yucheng song & zoltan toth 1 yucheng song and zoltan toth emc/ncep/nws/noaa update on the winter...
TRANSCRIPT
Yucheng Song & Zoltan Toth
1
Yucheng Song and Zoltan Toth
EMC/NCEP/NWS/NOAA
Update on the Winter Storm Reconnaissance Program
Meeting of the working group on Space-based Lidar winds, Miami, Florida Feb 6-9
Yucheng Song & Zoltan Toth
2
Acknowledgments
• NWS field offices, HPC/NCEP and SDMs• NOAA G-IV and the USAFR C-130 flight crews• CARCAH (John Pavone)• Sharanya Majumdar Univ. of Miami/CIMAS• Jack Woollen - EMC• Russ Treadon - EMC• Mark Iredell - EMC• Istvan Szunyogh – Univ. of Maryland• Wayman Baker, JCSDA• + others who have contributed!
Yucheng Song & Zoltan Toth
3
Collaborators
• Sharanya Majumdar Univ. of Miami/CIMAS• Craig Bishop Naval Research Lab• Christopher Velden Univ. of Wisc./CIMSS• Milija Zupanski Colo. State Univ./CIRA• Thomas Hamill NOAA/Climate Diagnostics Lab• Istvan Szunyogh Univ. of Maryland• Robert Atlas NASA/GSFC• David Emmitt, Simpson Weather Associates, SMA
Yucheng Song & Zoltan Toth
4
Winter Storm Reconnaissance (WSR 2006) Program
• Took place 27 Jan – 9 March 2006
• Dropwinsonde observations taken over the NE Pacific by aircraft operated by NOAA’s Aircraft Operations Center (G-IV) and the US Air Force Reserve (C-130s).
• Observations are adaptive –
– collected only prior to significant winter weather events of interest
– in areas that might influence forecast the most.
• Operational since January 2001
• 26 flights, around 342 dropsondes this winter which is reduced from 500 drops last year
Winter Storm Reconnaissance Program
Objective:
Improve Forecasts of Significant Winter Weather Events Through Targeted Observations in Data Sparse Northeast Pacific Ocean
Approach:
1) Collected Only Prior to Significant Winter Weather Events of Interest
2) Collected in Areas that Influence the Forecast the Most
Results:
About 70% of Targeted Numerical Weather Predictions Improve Significantly Due to Winter Storm Reconnaissance Program (Operational since January 2001)
Yucheng Song & Zoltan Toth
7
Overall results for Temperature
Of the cases:22 improved 1 neutral 7 degraded
RMS errors averaged in the verification regions for forecasts with and without WSR data
Yucheng Song & Zoltan Toth
8
Overall results for Vector Wind
Of all cases:19 improved 0 neutral11 degraded
RMS errors averaged in the verification regions for forecasts with and without WSR data
Yucheng Song & Zoltan Toth
9
Overall results for Specific humidity
Of all cases:19 improved 0 neutral11 degraded
RMS errors averaged in the verification regions for forecasts with and without WSR data
Yucheng Song & Zoltan Toth
10
Forecast Verification for Wind
RMS error reduction vs. forecast lead time
Yucheng Song & Zoltan Toth
11
Individual Case Comparison
1 denotes positive effect
0 denotes neutral effect
-1 denotes negative effect
19 OVERALL POSITIVE
0 OVERALL NEUTRAL
8 OVERALL NEGATIVE
70% improved 30 % degraded
VR OBSDATE P T V OVERALL REGION FHOUR W 20060127 1 1 1 1 124W ,45N 48 E 20060129 -1 -1 1 -1 75W ,40N 72 E 20060201 1 1 1 1 83W ,36N 72 W 20060203 -1 1 -1 -1 124W ,42N 36 E 20060203 1 1 1 1 84W ,35N 48 E 20060204 -1 -1 1 -1 68W ,45N 48 AK 20060206 -1 1 1 1 140W ,60N 24 W 20060209 -1 -1 -1 -1 122W ,45N 60 AK 20060210 1 1 1 1 131W ,55N 60 AK 20060212 1 -1 1 1 140W ,60N 12 E 20060214 -1 -1 -1 -1 88W ,42N 84 E 20060221 -1 1 1 1 87W ,42N 48 C 20060221 1 -1 1 1 95W ,33N 48 W 20060223 1 1 -1 1 120W ,47N 24 W 20060225 -1 -1 1 -1 122W ,40N 72 W 20060226 1 -1 -1 -1 122W ,40N 36 AK 20060226 1 1 1 1 150W ,60N 48 W 20060227 -1 1 1 1 118W ,34N 36 H 20060227 -1 1 1 1 158W ,22N 84 W 20060301 1 -1 1 1 124W ,42N 24 W 20060303 -1 1 1 1 123W ,45N 84 W 20060304 1 1 1 1 123W ,40N 48 W 20060305 -1 1 1 1 121W ,40N 36 W 20060307 1 1 -1 1 122W ,46N 48 C 20060307 1 1 1 1 90W ,37N 60 E 20060307 -1 -1 1 -1 85W ,45N 84 E 20060309 -1 1 1 1 80W ,50N 84
Yucheng Song & Zoltan Toth
12
Impact of WSR targeted dropsondes (2006)
1 Jan – 28 Feb 2006 00UTC Analysis
NOAA-WSRP 191 Profiles
Beneficial (-0.01 to -0.1 J kg-1)
Non-beneficial (0.01 to 0.1 J kg-1)Small impact (-0.01 to 0.01 J kg-1)
Binned Impact
Average dropsonde obs impact is beneficial and ~2-3x greater than average radiosonde impact (From Dr. Rolf Langland, Naval Research Lab.)
Composite summary maps
139.6W 59.8N 36hrs (7 cases) 92W 38.6N 60hrs (5 cases)
122W 37.5N 49.5hrs (8 cases) 80W 38.6N 63.5hrs (8 cases)
Verification Region
Verification Region
Shaded is the ETKF derived error reduction in Verification region and contours are sea level pressure
Yucheng Song & Zoltan Toth
14
3 649.5
60
63.5
0
1000
2000
3000
4000
5000
6000
0 20 40 60 80F o r e c a s t H o u r s
D i s t a n c e ( k m )
ETKF predicted signal propagation
Yucheng Song & Zoltan Toth
15
5143km63.5hrs (East Coast)
4064km60hrs (Central U.S)
2034km49.5hrs (West Coast)1422km36 hrs (Alaska)
Forecast hours vs. Distance
The above table listed forecast hours and the distance between the centers of sensitivity and verification regions
Yucheng Song & Zoltan Toth
16
WSR overall statistics (2004-2006)
Variable# cases
improved# cases neutral
#cases degraded
Surface pressure
21+20+13=54 0+1+0=1 14+9+14=37
Temperature 24+22+17=63 1+1+0=2 10+7+10=27
Vector Wind 23+19+21=63 1+0+0=1 11+11+6=28
Humidity 22+19+13=54 0+0+0=0 13+11+14=38
25+22+19 = 66 OVERALL POSITIVE CASES.
0+1+0 = 1 OVERALL NEUTRAL CASES.
10+7+8 = 25 OVERALL NEGATIVE CASES. 71.7% improved 27.1% degraded
Yucheng Song & Zoltan Toth
17
Future Work
• Improve ensemble products to give better guidance to HPC and SDMs– using “energy norms” in addition to the conventional products– Better ensemble products
• Improve targeting method based on ETKF method with increasing resolution and ensemble membership– Explore parallel structure of the codes
• Improve verification techniques• High resolution (T382L64?)• WSR 2007 increased more ensemble members from ECMWF,
NCEP • Moisture data will be assimilated in the verification• Tropical flareups and Arctic signals – T-PARC and IPY
Yucheng Song & Zoltan Toth
22
NRL P-3 and NRL P-3 and HIAPER with theHIAPER with theDLR Wind LidarDLR Wind Lidar
NRL P-3 and NRL P-3 and HIAPER with theHIAPER with theDLR Wind LidarDLR Wind Lidar
Upgraded Russian Radiosonde Network for IPY
Winter storm reconnaissanceand driftsonde
T-PARC Experiments and Collaborative Efforts
Yucheng Song & Zoltan Toth
23
Winter component of T-PARC
• December 08 to February 09
• Improving adaptive targeting methods for 3-5 days ET SV (possible collaboration with NRL)
• Adaptive use of DWL, Driftsondes, Radiosondes, special IPY observations as well as satellite data can enhance high impact weather events in Arctic and CONUS, Enhanced Siberian Observation network, MTSAT rapid scan, Polar orbiting platforms
Yucheng Song & Zoltan Toth
24
Goal
• Contribute to the design and utilization of optimal global observing network, including adaptive components
• Contribute to Global Earth Observation System of Systems (GEOSS)
• Improved forecasts translate into enhanced guidance for users and real life applications
• Develop new products (sea ice, freezing spray, storm surge)• Probabilistic approach to model uncertainties,
downscaling/post-processing better representation in model of high impact weather events
Yucheng Song & Zoltan Toth
25
Data issues (AMMA example)
Temperature and Wind Sounding Counts for Site 65578
0
2
4
6
8
10
12
14
16
jun05 jul05 aug05 sep05 jun06 jul06 aug06 sep06
Month and Year
Nu
mb
er o
f S
ou
nd
ing
s TS00
TS06
TS12
TS18
WS00
WS06
WS12
WS18
Daily take-in of the AMMA data for Temperature and wind at one station
An example showing Track-Check Problems with African AMDAR Data during AMMA period
Courtesy of Dr. Bradley Ballish
Yucheng Song & Zoltan Toth
26
Background information - ETKF
aTaafTff
fTfTffa
ZZPZZP
HPRHHPHPPP
while
operatorn observatio : H
matrix covarianceerror n observatio : R
matrix covarianceerror forecast :P
matrix covarianceerror analysis : P
(1) )(
f
a
1
Yucheng Song & Zoltan Toth
27
Continued
seigenvalue
rseigenvectoC
CCHZRH
HZRHZI
TZZ
onperturbatizz
zzzk
Z
zzzk
Z
kkfTfT
fTfT
fa
ai
fi
ak
aaa
fk
ffT
f
thecontainingmatrix diagonal:
and lorthonorma :
][Z andI)C(T
issolution One
][TT
have will we(1) into thissubstitute weIf
assumes one ETKF,In
analysis andforecast ensembleith theare ,
],....,,[1
1
],....,,[1
1
1 1/2-
11T
11
11