Evaluating processes controlling subtropical humidity

Evaluating processes controlling subtropical

humidity using water vapor isotope measurements

Camille Risi

CIRES, Boulder

Thanks to: David Noone, Sandrine Bony,

TES data: John Worden, Jeonghoon Lee, Derek Brown,

SCIAMACHY data: Christian Frankenberg,

ACE-FTS data: Kaley Walker, Peter Bernath,

ground-based FTIR: Debra Wunsh, Matthias Schneider

TES meeting, 17 June 2010

Uncertainties in humidity changes

I subtropical relative humidity strongly impacts

I water vapor feedback (Soden et al 2008)

I clouds feedbacks (Sherwood et al 2010)

I but high dispersion in climate models (Sherwood et al 2010)

I for present day relative humidityI for projectionsI no link between mean state and projections (Santer et al 2009)

I How to evaluate the credibility of these projections?

2/13

Uncertainties in humidity changes

I subtropical relative humidity strongly impacts

I water vapor feedback (Soden et al 2008)I clouds feedbacks (Sherwood et al 2010)

I but high dispersion in climate models (Sherwood et al 2010)

I for present day relative humidityI for projectionsI no link between mean state and projections (Santer et al 2009)

I How to evaluate the credibility of these projections?

2/13

Uncertainties in humidity changes

I subtropical relative humidity strongly impacts

I water vapor feedback (Soden et al 2008)I clouds feedbacks (Sherwood et al 2010)

I but high dispersion in climate models (Sherwood et al 2010)

I for present day relative humidityI for projections

I no link between mean state and projections (Santer et al 2009)

I How to evaluate the credibility of these projections?

2/13

Uncertainties in humidity changes

I subtropical relative humidity strongly impacts

I water vapor feedback (Soden et al 2008)I clouds feedbacks (Sherwood et al 2010)

I but high dispersion in climate models (Sherwood et al 2010)

I for present day relative humidityI for projectionsI no link between mean state and projections (Santer et al 2009)

I How to evaluate the credibility of these projections?

2/13

Uncertainties in humidity changes

I subtropical relative humidity strongly impacts

I water vapor feedback (Soden et al 2008)I clouds feedbacks (Sherwood et al 2010)

I but high dispersion in climate models (Sherwood et al 2010)

I for present day relative humidityI for projectionsI no link between mean state and projections (Santer et al 2009)

I How to evaluate the credibility of these projections?

2/13

Processes controlling relative humidity

I dispersion re�ects complexity of humidity controls

800

600

200

100

P (hPa)

0°

400

100030°N

⇒need complementary evaluation tools

3/13

Processes controlling relative humidity

I dispersion re�ects complexity of humidity controls

800

600

200

100

P (hPa)

0°

400

100030°N

large-s ale subsiden eSherwood et al 1996

⇒need complementary evaluation tools

3/13

Processes controlling relative humidity

I dispersion re�ects complexity of humidity controls

800

600

200

100

P (hPa)

0°

400

100030°N

large-s ale subsiden edry air intrusionsSherwood et al 1996Pierrehumbert and Ro a 98Galewsky et al 2005

⇒need complementary evaluation tools

3/13

Processes controlling relative humidity

I dispersion re�ects complexity of humidity controls

800

600

200

100

P (hPa)

0°

400

100030°N

large-s ale subsiden edry air intrusionslateral adve tionSherwood et al 1996Pierrehumbert and Ro a 98Galewsky et al 2005Pierrehumbert 1998

⇒need complementary evaluation tools

3/13

Processes controlling relative humidity

I dispersion re�ects complexity of humidity controls

800

600

200

100

P (hPa)

0°

400

100030°N

large-s ale subsiden edry air intrusionslateral adve tionsurfa e evaporationboundary layerSherwood et al 1996Pierrehumbert and Ro a 98Galewsky et al 2005Pierrehumbert 1998Couhert et al 2010

⇒need complementary evaluation tools

3/13

Processes controlling relative humidity

I dispersion re�ects complexity of humidity controls

800

600

200

100

P (hPa)

0°

400

100030°N

large-s ale subsiden edry air intrusionslateral adve tionsurfa e evaporationboundary layerrain evaporation

Sherwood et al 1996Pierrehumbert and Ro a 98Galewsky et al 2005Pierrehumbert 1998Couhert et al 2010Folkins and Martin 2005

⇒need complementary evaluation tools

3/13

Processes controlling relative humidity

I dispersion re�ects complexity of humidity controls

800

600

200

100

P (hPa)

0°

400

100030°N

large-s ale subsiden edry air intrusionslateral adve tionsurfa e evaporationboundary layerrain evaporation onve tive subsiden e

Sherwood et al 1996Pierrehumbert and Ro a 98Galewsky et al 2005Pierrehumbert 1998Couhert et al 2010Bretherton et al 2005Folkins and Martin 2005

⇒need complementary evaluation tools

3/13

Processes controlling relative humidity

I dispersion re�ects complexity of humidity controls

800

600

200

100

P (hPa)

0°

400

100030°N

condensate detrainementPierrehumbert and Emanuel 1994 large-s ale subsiden edry air intrusionslateral adve tionsurfa e evaporationboundary layerrain evaporation onve tive subsiden e

Wright et al 2009 Sherwood et al 1996Pierrehumbert and Ro a 98Galewsky et al 2005Pierrehumbert 1998Couhert et al 2010Bretherton et al 2005Folkins and Martin 2005

⇒need complementary evaluation tools

3/13

Processes controlling relative humidity

I dispersion re�ects complexity of humidity controls

800

600

200

100

P (hPa)

0°

400

100030°N

condensate detrainementPierrehumbert and Emanuel 1994 large-s ale subsiden edry air intrusionslateral adve tionsurfa e evaporationboundary layerrain evaporation onve tive subsiden e

Wright et al 2009 Sherwood et al 1996Pierrehumbert and Ro a 98Galewsky et al 2005Pierrehumbert 1998Couhert et al 2010Bretherton et al 2005Folkins and Martin 2005

⇒need complementary evaluation tools3/13

Water stable isotopes

I water isotopes: H162 O, H18

2 O,HDO

I fractionation

⇒ record phase changes along

air mass history

HDO

H162 O

⇒ Goal: use water isotopes to evaluate processes controlling

humidity in climate models

4/13

Water stable isotopes

I water isotopes: H162 O, H18

2 O,HDO

I fractionation ⇒ record phase changes along

air mass historyHDO

H162 O

⇒ Goal: use water isotopes to evaluate processes controlling

humidity in climate models

4/13

Water stable isotopes

I water isotopes: H162 O, H18

2 O,HDO

I fractionation ⇒ record phase changes along

air mass historyHDO

H162 O

800

600

200

100

P (hPa)

0°

400

100030°N

⇒ Goal: use water isotopes to evaluate processes controlling

humidity in climate models

4/13

Water stable isotopes

I water isotopes: H162 O, H18

2 O,HDO

I fractionation ⇒ record phase changes along

air mass historyHDO

H162 O

800

600

200

100

P (hPa)

0°

400

100030°N

onve tive subsiden eRisi et al 2008

⇒ Goal: use water isotopes to evaluate processes controlling

humidity in climate models

4/13

Water stable isotopes

I water isotopes: H162 O, H18

2 O,HDO

I fractionation ⇒ record phase changes along

air mass historyHDO

H162 O

800

600

200

100

P (hPa)

0°

400

100030°N

rain evaporation onve tive subsiden eWorden et al 2007Risi et al 2008

⇒ Goal: use water isotopes to evaluate processes controlling

humidity in climate models

4/13

Water stable isotopes

I water isotopes: H162 O, H18

2 O,HDO

I fractionation ⇒ record phase changes along

air mass historyHDO

H162 O

800

600

200

100

P (hPa)

0°

400

100030°N

ondensate detrainementrain evaporation onve tive subsiden eWorden et al 2007Risi et al 2008

Webster and Heims�led 2003

⇒ Goal: use water isotopes to evaluate processes controlling

humidity in climate models

4/13

Water stable isotopes

I water isotopes: H162 O, H18

2 O,HDO

I fractionation ⇒ record phase changes along

air mass historyHDO

H162 O

800

600

200

100

P (hPa)

0°

400

100030°N

ondensate detrainementlarge-s ale subsiden erain evaporation onve tive subsiden eWorden et al 2007Risi et al 2008Frankenberg et al 1999Webster and Heims�led 2003

⇒ Goal: use water isotopes to evaluate processes controlling

humidity in climate models

4/13

Water stable isotopes

I water isotopes: H162 O, H18

2 O,HDO

I fractionation ⇒ record phase changes along

air mass historyHDO

H162 O

800

600

200

100

P (hPa)

0°

400

100030°N

ondensate detrainementlarge-s ale subsiden erain evaporation onve tive subsiden eWorden et al 2007Risi et al 2008Frankenberg et al 1999Webster and Heims�led 2003

⇒ Goal: use water isotopes to evaluate processes controlling

humidity in climate models 4/13

Isotope dispersion in climate modelsI isotope-enabled GCMs intercomparison (SWING2)

5/13

Isotope dispersion in climate modelsI isotope-enabled GCMs intercomparison (SWING2)

P (

hPa)

nudged

free

−600 −400 −300 −200 −100−500

ECHAM

MIROC

CAM2

LMDZ AR4

LMDZ SWING

HadAM

GSM

1000

900

800

700

600

500

400

300

200

100Subtropi al mean 10◦N-30◦N

δD (h) annual5/13

Isotope dispersion in climate modelsI isotope-enabled GCMs intercomparison (SWING2)

20 25 30 35 40 45 50

−40

−20

0

20

40

60

P (

hPa)

nudged

free

−600 −400 −300 −200 −100−500

ECHAM

MIROC

CAM2

LMDZ AR4

LMDZ SWING

HadAM

GSM

1000

900

800

700

600

500

400

300

200

100Subtropi al mean 10◦N-30◦N

JJA-DJFδD

at450hPa(h)

δD (h) annual

NCEPrelative humidity at 450hPa (%)

5/13

Isotope dispersion in climate modelsI isotope-enabled GCMs intercomparison (SWING2)

20 25 30 35 40 45 50

−40

−20

0

20

40

60

P (

hPa)

LMDZ SWING

LMDZ AR4

nudged

free

−600 −400 −300 −200 −100−500

ECHAM

MIROC

CAM2

LMDZ AR4

LMDZ SWING

HadAM

GSM

1000

900

800

700

600

500

400

300

200

100Subtropi al mean 10◦N-30◦N

JJA-DJFδD

at450hPa(h)

δD (h) annual

NCEPrelative humidity at 450hPa (%)LMDZ-iso (Risi et al in press)

5/13

Isotope dispersion in climate modelsI isotope-enabled GCMs intercomparison (SWING2)

20 25 30 35 40 45 50

−40

−20

0

20

40

60

P (

hPa)

LMDZ SWING

LMDZ AR4

nudged

free

−600 −400 −300 −200 −100−500

ECHAM

MIROC

CAM2

LMDZ AR4

LMDZ SWING

HadAM

GSM

1000

900

800

700

600

500

400

300

200

100Subtropi al mean 10◦N-30◦N

JJA-DJFδD

at450hPa(h)

δD (h) annualwithsaturation limiterNCEP

relative humidity at 450hPa (%)(Codron and Sadourny 2002)adve t min(q,qs)LMDZ-iso (Risi et al in press)5/13

Isotope dispersion in climate modelsI isotope-enabled GCMs intercomparison (SWING2)

20 25 30 35 40 45 50

−40

−20

0

20

40

60

P (

hPa)

LMDZ SWING

LMDZ AR4

nudged

free

−600 −400 −300 −200 −100−500

ECHAM

MIROC

CAM2

LMDZ AR4

LMDZ SWING

HadAM

GSM

1000

900

800

700

600

500

400

300

200

100Subtropi al mean 10◦N-30◦N

JJA-DJFδD

at450hPa(h)

δD (h) annualwithsaturation limiterNCEP

relative humidity at 450hPa (%)withoutsaturationlimiteradve t q(Codron and Sadourny 2002)adve t min(q,qs)LMDZ-iso (Risi et al in press)5/13

Data for 3D isotope evaluation

800

600

200

100

P (hPa)

0° 30°N

400

1000

I quality screeningI GCM-data comparison: collocated with simulation nudged by

ECMWF; averaging kernels

6/13

Data for 3D isotope evaluation

TES(Worden et al 2007)

800

600

200

100

P (hPa)

0° 30°N

400

1000

I quality screeningI GCM-data comparison: collocated with simulation nudged by

ECMWF; averaging kernels

6/13

Data for 3D isotope evaluation

ACE−FTS

TES(Worden et al 2007)

(Nassar et al 2007)

800

600

200

100

P (hPa)

0° 30°N

400

1000

I quality screeningI GCM-data comparison: collocated with simulation nudged by

ECMWF; averaging kernels

6/13

Data for 3D isotope evaluation

ACE−FTS

TES(Worden et al 2007)

(Nassar et al 2007)

SCIAMACHY(Frankenberg et al 2009)

800

600

200

100

P (hPa)

0° 30°N

400

1000

I quality screeningI GCM-data comparison: collocated with simulation nudged by

ECMWF; averaging kernels

6/13

Data for 3D isotope evaluation

ACE−FTS

TES(Worden et al 2007)

(Nassar et al 2007)

SCIAMACHY(Frankenberg et al 2009)

(Wisconsin, Oklahoma)+ground−based FTIR

800

600

200

100

P (hPa)

0° 30°N

400

1000

I quality screeningI GCM-data comparison: collocated with simulation nudged by

ECMWF; averaging kernels

6/13

Data for 3D isotope evaluation

ACE−FTS

TES(Worden et al 2007)

(Nassar et al 2007)

in situ(GNIP, Angert et al 2007, Uemura et al 2007)

SCIAMACHY(Frankenberg et al 2009)

(Wisconsin, Oklahoma)+ground−based FTIR

800

600

200

100

P (hPa)

0° 30°N

400

1000

I quality screeningI GCM-data comparison: collocated with simulation nudged by

ECMWF; averaging kernels

6/13

Data for 3D isotope evaluation

ACE−FTS

TES

ground−based FTIR profile

(Worden et al 2007)

(Nassar et al 2007)

in situ(GNIP, Angert et al 2007, Uemura et al 2007)

(Izana:Schneider et al 2009)

SCIAMACHY(Frankenberg et al 2009)

(Wisconsin, Oklahoma)+ground−based FTIR

800

600

200

100

P (hPa)

0° 30°N

400

1000

I quality screeningI GCM-data comparison: collocated with simulation nudged by

ECMWF; averaging kernels

6/13

Data for 3D isotope evaluation

ACE−FTS

TES

ground−based FTIR profile

(Worden et al 2007)

(Nassar et al 2007)

in situ(GNIP, Angert et al 2007, Uemura et al 2007)

(Izana:Schneider et al 2009)

SCIAMACHY(Frankenberg et al 2009)

(Wisconsin, Oklahoma)+ground−based FTIR

800

600

200

100

P (hPa)

0° 30°N

400

1000

I quality screening

I GCM-data comparison: collocated with simulation nudged by

ECMWF; averaging kernels

6/13

Data for 3D isotope evaluation

ACE−FTS

TES

ground−based FTIR profile

(Worden et al 2007)

(Nassar et al 2007)

in situ(GNIP, Angert et al 2007, Uemura et al 2007)

(Izana:Schneider et al 2009)

SCIAMACHY(Frankenberg et al 2009)

(Wisconsin, Oklahoma)+ground−based FTIR

800

600

200

100

P (hPa)

0° 30°N

400

1000

I quality screeningI GCM-data comparison: collocated with simulation nudged by

ECMWF; averaging kernels 6/13

Multidataset evaluation: annual mean

Eq 30N60S 60N30S

Eq 30N60S 60N30S

Eq 30N60S 60N30S−100

−200

−300

−300

−200

−100

−200

−100

0

−400

−500

−600

−300

Eq 30N60S 60N30S

Eq 30N60S 60N30S−700

−400

−500

−600

LMDZ−AR4data

δD

(h)δD

(h)

δD

(h)Surfa e

Total olumn600 hPa

Zonal mean, annual average

δD

(h)

δD

(h)500-300hPa

300-200hPa7/13

Multidataset evaluation: annual mean

Eq 30N60S 60N30S

Eq 30N60S 60N30S

Eq 30N60S 60N30S−100

−200

−300

−300

−200

−100

−400

−500

−600

−300

Eq 30N60S 60N30S

Eq 30N60S 60N30S−700

−400

−500

−600

LMDZ−"moist bias"LMDZ−AR4data

−200

−100

0

Surfa eTotal olumn600 hPa

Zonal mean, annual averageδD

(h)δD

(h)

δD

(h)δD

(h)

δD

(h)500-300hPa

300-200hPa7/13

Multidataset evaluation: seasonal

Eq 30N60S 60N30S

Eq 30N60S 60N30S

−100

0

100

−100

0

100

−100

0

100

−100

0

100

−100

100

0

Eq 30N60S 60N30S Eq 30N60S 60N30S

Eq 30N60S 60N30S∆

δD

(h)∆

δD

(h)∆

δD

(h)

∆δD

(h)

∆δD

(h)

LMDZ-AR4data600 hPa

Total olumn and 800hPaSurfa e

300-200hPa500-300hPa

Zonal mean, JJA-DJF

8/13

Multidataset evaluation: seasonal

Eq 30N60S 60N30S

Eq 30N60S 60N30S

−100

0

100

−100

0

100

−100

0

100

−100

0

100

−100

100

0

Eq 30N60S 60N30S Eq 30N60S 60N30S

Eq 30N60S 60N30S∆

δD

(h)∆

δD

(h)∆

δD

(h)

∆δD

(h)

∆δD

(h)

LMDZ-AR4dataLMDZ-"moist bias"600 hPa

Total olumn and 800hPaSurfa e

300-200hPa500-300hPa

Zonal mean, JJA-DJF

8/13

Annual mean in TES

30S

Eq

60N

30N

60S

120W180W 060W 60E 120E 180E

120W180W 060W 60E 120E 180E

−230 −220 −210 −200 −190 −180 −170 −160 −150

30S

Eq

60N

30N

60S

TES data

LMDZ AR4 -31h

δD (h) 600hPa9/13

Annual mean in TES

120W180W 060W 60E 120E 180E

30S

Eq

60N

30N

60S

30S

Eq

60N

30N

60S

120W180W 060W 60E 120E 180E

120W180W 060W 60E 120E 180E

−230 −220 −210 −200 −190 −180 −170 −160 −150

30S

Eq

60N

30N

60S

same in CAM2(Lee et al 2009)

TES data

LMDZ AR4 -31h LMDZ "moist bias" -31h

δD (h) 600hPa9/13

Annual mean in TES

120W180W 060W 60E 120E 180E

30S

Eq

60N

30N

60S

30S

Eq

60N

30N

60S

120W180W 060W 60E 120E 180E

120W180W 060W 60E 120E 180E

−230 −220 −210 −200 −190 −180 −170 −160 −150

30S

Eq

60N

30N

60S

same in CAM2(Lee et al 2009)

TES data

LMDZ AR4 -31h LMDZ "moist bias" -31h

δD (h) 600hPa9/13

Seasonal variations in TES

−80 −50 −30 −20 −10 10 20 30 50 80

120W180W 060W 60E 120E 180E

30S

Eq

60N

30N

60S120W180W 060W 60E 120E 180E

30S

Eq

60N

30N

60S

120W180W 060W 60E 120E 180E

30S

Eq

60N

30N

60S

∆δD (h) JJA-DJF

LMDZ AR4 LMDZ "moist bias"

TES data

10/13

Cloud e�ect in TES

−50 −20 −10 −5 −2 5 10 20 502

120W180W 060W 60E 120E 180E

30S

Eq

60N

30N

60S120W180W 060W 60E 120E 180E

30S

Eq

60N

30N

60S

120W180W 060W 60E 120E 180E

30S

Eq

60N

30N

60S

∆δD (h) total- lear sky

TES data

LMDZ AR4 LMDZ "moist bias"

11/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

LMDZAR4

bias""moistLMDZ

− weak precip compositeStrong precip composite

∆δD (h)

∆δD (h)

12/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

LMDZAR4

bias""moistLMDZ

− weak precip compositeStrong precip composite

∆δD (h)

∆δD (h)

12/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

− weak precip compositeStrong precip composite

LMDZAR4

bias""moistLMDZ

saturationlimiter

∆δD (h)

∆δD (h)

12/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

− weak precip compositeStrong precip composite

LMDZAR4

bias""moistLMDZ

saturationlimiter

dry tropics

dry,depleted subtropics

∆δD (h)

∆δD (h)

12/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

− weak precip compositeStrong precip composite

LMDZAR4

bias""moistLMDZ

saturationlimiter

dry tropics

dry,depleted subtropics

∆δD (h)

∆δD (h)

12/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

− weak precip compositeStrong precip composite

LMDZAR4

bias""moistLMDZ

saturationlimiter

dry tropics

dry,depleted subtropicsmore enriched in summer

∆δD (h)

∆δD (h)

12/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

− weak precip compositeStrong precip composite

LMDZAR4

bias""moistLMDZ

saturationlimiter

dry tropics

dry,depleted subtropicsmore enriched in summer

∆δD (h)

∆δD (h)

12/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

− weak precip compositeStrong precip composite

LMDZAR4

bias""moistLMDZ

saturationlimiter

dry tropics

dry,depleted subtropics

moist, enriched subtropics

moist tropics

more enriched in summer

∆δD (h)

∆δD (h)

12/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

− weak precip compositeStrong precip composite

LMDZAR4

bias""moistLMDZ

saturationlimiter

dry tropics

dry,depleted subtropics

moist, enriched subtropics

moist tropics

more enriched in summer

∆δD (h)

∆δD (h)

12/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

− weak precip compositeStrong precip composite

LMDZAR4

bias""moistLMDZ

saturationlimiter

dry tropics

dry,depleted subtropics

moist, enriched subtropics

moist tropics

more depleted in summer

more enriched in summer

∆δD (h)

∆δD (h)

12/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

− weak precip compositeStrong precip composite

LMDZAR4

bias""moistLMDZ

saturationlimiter

Annual mean 15°N−30°N

moist bias 25%

at 500hPa:

dry tropics

dry,depleted subtropics

moist, enriched subtropics

moist tropics

moist bias 15%

more depleted in summer

more enriched in summer

∆δD (h)

∆δD (h)

12/13

Isotopes reveal dehydration pathways

1000

900

700

800

600

500

400

200

100

300

1000

900

700

800

600

500

400

200

100

300

−10 0 10 20 30−20 40

−10 0 10 20 30 40−20

− weak precip compositeStrong precip composite

LMDZAR4

bias""moistLMDZ

saturationlimiter

Annual mean 15°N−30°N

moist bias 25%

at 500hPa:

dry tropics

dry,depleted subtropics

moist, enriched subtropics

moist tropics

moist bias 15%

more depleted in summer

more enriched in summer

∆δD (h)

∆δD (h)

+0.5% hange in 4xCO2

-4% hange in 4xCO2

12/13

Conclusion/perspectives

I Water isotopes can help detect and understand biases in water

budgets in climate models

⇒ help evaluate the credibility of projected future relative

humidity changes

I Perspectives:

I Observations: maximum variations and biases in themid/upper troposphere

I Model-data comparison methodology: de�ne consistentlycloudy situations in data and GCMs⇒develop GCM simulators of satellite data

I Modelling: water isotopes in climate change inter-comparisonslike CMIP?⇒ use intra-seasonal, seasonal, inter-annual isotope data as anobservational test for model behavior in climate change?

13/13

Conclusion/perspectives

I Water isotopes can help detect and understand biases in water

budgets in climate models

⇒ help evaluate the credibility of projected future relative

humidity changes

I Perspectives:

I Observations: maximum variations and biases in themid/upper troposphere

I Model-data comparison methodology: de�ne consistentlycloudy situations in data and GCMs⇒develop GCM simulators of satellite data

I Modelling: water isotopes in climate change inter-comparisonslike CMIP?⇒ use intra-seasonal, seasonal, inter-annual isotope data as anobservational test for model behavior in climate change?

13/13

Conclusion/perspectives

I Water isotopes can help detect and understand biases in water

budgets in climate models

⇒ help evaluate the credibility of projected future relative

humidity changes

I Perspectives:

I Observations: maximum variations and biases in themid/upper troposphere

I Model-data comparison methodology: de�ne consistentlycloudy situations in data and GCMs⇒develop GCM simulators of satellite data

I Modelling: water isotopes in climate change inter-comparisonslike CMIP?⇒ use intra-seasonal, seasonal, inter-annual isotope data as anobservational test for model behavior in climate change?

13/13

Conclusion/perspectives

I Water isotopes can help detect and understand biases in water

budgets in climate models

⇒ help evaluate the credibility of projected future relative

humidity changes

I Perspectives:

I Observations: maximum variations and biases in themid/upper troposphere

I Model-data comparison methodology: de�ne consistentlycloudy situations in data and GCMs⇒develop GCM simulators of satellite data

I Modelling: water isotopes in climate change inter-comparisonslike CMIP?⇒ use intra-seasonal, seasonal, inter-annual isotope data as anobservational test for model behavior in climate change?

13/13

Conclusion/perspectives

I Water isotopes can help detect and understand biases in water

budgets in climate models

⇒ help evaluate the credibility of projected future relative

humidity changes

I Perspectives:

I Observations: maximum variations and biases in themid/upper troposphere

I Model-data comparison methodology: de�ne consistentlycloudy situations in data and GCMs

⇒develop GCM simulators of satellite dataI Modelling: water isotopes in climate change inter-comparisons

like CMIP?⇒ use intra-seasonal, seasonal, inter-annual isotope data as anobservational test for model behavior in climate change?

13/13

Conclusion/perspectives

I Water isotopes can help detect and understand biases in water

budgets in climate models

⇒ help evaluate the credibility of projected future relative

humidity changes

I Perspectives:

I Observations: maximum variations and biases in themid/upper troposphere

I Model-data comparison methodology: de�ne consistentlycloudy situations in data and GCMs⇒develop GCM simulators of satellite data

I Modelling: water isotopes in climate change inter-comparisonslike CMIP?⇒ use intra-seasonal, seasonal, inter-annual isotope data as anobservational test for model behavior in climate change?

13/13

Conclusion/perspectives

I Water isotopes can help detect and understand biases in water

budgets in climate models

⇒ help evaluate the credibility of projected future relative

humidity changes

I Perspectives:

I Observations: maximum variations and biases in themid/upper troposphere

I Model-data comparison methodology: de�ne consistentlycloudy situations in data and GCMs⇒develop GCM simulators of satellite data

I Modelling: water isotopes in climate change inter-comparisonslike CMIP?

⇒ use intra-seasonal, seasonal, inter-annual isotope data as anobservational test for model behavior in climate change?

13/13

Conclusion/perspectives

I Water isotopes can help detect and understand biases in water

budgets in climate models

⇒ help evaluate the credibility of projected future relative

humidity changes

I Perspectives:

I Observations: maximum variations and biases in themid/upper troposphere

I Model-data comparison methodology: de�ne consistentlycloudy situations in data and GCMs⇒develop GCM simulators of satellite data

I Modelling: water isotopes in climate change inter-comparisonslike CMIP?⇒ use intra-seasonal, seasonal, inter-annual isotope data as anobservational test for model behavior in climate change?

13/13