COAMPS Ducting Validation Wallops2000

1
COAMPS Ducting ValidationWallops-2000
William Thompson and Tracy Haack Naval Research
Laboratory Marine Meteorology Division Monterey,
CA
COAMPS is a registered trademark of the Naval
Research Laboratory
2

• Wallops-2000 Experiment
• April/May 2000

• Wallops-2000
• Experiment
• April/May 2000
• NSWCDD Microwave Propagation
• Measurement System II (MPMS)
• NSWCDD Sealion project boat
• JHU/APL Chessie Boat
• NPS Flux buoy (13 km shore)
• JHU/APL Helicopter
• SCSC resident AN/Spy-1A radar
• NASA Space and Range radar

3

• COAMPS Reanalysis
• Wallops-2000 Experiment
• April/May 2000

Nested domains

• Wallops 2000 Reanalysis
• April/May 2000 12-hr continuous
• forecasts
• 36, 12, 4-km resolution
• 71 levels w/ 20 levels below 1.5 km
• Sea Surface temperature analysis
• preformed at grid resolution
• Hourly 2D and 3D forecast fields

4 km
12 km
36 km

• COAMPS
• Non-hydrostatic, compressible equations
• Physical parameterizations include
• surface fluxes
• radiation
• turbulence
• microphysics
• Data assimilation incremental update
• Diagnostic computation of refractivity
• and ducting layers

282

NPS Buoy

NPS Buoy
290
286
294
4-km Grid
4
Modified Refractivity (M) andDucting Properties

• M 77.6/T P 4810 e/T z/Re x 1x106
• where
• T is temperature in K
• P is pressure in mb
• e is vapor pressure in mb
• z is elevation
• Re is the radius of the earth
• A duct exists when M decreases with elevation
• The strength of the duct is the difference
  between the relative minimum in M (Mm) and the
  relative maximum in M
• The base of the duct is the elevation below Mm at
  which the value of M is equal to Mm
• The thickness of the duct is the difference in
  elevation between Mm and the base of the duct

5
Modified Refractivity (M) and Ducting Properties
Z
Duct Strength
M
6
Animation of Duct Properties

• Hourly images of forecast duct strength, duct
  thickness, and duct base height
• Shading intervals
• Duct strength 2 M units
• Duct thickness 20 m
• Duct base height 25 m
• Forecast from 0000 UTC 1 May to 0000 UTC 2 May
  2000

7
Duct Strength
8
Duct Thickness
9
Duct Base Height
10
Forecast Cross Section of Potential Temperature
(K, contours) and Mixing Ratio (g/Kg, shaded)
valid 0600 UTC 1 May 2000
11
Forecast Cross Section of Modified Index of
Refraction Valid 0600 UTC 1 May 2000
12
Forecast Cross Section of Potential Temperature
(K, contours) and Mixing Ratio (g/Kg, shaded)
valid 2300 UTC 1 May 2000
13
Forecast Cross Section of Modified Index of
Refraction Valid 2300 UTC 1 May 2000
14
Duct Property Means and Time Series At NPS Buoy
Base Ht Thickness Strength
Frequency NPS 30.2 94.3 7.7 62
30km 27.0 107.8 8.3 58 60km
16.9 121.4 9.2 54 100km 25.8
128.7 9.8 58
Duct Thickness
Duct Strength
Duct Base
Frequency
15
Time Series at 30 km
Duct Thickness
Duct Strength
Duct Base
Frequency
16
JHU APL Helicopter Profiles
1 May 2000 1500 UTC 37.8o N, 75.4o W
Specific Humidity
Modified Refractivity
17
Ducting Statistics
MODEL Ducting STATS Duct Frequency ()
Num CMPMEAN CMPSTD 190
0.34 0.48 Duct Base
Height (m) Num CMPMEAN
CMPSTD 65 3.15 14.40
Duct Strength (M-units) Num
CMPMEAN CMPSTD 65 8.74
8.84 Duct Thickness (m)
Num CMPMEAN CMPSTD 65
77.18 45.38
OBS Ducting STATS Duct Frequency ()
Num OBSMEAN OBSSTD 190
65 48 Duct Base
Height (m) Num OBSMEAN
OBSSTD 123 1.40 4.88
Duct Strength (M-units) Num
OBSMEAN OBSSTD 123 6.99
6.53 Duct Thickness (m)
Num OBSMEAN OBSSTD 123
71.91 27.52
18
Observed vs Model Ducting Statistics
---------------------------------------------
--- Ducting Contingency Table
------------------------------------------------
Model Ducting Obs Ducting
Duct No Duct Total Duct 54
69 123 No Duct 11
56 67 Total 65
125 190 --------------------------------
---------------- Event freq 0.6473684
Percent correct 0.5789474 Mean Squ
Error 0.4210526 Hit Rate
0.4390244 Miss Rate 0.5609756
False alrm Rate 0.1641791 Corct Null
Rate 0.8358209 Discrim Score 0.2748453

Duct Frequency () Num
CMPMN OBSMN CMPSTD OBSSTD BIAS RMSE
190 0.34 0.65 0.48
0.48 0.31 0.65 Duct
Base Height (m) Num CMPMN
OBSMN CMPSTD OBSSTD BIAS RMSE 54
1.17 1.61 4.47
5.15 0.44 4.60 Duct
Strength (M-units) Num CMPMN
OBSMN CMPSTD OBSSTD BIAS RMSE 54
9.77 7.65 9.32
7.46 -2.11 10.37 Duct
Thickness (m) Num CMPMN
OBSMN CMPSTD OBSSTD BIAS RMSE 54
75.37 70.89 46.53 33.19
-4.48 49.20
both obs and model have duct within depth of obs
profile
19
Meteorology Validation At NPS Buoy
Solid obs Dashed - model
20
Conclusions

• Forecast ducting properties over the week of the
  experiment varied substantially.
• Duct location and strength were strongly
  dependent on the vertical gradient in mixing
  ratio
• The onset of ducting in the morning is often
  missed due to excess moisture above the surface -
  frequently observed during southerly flow
  conditions
• Transitions in ducting are missed or incorrectly
  timed

21
Conclusions (continued)

• Verification of ducting against the helicopter
  was hampered by limited sampling times
• Mean duct properties were in fair agreement but
  the standard deviations were in poor agreement
  for duct base height and duct thickness
• The overall percent correct in duct occurrence
  was 57 with a false alarm rate of 16 and a
  correct null rate of 84
• Verification of the model at the location of the
  NPS buoy showed a slight cold bias and relatively
  good agreement in terms of relative humidity,
  wind speed and direction, and pressure

22
Extra Slides
23
Why is microwave refractivity sensitive to water
vapor distribution?

• Ducts are frequently associated with large
  vertical gradient in water vapor
• In dry air, microwave refractivity is only
  influenced by atmospheric density
• At microwave frequencies, the water dipole
  reorients itself rapidly enough to follow and
  therefore modify the electromagnetic field
• At higher frequencies, e. g. visible light, this
  does not occur
• Thus, the refractive index of a moist atmosphere
  is larger for microwaves than for shorter
  wavelengths
• from Haack and Burk (2001)