bidlot ocean wave analysis - ecmwf · reanalysis workshop slide 3 background : zecmwf has been...

Ocean Wave Analysisand

use of surface wind observationsover the oceans

Jean-Raymond BidlotPeter JanssenSaleh Abdalla

Hans Hersbach

European Centre for Medium Range Weather Forecasts, Reading, U.K.

[email protected]

ECMWF/GEO Workshop on Atmospheric Reanalysis. 19-22 June 2006

Reanalysis Workshop Slide 2

Outline :

Background.

Wave data available for reanalysis.

Wave data analysis at ECMWF.

ERA-40 wave data problem.

Wave model for the interim reanalysis.

Wave data for the interim reanalysis.

A few comments on in-situ surface wind.


Background :ECMWF has been running operational wave models since June 1992.

Two-way coupling between the wave model and the atmospheric model was introduced in June 1998 with benefits for both wave model (via high temporal wind field resolution) and atmospheric model (via wave dependent surface roughness).

More than just winds are now exchanged between the two models and more wave dependent processes could be added (i.e. sea-state dependent satellite retrievals, wave-current interactions).

Reanalysis wave data and wind data have many applications, quite often in coastal areas, hence the need for an increase in horizontal resolution.


Reanalysis wave data :

Wave height 100 year return value as derived from ERA-40as available in the KNMI wave and wind atlas:http://www.knmi.nl/waveatlas


Lon: 342-351Lat: 54N-45NDir: 0.- 45.Total of:67718. observations,4.%

Tm \ Hs 0.-1. 1.- 2. 2.- 3. 3.- 4. 4.- 5. 5.- 6. 6.-

7.7.-8.

8.-9.

9.-10.

10.- 11.

11.- 12.

12.- 13.

13.- 14.

14.- 33. sum

0.- 3. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.3.- 4. 2. 15. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 17.4.- 5. 24. 883. 122. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1029.5.- 6. 36. 1016. 1625. 172. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 2849.6.- 7. 3. 647. 1013. 960. 242. 2. 0. 0. 0. 0. 0. 0. 0. 0. 0. 2868.7.- 8. 0. 259. 697. 432. 379. 166. 13. 0. 0. 0. 0. 0. 0. 0. 0. 1946.8.- 9. 0. 43. 355. 223. 92. 92. 74. 21. 1. 0. 0. 0. 0. 0. 0. 901.9.- 10. 0. 1. 69. 148. 38. 19. 12. 18. 12. 2. 0. 0. 0. 0. 0. 319.10.- 11. 0. 0. 1. 16. 27. 14. 3. 1. 1. 2. 1. 0. 0. 0. 0. 67.11.- 12. 0. 0. 0. 0. 0. 3. 0. 0. 0. 0. 0. 0. 0. 0. 0. 3.12.- 13. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.13.- 33. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.sum 65. 2864. 3882. 1952. 779. 296. 102. 41. 14. 3. 1. 0. 0. 0. 0. 10000.

i 4 90

Reanalysis wave data :

Tabular wave height and mean period bivariate histogramfor an area west of Ireland as derived from ERA-40:http://www.knmi.nl/waveatlas


Wave data for reanalysis : from satellites

Radar altimeter wave height observations:

- Seasat (1978), Geosat (1985-1986), ERS-1* (1991-1996), Topex(1992-2002),

- ERS-2* (1995-), GFO (2000-), Jason* (2002-), ENVISAT* (2002-)

Note: the wave model predicts the evolution of the wave spectrum. Wave height is connected to the total energy, which isproportional to the integral of the wave spectrum. Hence some assumptions are required to go from analysed wave height to analysed spectrum.

* Currently available in the ECMWF archives



Note: radar altimeter are nadir looking instrument, with a very narrow swath.Their global coverage is therefore limited.

ENVISAT in 24 hrs Jason in 24 hrs



SAR/ASAR ocean mode data:

- ERS-1* and 2* SAR

- ENVISAT* ASAR

Note: SAR does not unfortunately observe the full two-dimensional wave spectrum. In practice SAR wave spectra are retrieved using a model first guess. For these reasons, the impact of SAR assimilation is much reduced than originally hoped.

* Currently available in the ECMWF archives


Wave data for reanalysis verification: in-situ

40°S40°S

30°S 30°S

20°S20°S

10°S 10°S

0°0°

10°N 10°N

20°N20°N

30°N 30°N

40°N40°N

50°N 50°N

60°N60°N

160°W

160°W 140°W

140°W 120°W

120°W 100°W

100°W 80°W

80°W 60°W

60°W 40°W

40°W 20°W

20°W 0°

0° 20°E

20°E 40°E

40°E 60°E

60°E 80°E

80°E 100°E

100°E 120°E

120°E 140°E

140°E1991-08-01 to 1996-06-03

Number of collocations between ERS-1 altimeter and buoys

0

150

300

450

600

750

900

1050

1200

1350

1500

1650

1747.3

40°S40°S

30°S 30°S

20°S20°S

10°S 10°S

0°0°

10°N 10°N

20°N20°N

30°N 30°N

40°N40°N

50°N 50°N

60°N60°N

160°W

160°W 140°W

140°W 120°W

120°W 100°W

100°W 80°W

80°W 60°W

60°W 40°W

40°W 20°W

20°W 0°

0° 20°E

20°E 40°E

40°E 60°E

60°E 80°E

80°E 100°E

100°E 120°E

120°E 140°E

140°E2003-07-21 to 2006-05-13

Number of collocations between ENVISAT altimeter and buoys

0

150

300

450

600

750

900

1018.4

Buoy reporting wave observationshave a relatively limited geographical coverage.They are used for verification only.

Jun-

1987

Jun-

1988

Jun-

1989

Jun-

1990

Jun-

1991

Jun-

1992

Jun-

1993

Jun-

1994

Jun-

1995

Jun-

1996

Jun-

1997

Jun-

1998

Jun-

1999

Jun-

2000

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

months

0

10000

20000

30000

40000

50000

Num

ber o

f col

loca

tions

with

mod

el

Comparison with buoy data3 month collocation period

1987 2006The number of wave buoys available on the GTS has steadily increased over the years.

40°S40°S

30°S 30°S

20°S20°S

10°S 10°S

0°0°

10°N 10°N

20°N20°N

30°N 30°N

40°N40°N

50°N 50°N

60°N60°N

160°W

160°W 140°W

140°W 120°W

120°W 100°W

100°W 80°W

80°W 60°W

60°W 40°W

40°W 20°W

20°W 0°

0° 20°E

20°E 40°E

40°E 60°E

60°E 80°E

80°E 100°E

100°E 120°E

120°E 140°E

140°E1996-06-03 to 2003-06-22

Number of collocations between ERS-2 altimeter and buoys

01503004506007509001050120013501500165018001950210022502400


Wave data analysis:

The wave model data analysis is based on an optimal interpolation scheme for wave height as originally developed by Lionello et al. 1992.

The scheme was extended to assimilate wave systems as derived from SAR wave spectra.

It is not yet directly coupled to the atmospheric scheme.


ERA-40 :

1.5ºx1.5º wave model coupled to T159 atmospheric model.

Only deep water physics.

Wave spectra discretised with 12 directions and 25 frequencies.

Near-real time operationally obtained ERS-1 (1991-1996) and ERS-2 (1996-2002) altimeter data were used.

An extensive comparison with buoy data and Topex altimeter data was carried by a KNMI team http://www.knmi.nl/waveatlas


Too low wind in ERA-40 :

Jun-

1987

Jun-

1988

Jun-

1989

Jun-

1990

Jun-

1991

Jun-

1992

Jun-

1993

Jun-

1994

Jun-

1995

Jun-

1996

Jun-

1997

Jun-

1998

Jun-

1999

Jun-

2000

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

months

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

Win

d Sp

eed

Bias

(mod

el-b

uoy)

(m/s

)

ERA-40 (e4 0001)ops (od 0001)

Comparison with buoy data3 month running average

1987 2006

Wind speed bias

Jun-

1987

Jun-

1988

Jun-

1989

Jun-

1990

Jun-

1991

Jun-

1992

Jun-

1993

Jun-

1994

Jun-

1995

Jun-

1996

Jun-

1997

Jun-

1998

Jun-

1999

Jun-

2000

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

months

-0.4

-0.2

0

0.2

0.4

Wav

e H

eigh

t Bia

s (m

odel

-buo

y) (m

)

CY30R1 stand alone hindcast (rd era5)


1987 2006

Hs bias

Results from a wave model hindcast run(uncoupled, no data assimilation),with the latest operational model version (CY30R1),forced by ERA-40 winds until August 2002, thenoperational analysis winds.


Problem with wave data in ERA-40 :

Jun-

1987

Jun-

1988

Jun-

1989

Jun-

1990

Jun-

1991

Jun-

1992

Jun-

1993

Jun-

1994

Jun-

1995

Jun-

1996

Jun-

1997

Jun-

1998

Jun-

1999

Jun-

2000

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

months

-0.4

-0.2

0

0.2

0.4

Wav

e H

eigh

t Bia

s (m

odel

-buo

y) (m

)

ERA-40 (e4 0001)ops (od 0001)


1987 2006Dec. 1991

May 1993

Between Dec. 1991 and May 1993low quality ERS-1 altimeter datawere wrongly assimilated.

Jun-

1987

Jun-

1988

Jun-

1989

Jun-

1990

Jun-

1991

Jun-

1992

Jun-

1993

Jun-

1994

Jun-

1995

Jun-

1996

Jun-

1997

Jun-

1998

Jun-

1999

Jun-

2000

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

months

-0.5

0

0.5

1

1.5

2

Wav

e M

ean

Perio

d B

ias

(mod

el-b

uoy)

(s) ERA-40 (e4 0001)

ops (od 0001)


1987 2006Dec. 1991

May 1993

Between Dec. 1991 and May 1993low quality ERS-1 altimeter datawere wrongly assimilated.

Hs bias Tz bias

Note: this problem was originally spotted by KNMI. This emphasises theneed for ‘in-house’ monitoring tools.


Better surface winds for the interim reanalysis :

T159 T255 better (ideally more)

Benefit from using 4dvar.

Improved atmospheric model physics.

The latest physics changes (CY31R1) should be even better.

Jun-

1987

Jun-

1988

Jun-

1989

Jun-

1990

Jun-

1991

Jun-

1992

Jun-

1993

Jun-

1994

Jun-

1995

Jun-

1996

Jun-

1997

Jun-

1998

Jun-

1999

Jun-

2000

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

months

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

Win

d Sp

eed

Bias

(mod

el-b

uoy)

(m/s

)

ERA-40 (e4 0001)ops (od 0001)preliminary interim reanalysis (e4 0060)interim reanalysis prototype (ei 1001)interim analysis prototype (ei 1103)


1987 2006

Wind speed bias


Improved wave model for the interim analysis :

1.5º 1.0º

24 directions and 30 freq.

Shallow water physics.

Improved wave model.-Unresolved bathymetry scheme.

-New dissipation source term.

-…

Jun-

1987

Jun-

1988

Jun-

1989

Jun-

1990

Jun-

1991

Jun-

1992

Jun-

1993

Jun-

1994

Jun-

1995

Jun-

1996

Jun-

1997

Jun-

1998

Jun-

1999

Jun-

2000

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

months

10

12

14

16

18

20

22

24

26

Wav

e M

ean

Perio

d Sc

atte

r Ind

ex (%

)

ERA-40 (e4 0001)ops (od 0001)CY30R1 hindcast (rd era5)preliminary reanalysis (e4 0060)interim prototype (ei 1001)interim reanalysis prototype (ei 1103)


1987 2006

Tz bias


Wave data for the interim reanalysis :

Because of the inhomogeneity of the ERS-1 and also some other known problems with the NRT ERS data, we have obtained OPR ERS-1 and 2 data from ESA.

Some works was necessary to convert the data from CD’s into data in a format more compatible with ECMWF system (i.eBUFR)!

NRT ENVISAT altimeter data are already of good quality. Note that some orbits are missing, especially in early days of the mission due to NRT transmission problems. Access to off-lindata could prove useful.

NRT Jason altimeter data are also available.

SAR data from ERS-2 and ASAR data from ENVISAT will also be used.



Because the OPR ERS-1 and 2 have never been used by our data assimilation system, we have run passively the altimeter data through a long wave model hindcast with the latest operational wave model version (CY30R1) using ERA-40 and operational winds.

A comprehensive buoy data set was generated by merging NDBC (US) and MEDS (Canada) data sets with GTS data from the ECMWF archives.

This comprehensive buoy data set was collocated with the altimeter data and the model counterparts.

Triple collocation techniques can be used to infer relative biases and error estimates for each data set.



ERS-1 OPR wave heights are stillbiased low when compared to buoys

ERS-2 OPR wave heights are slightlybiased low when compared to buoys

ENVISAT wave heights are slightlybiased high when compared to buoys

0 1 2 3 4 5 6 7 8 9 10 11 12 13GTS wave heights

0123456789

10111213

Cor

rect

ed a

lt. w

ave

heig

hts ENTRIES:

1 - 33 - 66 - 12

12 - 26

26 - 59

59 - 133

133 - 300

ERS-1

1991-08-01 to

1996-06-03

Buoy wave heights

0 1 2 3 4 5 6 7 8 9 10 11GTS wave heights

0

1

2

3

4

5

6

7

8

9

10

11

Cor

rect

ed a

lt. w

ave

heig

hts ENTRIES:

1 - 33 - 77 - 15

15 - 37

37 - 90

90 - 220

220 - 540

ERS-2

1996-06-03to

2003-06-22

Buoy wave heights

0 1 2 3 4 5 6 7 8 9 10 11GTS wave heights

0

1

2

3

4

5

6

7

8

9

10

11

Cor

rect

ed a

lt. w

ave

heig

hts ENTRIES:

1 - 33 - 55 - 10

10 - 22

22 - 46

46 - 98

98 - 210

ENVISAT

2003-07-21to

2006-05-13

Buoy wave heights



Using a non-parametric bias estimation technique, we have determined the relative bias with respect to the buoys (work still under way).

Initial results are quite satisfactory. More test will follow.0 1 2 3 4 5 6 7

wave height (m)

-0.8

-0.6

-0.4

-0.2

0

0.2

BIA

S fo

r alti

met

er w

ave

heig

ht a

s pr

oces

sed

by W

AM

ERS-1ERS-2ENVISAT

BIAS as derived from non-parametric estimate


Wave data for interim reanalysis :

Jun-

1987

Jun-

1988

Jun-

1989

Jun-

1990

Jun-

1991

Jun-

1992

Jun-

1993

Jun-

1994

Jun-

1995

Jun-

1996

Jun-

1997

Jun-

1998

Jun-

1999

Jun-

2000

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

months

14

16

18

20

22

24

26

28

30

Wav

e H

eigh

t Sca

tter I

ndex

(%)

ERA-40 (e4 0001)CY30R1 stand alone hindcast (rd era5)interim reanalysis prototype (ei 1103)


1987 2006

with bias corrected ERS-1 data


In-situ surface wind data over the ocean :

The quality of any wave reanalysis is largely determined by the quality of the wind fields used to force the model.

More emphasis should be put on making a better use of wind observations over the oceans.



-0.50

0

0

0 0

0

0

0

0

0

00

0

0

0 0

0

0

0

41001

41004

410064100941010

41018

42002 4200342019

42020

42035 420364203942040

44004

4400544007

44009

44011

44014

44025

44140

46001

46002

4600346004

46005

46006

4601346014

46023

46030

4603646041

46042

46050

46059

46132

4614546184

46204

4620646207

46208

51001

5100251003

51004

51026

10°N 10°N

15°N15°N

20°N 20°N

25°N25°N

30°N 30°N

35°N35°N

40°N 40°N

45°N45°N

50°N 50°N

55°N55°N

60°N 60°N160°W

160°W 155°W

155°W 150°W

150°W 145°W

145°W 140°W

140°W 135°W

135°W 130°W

130°W 125°W

125°W 120°W

120°W 115°W

115°W 110°W

110°W 105°W

105°W 100°W

100°W 95°W

95°W 90°W

90°W 85°W

85°W 80°W

80°W 75°W

75°W 70°W

70°W 65°W

65°W 60°W

60°W 55°W

55°W 50°W

50°W

from 19960201 to 19960229Mean analysed minus forecast 10m wind speed for expver:18

-1

-0.5

-0.5

-0.5

0

0

0

0

00

0

0

0

0

0

0

0

0

0

0

0

0.5

0.5

0.5

0.5

1

62026

62081

62103

62105

62106

62107

62108

62109

62112

62162

62163

62165

6230362304

62305

63103

63111

63113

64045

LF3F

LF3J

LF3N

LF4C

LF5T

LF5U

LDWR

40°N 40°N

45°N45°N

50°N 50°N

55°N55°N

60°N 60°N

65°N65°N

70°N 70°N

25°W

25°W 20°W

20°W 15°W

15°W 10°W

10°W 5°W

5°W 0°

0° 5°E

5°E 10°E

10°E

from 19960201 to 19960229Mean analysed minus forecast 10m wind speed for expver:18

During the ERA-40 production, thefollowing mean analysis incrementsin surface winds were found.For a large part, they collocated withlocations where meteorological buoyswere deployed !

NB: maps of mean analysis increments are powerful tools



0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

00

0

0

0

0

0

0

0

0.1

0.1

41001

41004

410064100941010

42002 4200342019

42020

42035 420364203942040

44004

4400544007

44009

44011

44014

44025

46001

46002

4600346004

46005

46006

4601346014

46023

46030

4603646041

46042

46050

46059

46132

4614546184

46204

46206

4620746208

51001

5100251003

51004

5102620°N 20°N

25°N25°N

30°N 30°N

35°N35°N

40°N 40°N

45°N45°N

50°N 50°N

55°N55°N

160°W

160°W 155°W

155°W 150°W

150°W 145°W

145°W 140°W

140°W 135°W

135°W 130°W

130°W 125°W

125°W 120°W

120°W 115°W

115°W 110°W

110°W 105°W

105°W 100°W

100°W 95°W

95°W 90°W

90°W 85°W

85°W 80°W

80°W 75°W

75°W 70°W

70°W 65°W

65°W 60°W

60°W 55°W

55°Wfrom 19960201 to 19960228

Mean analysis difference in 10m wind speed: 314 - 315

0

0

0

0

0

0

0

0

0

0

0 0

0

00

0

0

0

0

0

0

0

0

0.1

0.1

0.1

0.2

62026

62081

62103

62105

62106

62107

62108

62109

62112

62162

62163

62165

62303 6230462305

63103

63111

63113

64045

LF3F

LF3J

LF3N

LF4C

LF5T

LF5U

LDWR

40°N 40°N

45°N45°N

50°N 50°N

55°N55°N

60°N 60°N

65°N65°N

70°N 70°N

35°W

35°W 30°W

30°W 25°W

25°W 20°W

20°W 15°W

15°W 10°W

10°W 5°W

5°W 0°

0° 5°E

5°E 10°E

10°E 15°E

15°Efrom 19960201 to 19960228

Mean analysis difference in 10m wind speed: 314 - 315

Further analysis, indicated that thescheme used to provide the dataanalysis with the actual height of theanemometer did not work !

Most buoys observes winds at height ~4-5 m. By default the analysis assumes a height of 10m(25m for ships).Once corrected, the expected impact was visible (ERA-40 was modified ~1997).Is it the end of it ?



NHOV 00

BL

NTSG 00

BL

CHLV2

BL

CLKN7

BL

FBIS1BL

IOSN3

BL

DSLN7

BL

TPLM2

BL

FPSN7

BL

IOSN3

BL

CHLV2

BL

CLKN7

BL

FBIS1BL

ALSN6

BL

DBLN6BL

KIBK

TPLM2

BL

IOSN3

BL

CHLV2

BL

CLKN7

BL

FBIS1BL

ALSN6

BL

DBLN6BL

DSLN7

BL

TPLM2

BL

FPSN7

BL

IOSN3

BL

44013

BL

44011

44014

BL

44004

41001

44025

BL

41004

BL

44009

BL

41002

44005

44004

41001

44025

BL

41004

BL

44009

BL

41002

44005

44004

41001

44011

44009

BL

44014

BL

CG268

BL

DQFV

BL

ZCAM9

BL

44025

BL

41004

BL

44009

BL

41002

44005

44014

BL

44025

BL

41004

BL

44011

44014

BL

44004

41001

NJCS 00

BL

DSLN7

BL

TPLM2

BL

FPSN7

BL

ALSN6

BL

DBLN6BL

CHLV2

BL

CLKN7

BL

FBIS1BL

ALSN6

BL

DBLN6BL

DSLN7

BL

TPLM2

BL

FPSN7

BL

IOSN3

BL

DSLN7

BL

FPSN7

BL

FBIS1BL

ALSN6

BL

DBLN6BL

DSLN7

BL

TPLM2

BL

FPSN7

BL

IOSN3

BL

CHLV2

BL

CLKN7

BL

FBIS1BL

ALSN6

BL

DBLN6BL

44025

BL

41004

BL

41002

44005

44009

BL

44013

BL

44011

44014

BL

44013

BL

44011

44014

BL

44004

41001

44013

BL

44011

44004

41001

44013

BL

41002

44005

44009

BL

41001

44014

BL

41004

BL

4401144028

BL

41001

44014

BL

41004

BL

4401144028

BL

4400444009

BL

41002

4401144028

BL

41001

44014

BL

41004

BL

44005

BL

44013

BL

44025

BL

41001

44014

BL

41004

BL

44005

BL

44013

BL

44025

BL

41001

44014

BL

41004

BL

44005

BL

44013

BL

44025

BL

4400444009

BL

41002

44005

BL

44013

BL

44025

BL

4400444009

BL

41002

44005

BL

44013

BL

44025

BL

41001

44014

BL

41004

BL

44005

BL

44013

BL

44025

BL

4400444009

BL

41002

4401144028

BL

4400444009

BL

41002

4401144028

BL

4400444009

BL

41002

4401144028

BL

40°N 40°N

for 19960301 0 UTCERA40 SHIP and DRIBU observations

SYNOP: 114 DRIBU: 0 AIREP: 0 SATOB: 0 TEMP: 0 PILOT: 0 SATEM: 0 CORRECT 29

POSS ERR 0 PROB ERR 0 ERROR: 85 SYNOP: 0 DRIBU: 66 AIREP: 0 SATOB: 0 TEMP: 0

It was not realised then that some buoy data are available both as SHIP and DRIBU, because of the blend of ECMWF (SHIP) and NCEP (DRIBU) data sets.

The anemometer height scheme only worked on SHIP data !

It now works for DRIBU dataas well (since end of August 2005 in operations) and in the interim reruns (finally)!



9N140W9N140W9N140W

8S180W8S180W8S180W8S165E8S165E8S165E 8S155W8S155W8S155W 8S125W8S125W8S125W 8S110W8S110W8S110W

8N170W8N170W8N170W8N165E

FG

8N165E

FG

8N165E 8N125W8N125W8N125W 8N110W8N110W8N110W

5S180W5S180W5S180W5S156E

BL

5S156E

BL

5S156E

BL

5S155W

BL

5S155W

BL

5S155W

BL

5S140W5S140W5S140W 5S110W5S110W

5S110W

5N180W5N180W5N180W 5N170W5N170W5N170W5N165E5N165E5N165E5N155W5N155W5N155W

5N140W5N140W5N140W 5N125W5N125W5N125W 5N110W5N110W5N110W2N180W2N180W2N180W 2N170W2N170W2N170W2N165E2N165E2N165E 2N155W2N155W2N155W2N147E2N147E2N147E 2N125W2N125W2N125W

2S95W2S95W2S95W2S180W2S180W2S180W 2S170W

BL

2S170W

BL

2S170W

BL

2S155W2S155W2S155W 2S140W2S140W2S140W 2S125W2S125W2S125W 2S110W2S110W2S110W0N180W0N180W0N180W 0N170W0N170W0N170W0N165E0N165E0N165E0N156E0N156E0N156E 0N155W0N155W0N155W0N143E

BL

0N143E

BL

0N143E

BL

0N140W0N140W0N140W 0N125W0N125W0N125W 0N110W0N110W0N110W2256

4DDEEK

91338

DZOD

BL

JADC

BL91222

BL

91352

DVRF

BL

CPCL

BL

121691338

91251

BL

SANF1

BL

91222

BL

1136

91377

BL

9132891251

BL91222

BL

FNOM

BLYJZH9

C6IC

BL

1466

KTDQ1225

NCJE

29891395DUMF

NCDWDZOD

P3XM4ELNG9JRKN

JADCKRJFKGXA

ELIG8BL6ZXGBL

DVXENCVVWJLX

9134391355

BL

1217

CBGR

BL

91377

BL

91338

VVGC

9134391355

BL

2976

MYRF

91377

BL

91338

91251

BL91222

BL

5100251004

SKFQJJRQ

BL

DVRF

BL

51002510045100251004

VRGW

BL

PGDX

VNGK

BL

C6KV2BL

5100251004

C6HL8DNKL

ELKD6

4XGVVPIB

C6KF2

UHVW

JJRQ

BL

8JPX

PJNL

C6HK9

BL

C6IC

C6MD6

WNJG

BL

PFPT

BL

GCCCPGLA

ELJS5

BLOYMO2

CPCL

BL

HRRF

BL

5100251004

51001BL

52009

BL

9N140W9N140W9N140W

8S180W8S180W8S180W8S165E8S165E8S165E 8S155W8S155W8S155W 8S125W8S125W8S125W 8S110W8S110W8S110W

8N170W8N170W8N170W8N165E

FG

8N165E8N165E 8N125W8N125W8N125W 8N110W8N110W8N110W

5S180W5S180W5S180W5S156E

BL

5S156E

BL

5S156E

BL

5S155W

BL

5S155W

BL

5S155W

BL

5S140W5S140W5S140W 5S110W5S110W5S110W

5N180W5N180W5N180W 5N170W5N170W5N170W5N165E5N165E5N165E5N155W5N155W5N155W

5N140W5N140W5N140W 5N125W5N125W5N125W 5N110W5N110W5N110W2N180W

FG

2N180W

FG

2N180W 2N170W2N170W2N170W2N165E2N165E2N165E 2N155W2N155W2N155W2N147E2N147E2N147E 2N125W2N125W2N125W

2S95W2S95W2S95W2S180W2S180W2S180W 2S170W

BL

2S170W

BL

2S170W

BL

2S155W2S155W2S155W 2S140W2S140W2S140W 2S125W2S125W2S125W 2S110W2S110W2S110W0N180W0N180W0N180W 0N170W0N170W0N170W0N165E0N165E0N165E0N156E0N156E0N156E 0N155W0N155W0N155W0N143E

BL

0N143E

BL

0N143E

BL

0N140W0N140W0N140W 0N125W0N125W0N125W 0N110W0N110W0N110W

139891377

BL

9132891251

BL

SANF1

BL

JRKN

9134391355

BL2837

CBGR

BL

91377

BL

91328

J8PD

BL

913439135291355

BL

2534

2931

2984

91338

1400

SANF1

BL

913439135291355

BL

1124

1227

FAREBL1218

C6KB6

2811ELEU

259191377

BL

9133891328

WFQE

BL

LACP4LACP4

91251

BL

NUOT

BL

ELLE2JCOY

DVCKBL3EPU4

NPFDBLNAON

91222

BL

91352

1124

DVRF

BL

CPCL

BL

9132891251

BL

SANF1

BL

91222

BL

91352

1223

2984

2837

HRRF

BL

91328

SANF1

BL

913439135291355

BL

32302

51026

BL

51003

51001BL

JPTFJCOY

JBTQDZOD

BL

CPCL

BL

51026BL

51003

51001BL 51026BL51001BL

VRSPBL

YJYY2

BL

VRSB

BL

VROC32302

V2EFC6KB4

51026BL51001BL

KCGH

VVGCJKPS

JPTF

JBTQ VRIY

WXWL

6YSA

ELMH5

ZHET5

BLZCKUBL

SHIP

BL

DVRF

BL

MYRF

51026BL51001BL

5100251004

32302

51026

BL

5100451003

51001BL

3232232323

3231851308

51009

51302

51010

52313

52306

BL

51011

51307

51311

51006

51023

51303

513055231152321

51002

3232232323

3231851308

51009

51302

51010

5231352007

52010

BL52306

BL 3231543001

51015

51016

5100751020

51021

51309

5231052311

52006

FG

52003

32302

51026BL

51002

32319

32317510175101451019

BL52306

BL 3231543001

51015

51016

5100751020

5231752301

32302

5100451003

51001BL

32322

32319

32317510175101451019

BL

51306

BL

52313

52306

BL

51011

51307

51311

51006

51023

51303

51305

52309

52006

FG

52003

52317

5100451003

51001BL

32319

3231751017

51302

5231352007

52010

BL 51011

51307

51311

5130352309

5232152001

32302

51026BL

5100451003

51001BL

3232351010

5231352007

52010

BL52306

BL 32315

51311

51303

51305

52309

5232152001

52009

BL

52007

52301

5200352008FG

52313

5131052316

51022

52007

52525

52310

FG

5130551023

5200352006

FG

51302

510095101051306

BL

5101451022 51009

5101451019

BL

3231932318

51308

51006

51007

5130551023

51006

5102351016

43001

51002

52525

32319

32317510175101451019

BL

5102251306

BL

52312

52316

3231543001

51015

51016

5100751020

51021

5130952309

52317

32302

51026

BL

5100451003

51001BL

32319

32317510175101451019

BL

5102251306

BL

52312

52316

51011

51307

51311

51006

51023

51303

51305

52309

5232152001

5231752301

52525

5100451003

51001BL

3232232323

3231851308

51009

51302

52010

BL 51011

51307

51311

51006

5102152321

51026BL

51002

3232332318

51308

510095102251010

523165200752010

BL

3231543001

51015

51016

5100751020

51021

51309

52310523215200152301

32302

51026BL

51002

3232232323

3231851308

52312

52316

52306

BL 3231543001

51015

51016

51020

5231052311

5200652003

5231752301

51002

32322

51302

51306

BL

52312

52316

51011

5102051309

5231052311

5200652003

5231752301

52012

BL

52010

BL 52310

FG

52011 52001

52006

FG

523125101051306

BL51303

52311

5230951309

52001

5231252311

51019

BL

51310

510185101851022 51017 32317

51021

5130951007

51021

5130751015

10°S10°S

0° 0°

10°N10°N

20°N 20°N

140°E

140°E 160°E

160°E 180°

180° 160°W

160°W 140°W

140°W 120°W

120°W 100°W

100°W

for 19940301 18 UTCERA40 SHIP and DRIBU observations

SYNOP: 468 DRIBU: 0 AIREP: 0 SATOB: 0 TEMP: 0 PILOT: 0 SATEM: 0 CORRECT 357POSS ERR 5 PROB ERR 0 ERROR: 106 SYNOP: 0 DRIBU: 311 AIREP: 0 SATOB: 0 TEMP: 0

PILOT: 0 SATEM: 0 CORRECT 266 POSS ERR 5 PROB ERR 2 ERROR: 38

The anemometer height scheme relies on a list connecting the station identifier (buoys, ship and platforms) with the height of the anemometer. It has been built by gathering information fromdifferent sources. It will requires more effort for older data sets.

e.g. earlier TAOdata both appear as SHIP and DRBU but with different station identifier:e.g.51014 <=> 5S140W



Jun-

1992

Jun-

1993

Jun-

1994

Jun-

1995

Jun-

1996

Jun-

1997

Jun-

1998

Jun-

1999

Jun-

2000

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

months

0

5000

10000

15000

20000

Num

ber o

f colloca

tions

with

mod

el

ERA-40 (e4 0001)ops (od 0001)

Comparison with TAO buoys as obtained by GTS3 month running average

Number of data

Limited past GTS dissimination of certain data should be complemented by off-line data as they are very useful data.e.g. TAO data for more recentyears before the upgrade in earlier 2005.

Jun-

1992

Jun-

1993

Jun-

1994

Jun-

1995

Jun-

1996

Jun-

1997

Jun-

1998

Jun-

1999

Jun-

2000

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

months

-1.5

-1.25

-1

-0.75

-0.5

-0.25

0

Win

d Sp

eed

Bias

(mod

el-b

uoy)

(m/s

)ERA-40 (e4 0001)ops (od 0001)


bias

1992 2006

Jun-

1992

Jun-

1993

Jun-

1994

Jun-

1995

Jun-

1996

Jun-

1997

Jun-

1998

Jun-

1999

Jun-

2000

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

months

10

15

20

25

30

Win

d Sp

eed

Scat

ter I

ndex

(%)

ERA-40 (e4 0001)ops (od 0001)


S.I.

1992 2006



There is a need to sort out in-situ observations over the oceans to avoid unnecessary duplications.

It is unfortunate that potentially very valuable data from moored buoys are mixed with other type of data of the same nature (e.g.with ship and fixed platforms (SHIP) or with drifting buoys (DRIBU)). This is reflected by the relatively large error assigned to these data. A separate data type should be created (moored buoys) and bit of data mining should take place to bring the data to ECMWF.


Conclusions :

Wind and waves should be better analysed in the interim analysis following improvements in both atmospheric and wave models.

A more consistent altimeter wave height data base has been created with the acquisition of OPR ERS data and the homogenization of the data following bias estimation from a triple collocated data set.

A bit more work (coordination) on a surface wind data base could be beneficial to any future reanalysis.

bidlot ocean wave analysis - ecmwf · reanalysis workshop slide 3 background : zecmwf has been...

Documents