environmental modeling center (emc)€¦ · adopted by council at its 62nd extraordinary session on...

Visit to NCEP 2 May 2006

Slide 1

HB 1

��

��

[email protected]

With input from many colleagues


Slide 2

HB 2

��

Ø Amendment of the Convention

Ø Changes and upgrades in the operational forecast system

Ø Deterministic and EPS performance

Ø Monthly/seasonal forecast system

Ø Products and dissemination

Ø Computing aspects

Ø Planned developments


Slide 3

HB 3

��

Ø Amendments to the ECMWF Convention were unanimously adopted by Council at its 62nd extraordinary session on 22 April 2005

Ø Process of acceptance of the amendments by Member States started immediately after: the procedure varies from one country to the other, but most will need Parliament ratification

Ø First acceptance was received in July (Finland)

Ø Finalization of the process is expected by end 2006 at the earliest


Slide 4

HB 4

��

Ø The adopted amendments concern mainly:

§ Allowing new Member States

§ Enlarging ECMWF’s mission to environmental monitoring

§ Redefining some decision processes (voting rights)

§ Widening the possibility for externally funded activities (e.g. by the European Union)

§ Extending official languages to all the official languages of the Member States, on a request-and-pay basis

Ø The amended Convention is available on ECMWF’s website


Slide 5

HB 5

�� ! ��"# ��$�� $

Early delivery systemCy28r229 Jun 2004

Wide range of changes including rain-

affected SSM/I radiancesCy29r228 Jun 2005

High resolution forecasting systemCy30r11 Feb 2006

New moist boundary layer scheme, wavelet

Jb, adaptive bias correctionCy29r15 Apr 2005

All BC analyses use background fields from

the latest 4D-Var9 Nov 2004

Harmonized bias corrections for AMSU and

HIRS radiances10 Oct 2004

Monthly forecasts run operationally7 Oct 2004

Wide range of changes affecting physics, use

of satellite data, EPSCy28r328 Sep 2004


Slide 6

HB 6

%��$��&�! '�(��)�%

Ø TL799 horizontal resolution for deterministic forecast

Ø 4D-Var increments at TL255 (30min time step)

Ø 91-level vertical resolution

Ø Changes to the wave model§ Grid spacing reduced from 0.5° to 0.36°

§ Use of Jason altimeter wave height data and ENVISAT ASAR spectrain the wave model assimilation

Ø TL399 L62 resolution for EPS

§ Wave model grid unchanged at 1° , but number of frequencies increased from 25 to 30, and number of directions changed from 12 to 24


Slide 7

HB 7

L60

L91'��

��

��


Slide 8

HB 8

* �� $�� +�� * �� $�� +�� * �� $�� +�� * �� $�� +��

40°N

90°W

90°W

85°W

85°W

80°W

80°W 75°W

75°W

40°N

90°W

90°W

85°W

85°W

80°W

80°W 75°W

75°W

Total # of points in TL511 grid: 348,528

Total # of points in TL799 grid: 843,490

TL511

TL799


Slide 9

HB 9

# �� ,��

��

��

T511 L60

T799 L91

Day

500hPa height anomaly correlationfor Europe

47 00UTC cases from 17 October 2005

T399 L62

T255 L40

Ensemble mean

Deterministic


Slide 10

HB 10

# �� ,��

T255 L40T399 L62

Day

47 00UTC EPS cases from17 October 2005

RP

SS

N Hem

Europe

RP

SS

T255 L40

T399 L62


Slide 11

HB 11

��-�� %��./�0��1��

OperationalT511 L60

72h forecast

36h forecast

OperationalT511 L60

TestT799 L91

TestT799 L91

+

++

+


Slide 12

HB 12

��


Slide 13

HB 13

�� 2��


Slide 14

HB 14

�� "��

Thanks to Adrian Simmons


Slide 15

HB 15

(��3"��4��

��

EUROPE LAT 35.000 TO 75.000 LON -12.500 TO 42.500

ANOMALY CORRELATION FORECAST

500hPa GEOPOTENTIAL

ECMWF FORECAST VERIFICATION 12UTC

Forecast Day MA = 12 Month Moving Average

198019811982198319841985198619871988198919901991199219931994199519961997199819992000200120022003200420054

5

6

7

8

9

10

SCORE REACHES 60.00

SCORE REACHES 60.00 MA


Slide 16

HB 16

Population: 92,92,92,92,92,92,92,92,92,92 (averaged)Mean calculation method: standard

Date: 20030601 00UTC to 20050831 00UTCEurope Lat 35.0 to 75.0 Lon -12.5 to 42.5

Anomaly correlation forecast500hPa Geopotential

Mean curves

1 2 3 4 5 6 7 8 9 10

Forecast Day20

30

40

50

60

70

80

90

100

2003

2004

2005

(�� 355�4��)6� 6�"��


Slide 17

HB 17

5

5

10°N

10°N

10°N

20°N

20°N

20°N

30°N

40°N

50°N

60°N

70°N

80°N

180°160°W140°W

120

°W10

0°W

80°W

60°W

40°W 20°W 0° 20°E 40°E

60°E

80°E

100°E

120°E

140°E160°ERMSE Difference Z500 JJA 2005-2004: D+5 FC

-40-30-25-20-15-10-5

5101520253040

-8

-2

-2

-2

-2

2

10°N

10°N

10°N

20°N

20°N

20°N

30°N

40°N

50°N

60°N

70°N

80°N

180°160°W140°W

120°

W1

00°

W

80°W

60°W

40°W 20°W 0° 20°E 40°E

60°E

80°E

100°E

120°E

140°E160°EEPS Spread Z500 JJA 2005-2004: D+5 EPS Control

-25-20-16-12-8-4-2

24812162025

7�(�� $��55��

��"��


Slide 18

HB 18

(�� "8

�� .-��


Slide 19

HB 19

Population: 178,177,177,177,177,177 (averaged)Mean calculation method: standard

Date: 20041201 00UTC to 20060228 00UTCEurope Lat 35.0 to 75.0 Lon -12.5 to 42.5


Mean curves

1 2 3 4 5 6

Forecast Day70

75

80

85

90

95

100

ec 2004/2005

ec 2005/2006

uk 2004/2005

uk 2005/2006


Slide 20

HB 20


Date: 20041201 00UTC to 20060228 00UTCEurope Lat 35.0 to 75.0 Lon -12.5 to 42.5Root mean square error forecast

500hPa GeopotentialMean curves

1 2 3 4 5 6

Forecast Day0

10

20

30

40

50

60

70

80

90

100

ec 2004/2005

ec 2005/2006

uk 2004/2005

uk 2005/2006


Slide 21

HB 21


Date: 20060201 00UTC to 20060420 00UTCN.hem Lat 20.0 to 90.0 Lon -180.0 to 180.0


Mean curves

1 2 3 4 5 6

Forecast Day75

80

85

90

95

100

ec

ex

nc

ensmn


Slide 22

HB 22

1 2 3 4 5 6 7 8

Forecast Day0

20

40

60

80

100

120

Mean values 200408 to 200507RMSE (m)

500 hPa GEOPOTENTIAL HEIGHTVERIFICATION AGAINST RADIOSONDES

EUROPEVERIFICATION TO W.M.O. STANDARDS ECMWF 00

DWD 00 FRANCE 00

UK 00 NCEP 00 CANADA 00

1 2 3 4 5 6 7 8

Forecast Day3

4

5

6

7

8

9

10

11

Mean values 200408 to 200507RMSEV (m/s)850 hPa WIND

VERIFICATION AGAINST RADIOSONDESEUROPE

VERIFICATION TO W.M.O. STANDARDS ECMWF 00

DWD 00 FRANCE 00

UK 00 NCEP 00 CANADA 00

850hPa wind 500hPa height

��6�9(�:��


Slide 23

HB 23

0.1

0.2

0.3

0.4

0.5

TS

S

1993J

1994J

1995J

1996J

1997J

1998J

1999J

2000J

2001J

2002J

2003J

2004 2005

t + 42 t + 90

0.2

0.3

0.4

0.5

0.6

0.7

TS

S

0.2

0.3

0.4

0.5

0.6

0.7

1993D

1994D

1995D

1996D

1997D

1998D

1999D

2000D

2001D

2002D

2003D

2004 2005

precipitation exceeding 10.0 mm/24h

t + 42 t + 90

Europe 10mm/24h

��

� � ��

(;+��</(Winter months (DJF)

Summer months (JJA)

0.3

0.7

0.2

0.4

TSS = H - F


Slide 24

HB 24

�� $�� (��


Slide 25

HB 25

JJAOPER T+168 12 UTC

Europe Lat 35.0 to 75.0 Lon -12.5 to 42.5850hPa Temperature

Anomaly correlation forecastCumulative distribution of

1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 20060

20

40

60

80

100

< -40.0-40.0 to -10.0-10.0 to 10.010.0 to 20.020.0 to 30.030.0 to 40.0

40.0 to 50.050.0 to 60.060.0 to 70.070.0 to 80.080.0 to 90.0> 90.0

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

EPS Mean

(�� "(�� "(�� "(�� "Europe


Slide 26

HB 26

9 �(= ��(��% ��(��

� ��>"�� ?

Black -8KBlue -4kRed +4kGreen +8K

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1994A

1995A D A

1996A D A

1997A D A

1998A D A

1999A D A

2000A D A

2001A D A

2002A D A

2003A D A

2004A D A

2005A

Brier skill score (long term clim) fc step 144 T850 anomaly exceedingProbability forecast verification against an ( 3-M. moving sample)

-8 K -4 K 4 K 8 K


Slide 27

HB 27

��(��


Slide 28

HB 28

��

Ø Weekly 32-day forecasts T159L62For

Ø Coupled to HOPE ocean model

Ø 51 member ensemble

Ø Compared to 5 forecasts for same day over last 12 years

§ 60 member ensemble

Ø Results interpreted in terms of anomalies

§ For example: probability that 2m temperature averaged over day 12 to 18 is in the upper/middle/lower tercile

Ø Products become available every Thursday at 22UTC


Slide 29

HB 29

��


Slide 30

HB 30

(�� 8�

�.7�('�� 2� �� $�

Ø Three models running at ECMWF:

§ ECMWF – System 2

§ Met Office – HADCM3 model, Met Office ocean analyses

§ Meteo-France – Arpege/Climat, Mercator ocean analyses

§ Spain + Germany may join

Ø Unified system

§ All data in ECMWF operational archive

§ Common operational schedule (products released at 12UTC on the 15th of each month)

Ø Common products will be available soon

§ ECMWF release of web products expected this autumn

Ø EUROSIP appears to be better than the individual systems


Slide 31

HB 31

30°N 30°N

45°N45°N

60°N 60°N

30°W

30°W 0°

0° 30°E

30°E

30°N 30°N

45°N45°N

60°N 60°N

30°W

30°W 0°

0° 30°E

30°E

Probability of 2m Temp > upper tercileof the model climate for JJA 2005

Seasonal forecasts from 1 May 2005Met Office

ECMWF

Météo France


Slide 32

HB 32

�� *5��&8�� *5��&8�� *5��&8�� *5��&8�

?� ?� ?� ?�?�+��"?�+��"?�+��"?�+��"

ECMWF

Met Office

Météo France

-4

-3-3

-3

-3

-2

-2 -2 -2

-2

-2

-1

-1 -1

-1

-1

-1

-1

-1

1

1

1

1

11

22

22

22

2

3

3 3

33 3

34

4

4

44

4

55

555

5 5

10

10

-2-2

-2

-2 -2

-2

-2

-2-2

-2

-2

2

2

2

22 2

2

2 2

22 4

4

4 6

-10

-5

-4

-3

-2

-1

1

2

3

4

5

10

Observed anomalies


Slide 33

HB 33

* �� 2��


Slide 34

HB 34

'��

Ø Global Extreme Forecast Index

§ Anomalies for 2m temperature, 10m wind and precipitation

§ Will be made available as a web product

§ Verification against observations has started

Ø Products from the monthly forecasting system were added to the dissemination and the Catalogue

§ Includes re-forecasts for both atmospheric and wave products

Ø Wave kurtosis parameter

§ Available from deterministic and ensemble system, can be used to estimate the likelihood of a freak wave occurring in the oceans

Ø CAPE parameter

§ Available from deterministic and ensemble system


Slide 35

HB 35

H

H

L

L

0

0

0

00

8

8

8

8

88

8

16

16

16

16

16

16

1616

16

24

24

24

24

32

3240°W

40°W

0°

0° 40°E

40°E

VT: 200411141000 hPa Z ensemble mean and EFI values for 24h TP, wind gust and 2m temperature

Weather anomalies predicted by EPS: 20041112 0 UTC

warm extreme warm extreme cold cold windy extreme wind Heavy precip Extreme precip��'�

��

% 6%%6�


Slide 36

HB 36

1038

1033

975

1001

1012

1014

995

99 5

1005

1005

1015

1015

1015

1015

1015

1015

1 01 5

1025

1025

1025

1025

1025

1025

1035

15.0m/s

80°W

80°W

40°W

40°W

Thursday 14 April 2005 00UTC ECMWF Forecast t+36 VT: Friday 15 April 2005 12UTC Surface: Wave spectral kurtosisThursday 14 April 2005 00UTC ECMWF Forecast t+36 VT: Friday 15 April 2005 12UTC Surface: mean sea level pressure/Surf: 10 mtr v

0.05

0.075

0.1

0.125

0.15

0.175

0.2

0.225

0.25

0.275

0.3

��

=��

%"6 6�"

Freak waves


Slide 37

HB 37

��


Slide 38

HB 38

��


Slide 39

HB 39

'9��#��)

Ø Two identical clusters, each one comprising

§ 68 compute nodes

§ Each node has 32 Power4+ processors and 32 GB of memory

Ø Dual-cluster has been in full production since early 2005

Ø System reliability is high (99.6% and 98.9% respectively)


Slide 40

HB 40

'9��#�� $��

Ø Council approved in December 2004 a two-year extension of the contract with IBM

Ø The extended contract includes the replacement of the current clusters with two Power5+ based clusters

Ø The aggregate performance on ECMWF’s benchmark codes will increase from 2.5 to 4.5 teraflops sustained

Ø There are no further upgrades planned under the revised contract and the Phase 4 system will remain at ECMWF until March 2009

Ø The next HPC procurement will take place in 2007


Slide 41

HB 41

��


Slide 42

HB 42

@�7��(�3�� &4

Ø Plan to extend forecast range of EPS to day 15

Ø Two additional fixed-resolution calibration forecasts will be run

Ø Current products to day 10 will be unaffected

Ø T799 deterministic will remain as now (days 1-10 only)

TL2551CL-LRES

TL3991CL-HRES

TL255TL39950VAR-PF

TL255TL3991VAR-CF

day 10-15day 0-10# fc


Slide 43

HB 43

@�7��(�3�� 4

Ø EPS products to day 15 will be developed

Ø 15-day VAREPS will be combined with the monthly system (coupling with ocean model)

Ø Current implementation with change of resolution at day 10 is likely to be revised in the light of experience

Ø Depending on ongoing research, the change in resolution may be brought forward in time (e.g. to day 7)


Slide 44

HB 44

�� :��! '�(��)��35��&4

Ø Variational radiance bias correction

Ø Thinning of low-level AMDAR data

Ø Improved treatment of ice sedimentation and autoconversion to snow in cloud scheme

Ø Implicit treatment of convective momentum transport

Ø Includes changes for EPS extension to day 15

§ T255 perturbed forecasts from day 10 to day 15

§ T399/255 control to day 10/15

§ Also uniform T399 control to day 15 and T255 control to day 21

Ø To be used for T255L91 interim reanalysis for 1989 - ….

Ø To be used in version 3 of Seasonal Forecasting System?


Slide 45

HB 45

��A��)A��&8

Ø Introduction of turbulent orographic drag scheme

Ø Use of albedo fields from MODIS

Ø Use of high-resolution NCEP SST fields

Ø Refinements to stratospheric analysis

Ø Recalibrated radiosonde temperature bias corrections


Slide 46

HB 46

�� A��&8

Ø 4D-Var changes:§ 3rd inner loop§ revised trajectory interpolation§ revised data usage, including modified Var QC§ new cloud and convection schemes in minimization

Ø Upgrade fast radiative transfer to RTTOV-9

Ø Change model short-wave radiation scheme to RRTM-SW

Ø Unified medium-range/monthly EPS

Ø Start of GEMS constituent reanalyses for 2003/4

And over course of year:

Ø Monitoring and later assimilation of data from:§ AMSR-E, CHAMP, COSMIC, FY-2C, METOP ATOVS + … , MET9, MTSAT,

SSMIS, TMI


Slide 47

HB 47

Ø The interim reanalysis§ will start soon with T255L91, 4D-Var and cycle 30r2 (?)

§ will run from 1989 onwards, and be continued in close to real time

§ will use mostly same pre-2002 data as ERA-40, but will include

• reprocessed winds from EUMETSAT

• reprocessed ERS altimeter data

• GOME profile data from RAL

• improved radiosonde bias corrections

Ø European Regional Reanalysis (EURRA)

§ potential for a European regional reanalysis project (EURRA) is being explored with NMSs and EEA

Ø A new major reanalysis (ERA-65 or -75) in due course§ funding has to be secured

ECMWF plans for reanalysis


Slide 48

HB 48

��(��&2��%"

Ø The principal goal of ECMWF in the coming ten years will be to maintain the current, rapid rate of improvement of its global, medium-range weather forecasting products, with particular effort on early warnings of severe weather events

Ø Complementary goals are:§ To improve the quality and scope of monthly and seasonal-to-interannual

forecasts

§ To enhance support to Member States national forecasting activities by providing suitable boundary conditions for limited-area models (and regional CTMs)

§ To deliver real-time analyses and forecasts of atmospheric composition

§ To carry out climate monitoring through regular reanalyses of the Earth-system

§ To contribute towards the optimization of the Global Observing System


Slide 49

HB 49

B��C�� 0��=��D

environmental modeling center (emc)€¦ · adopted by council at its 62nd extraordinary session on...

Documents