24
Development of Water Isotope Ra2o Data Assimila2on System with Ensemble Kalman Filter Kei Yoshimura AORI, Univ Tokyo Yoshimura, Miyoshi, Kanamitsu, 2013 Yoshimura, Miyoshi, Kanamitsu, 2014 H 18 O H H D 16 O

Development*of*Water*Isotope*Ra2o** Data*Assimila2on

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Development  of  Water  Isotope  Ra2o    Data  Assimila2on  System    

with  Ensemble  Kalman  Filter

Kei  Yoshimura  AORI,  Univ  Tokyo  

 Yoshimura,  Miyoshi,  Kanamitsu,  2013  Yoshimura,  Miyoshi,  Kanamitsu,  2014

H   18O H  

H  D  16O

Page 2: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Stable  Water  Isotopes  and  Hydrologic  Cycle

•  SWI  have  integrated  records  of  phase  changes  during  its  transport.  

Iso  Var  

Iso.  Var.  

River  Flow  

H  

H  

 18O  

16O  H  

H  

<  

H  D  16O  

0.20% 0.016%

H2O

H218O HDO

(‘Heavy’ Water) Easy to condense

(‘Light’ Water) Easy to evaporate

“Fractionation” causes large heterogeneity in time and space.

Page 3: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Jasechko  et  al.,  2013

• TranspiraIon  represents  80~90  %  of  terrestrial  ET.    

⎥⎦

⎤⎢⎣

−=

ET

EET

ETT

δδδδ ET:ET T:Transp. E:Evap.

δET,  δE,  δT:Respective  isotope  ratios

Transpiration δT

≈δS

Evaporation δE

≈δS-εET δET

Soil moisture δS

Page 4: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Isotopes  in  GCM/RCM

• Incorporate  water  isotopes  as  passive  tracers  in  GCMs/RCMs.  Whenever  water  phase  change  takes  place,  isotopic  water  (HDO,  H2

18O)  behave  differently  to  ordinary  water  (H2O).  

Risi  et  al.  2008  Courtesy  of  JMA  

H18OH

HD16O

Typical  convecIve  precipitaIon  process  

Page 5: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Valida2on:  Comparison  in  δ18O  in  Precip

Page 6: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

SWING-2 o  Kick-off in 17-19 November 2008 in IAEA

HQ; chaired by C. Sturm, K. Yoshimura & D. Noone.

o  More isotopic AGCMs (at least 9) and 2 isotopic RCMs.

o  Add nudging experiments to focus on only isotopic parameterizations and on more realistic reconstruction of isotopic variations.

o  More focused on hydrologic cycle than climatology

o  Endorsed by GHP/CEOP in 2008-2010

Page 7: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Courtesy  of  L.  StoZ

Forward  Proxy  Modeling  of    δ18O  in  cellulose

Page 8: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Sea  water  δ18O  derived  from  coral  and  model  (temperature  effect  removed  by  Sr/Ca)

           δ18Osw  records  are  well  reproduced  both  seasonally  and  inter-­‐annually  in  various  sites.

-­‐0.5  

0  

0.5  

-­‐0.3  

0.7  

Nauru  (1S  166E)                                      r=0.52

Papua  NewGuinea  (5S  144E)        r=0.73

Vanuatu  (16S  167E)                r  =  0.42 0  

0.5  

1  Seychells  (5S  54E)                r  =  0.41  

-­‐0.3  -­‐0.1  0.1  0.3  0.5  

1975 1980 1985 1990 1995 2000

Clipperton  (10N  108W)        r=  0.46

:Model :Coral  record

δ18 Osw

(‰)

-­‐0.5  

0  

0.5  

1  

1975 1980 1985 1990 1995 2000

-­‐2  

0  

2  Pohnpei  (6N  158E)                                      r=0.52

Courtesy  of  K.  Kojima

Page 9: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Way  forward:  Isotope  Reanalysis

H18OH

δ18O,  δD T,  P,  etc  

Regression  Analysis,  etc

ConvenIonal  paleo-­‐climatology  Empirical  relaIonship  between  δ-­‐T  or  -­‐P.

Ongoing  approach:  Forward  proxy  modeling  Proxy-­‐Climate  relaIonship  may  vary  in  space  and  Ime

Forward  Proxy  Modeling  with  GCMs

ObjecIve  analysis  with  DA

Page 10: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Toward  “Real”  Isotope  Reanalysis:  Data  Assimila2on  of  Isotope

R.S. HDO

H18OH

HD16O

IsoGSM

LETKF

[ ]

afofa

Tf

Tf

m

H δXXyKXX

RXHδUUDδXKXXδXxxX

+−+=

=

−==−−

))((

)(

,11

1 !

Targets: ü First global 4D analyses for vapor isotopes.

ü Accurate Precip. isotopes in fine resolution.

ü Possibility of improvement on other dynamical fields.

Page 11: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

SCIAMACHY/Envisat:  surface  vapor  HDO    (Frankenberg  et  al.,  2009,  Science)

Page 12: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

TES/Aura:  mid  troposphere  vapor  HDO  (Worden  et  al.,  2007,  Nature)

12

Page 13: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Local  Ensemble  Transformed  Kalman  Filter  (Miyoshi  and  Yamane,  2008)

• Not  only  the  assimilated  variables,  but  also  other  variables  will  be  corrected  to  be  a  consistent  field.  

Page 14: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Idealized  Experiments  (OSSE)

• Assume  one  realizaIon  of  AMIP  runs  as  truth.  

Page 15: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Grid  with  denser  observa2ons,  δ18O

NoObs

DeltaOnly

Page 16: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Grid  with  denser  observa2ons,  T

NoObs

DeltaOnly

Page 17: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

(a)  Bias,  NoObs

(b)  RMSD,  NoObs

(c)  Bias,  DeltaOnly

(d)  RMSD,  DeltaOnly

T

Page 18: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

What  about  more  realis2c  situa2on?    Experiments  with  conven2onal  measurement  system  

σ=0.995 U  [m/s]

V  [m/s]

T  [K]

q  [g/kg]

Ps  [hPa]

δ18O  [‰]

δD  [‰]

UVTq 1.33   1.30   0.40   0.42   1.04   0.98   7.23  

UVTq+δD 1.27   1.25   0.40   0.41   0.99   0.93   6.94  

σ=0.8835 U  [m/s]

V  [m/s]

T  [K]

q  [g/kg]

δ18O  [‰]

δD  [‰]

UVTq 1.49   1.39   0.55   0.69   1.41   10.77  

UVTq+δD 1.42   1.34   0.53   0.68   1.35   10.35  

Global  RMSD

Page 19: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Ul2mate  goal:  Climate  Reanalysis  

• Much  longer  records  than  man-­‐made  observaIon  • Oceanic  sediment  δ18O  (millions  yBP)  • Icesheet  cores  δ18O・δD  (~800  kyBP)  • Icecap  cores  δ18O・δD  (~20  kyBP)  • Speleothem  δ18O  (~2000  yBP)  • Treering  δ18O  (~1000  yBP)  • Coral  δ18O  (~400  yBP)  

• Bridging  data  and  physics,    consistently!    

19

H18OH

Page 20: Development*of*Water*Isotope*Ra2o** Data*Assimila2on
Page 21: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Summary

• Isotopic  Data  as  input  observaIon  had  posiIve  impact  on  not  only  isotopic  fields  but  also  dynamical  fields.    • (Selfish)  suggesIon  for  new  observaIons:    • Accuracy  <  Number  of  data  • Temporal  resoluIon  <  Longer  data  • Dense  coverage  <  Sparse  but  equally  distributed  

• There  is  potenIal  for  dynamical  constraint  by  isotopic  proxy  data  for  the  past,  but  lots  of  technical  obstacles  exist.  

Page 22: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Paired  isotopic  proxy  data  since  8,000yBP

Liu  et  al.,  2014,  Nature  Comm.

Page 23: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

LH  minus  MH  (‰) PNA+  minus  PNA-­‐(‰) IsoGSM  simulated  precipitaIon  δ18O  difference  

IsoGSM  simula2ons

•  Time-­‐slice  runs  for  MH  and  LH  for  30  years,  respecIvely.    •  SST  anomaly  simulated  by  IPSL-­‐CGCM  was  forced.  

Page 24: Development*of*Water*Isotope*Ra2o** Data*Assimila2on

Result  (Global  RMSE  for  δ18O,  Wind,  Temp,  and  Surface  Pressure)

3.3  

3.4  

3.5  

3.6  

3.7  

NoObs   SCIA   TES   GNIP   ALL  

δ18O  (‰)  at  2m  

4.15  

4.2  

4.25  

4.3  

4.35  

4.4  

NoObs   SCIA   TES   GNIP   ALL  

Zonal  Wind  (m/s)  at  boZom  

2  2.1  2.2  2.3  2.4  2.5  2.6  

NoObs   SCIA   TES   GNIP   ALL  

Air  Temperature  (K)  at  boZom  

540  

550  

560  

570  

580  

590  

NoObs   SCIA   TES   GNIP   ALL  

Surface  Pressure  (Pa)