• better simulation of shallow boundary
• Perform 3 simulations varying only SST res
• - 1, 1/2, and 1/8º (obs verified improving
  accuracy)
• Kain-Fritsch cumulus parameterization
• MRF PBL
• Goddard Microphysics
• Cloud Radiation
• OSU LSM

14
Model Domain

• 12 km horizontal resolution and 37 vertical
  levels (16 below 850 mb)
• initial conditions and lateral boundary
  conditions from Eta-212

15
SST Comparison
1º SST
1/8º SST
1/8º SST 1º SST
16
Results
36-h Forecast Valid 17 Dec 2000, 00 UTC 1/8º SST
36-h Forecast Valid 17 Dec 2000, 00 UTC 1º SST
Improvements with 1/8º SST Data

• better SST gradient at coast and Gulf Stream
• better sfc sensible heat fluxes
• better post-frontal air temps
• better simulation of shallow boundary

Manual Analysis Valid 17 Dec 2000, 00 UTC
17
Results may have been even better
if SST field off SC were
closer to obs.
CHS
18
MM5 LSM Sensitivity Test

• same model set-up
• 1/8º SST
• Run 1 OSU LSM on
• Run 2 OSU LSM off

Results were not significantly different!
19
MM5 PBL Sensitivity Test

• same model set-up
• 1/8º SST
• OSU-LSM turned off
• Run 1 MRF PBL
• Run 2 Blackadar PBL

20
Results
30-h Forecast Valid 16 Dec 2000, 18 UTC MRF PBL
30-h Forecast Valid 16 Dec 2000, 18 UTC Blackadar
PBL
Improvements with Blackadar Scheme

• better trough/ridge amplitude
• better frontal temp gradient
• better temps in CAD region
• better coastal front location

Manual Analysis Valid 16 Dec 2000, 18 UTC
21
2-m Air Temp Difference ºC (Blackadar MRF)
Why such a large air temp difference?

• flux calculation methods nearly identical
• non-linearity of processes may play small role
• probable cause different mixing methods

22
Acknowledgements

• Dr. Riordan
• Dr. Lackmann and Dr. Xie
• Mike Brennan and Students in the Forecasting Lab
• NWSFOs esp. RAH, ILM and MHX
• NOAA CSTAR Grant NA-07WA0206