Te - DecVar_

Document Sample
Te - DecVar_ Powered By Docstoc
					 Decadal Changes in Properties of
El Niño-Southern Oscillation (ENSO)




               Rong-Hua Zhang
Earth System Science Interdisciplinary Center (ESSIC),
         University of Maryland, College Park
  Observations => El Nino Property shift
                  in the late 1970s

Pre-shift period: 2-3 yr oscillation; westward propagation
              of SSTA; weak SSTA in far east. Equ. Pacific

Post-shift period: 4-5 yr oscillation; eastward propagation
             of SSTA; strong SSTA in far east. Equ. Pacific


    What caused the decadal changes ?
Mechanisms for Decadal ENSO Changes
• Random factors
  Stochastic forcing, Nonlinearity, Residual effect, …


• Deterministic factors
 Changes in mean climate states:
   Atmosphere: wind stress
   Ocean     : the thermocline depth, stratification, SST, …

 Changes in the spatial structure of subsurface
                ocean thermal fields
The role of atmos stochastic forcing (SF)
A null hypothesis for causing ENSO irregularity
       Significant effects on ENSO & its changes
              ENSO characteristics, Predictability, …

Limitations to previous studies:
  Random factor  systematic & coherent way in Obs?
  Focusing on amplitude & period; propagation ?
  Decadal subsurface temperature changes?

  ICM : Zebiak-Cane model: Te specified
  HCM: Statistical + OGCM

What are the respective roles relative to Te ?
  Observed decadal changes in ocean
 temperature in the Pacific in the late 1970s

Data: World Ocean Database
      at NODC/NOAA (Levitus et al.)

Decadal variability in the Pacific basin:
 1. The Midlatitude/extratropics
 2. The tropics
 3. Oceanic connections from the
     midlatitude/extratropics to the tropics
Temp anomaly at 250 m depth
157W-22N




172W-2N




157W-2N
   Decadal change in entrainment
 temperature (Te) & its role in ENSO
Hypothesis:
   Subsurface thermal structure change
       => ENSO changes

Testing using coupled models
      The Temperature of subsurface water
          entrained into the mixed layer (Te)


     Decadal changes in the structure of Te
Te : The Temperature of subsurface
water entrained into the mixed layer
                   Wind stress (τ)


                   SST
mixed layer


    Te        We
                              Te
             SST anomaly model
T '        T              T '      T              T '
      u '     (u  u ' )       v'     (v  v ' )
 t         x              x        y              y

         
        ( w  w ) M (  w  w )  wM (  w)
                   '               '
                                                    
                                             (Te  T )
                                                H
                                 (Te  T ' )
                                          '
        ( w  w ) M ( w  w )
                  '            '
                                              T '
                                     H
         h                        2 v
             h  ( H hT )                (Te  T ' )
                          '                     '

         H                     H (H  H 2 )

          Te: The Temperature of subsurface water
               entrained into the mixed layer
                              Outline
• Introduction
• Intermediate ocean model (IOM) with an empirical
  parameterization for Te
• The Role of Subsurface Entrainment Temperature (Te)
 (1) Intermediate coupled model ( ICM; Zhang and Busalacchi, 2005)
       Ocean:   Intermediate ocean model (IOM) + an SST anomaly model + Te
       Atmospheric wind stress (τ) model: SVD-based statistical

 (2) Hybrid coupled model (HCMAGCM )
       Ocean:        The IOM as in the ICM
       Atmosphere:    AGCM (ECHAM4.5)


• The respective roles of atmospheric stochastic forcing (SF) & Te
• Summary
                          Intermediate ocean model (IOM):
                    Ocean dynamical model + a SST anomaly model
Linear part:
       u t  fv   p x  v h  2 u  (v v u z ) z
                                h                                 vt  fu   py  vh2v  (vvuz ) z
                                                                                      h



         p z  g  0               u  0

              w 2
       t      N   h  2   ( ) zz
                          h
              g

 Non-linear part

       utnl  u  (u )  vh  2 (u nl )  (vv u z ) z
                               h
                                                 nl
                                                                   u
                                                                         nl
                                                                              0


Vertical modal decomposition

      u  n 1 un n                           v  n 1 vn n                    p  n1 pn n
               10                                        10                               10




     w  n 1 wn   n dz                        n1  n nz
              10      z                                  10

                     H
             SST anomaly model
T '        T              T '      T              T '
      u '     (u  u ' )       v'     (v  v ' )
 t         x              x        y              y

         
        ( w  w ) M (  w  w )  wM (  w)
                   '               '
                                                    
                                             (Te  T )
                                                H
                                 (Te  T ' )
                                          '
        ( w  w ) M ( w  w )
                  '            '
                                              T '
                                     H
         h                        2 v
             h  ( H hT )                (Te  T ' )
                          '                     '

         H                     H (H  H 2 )

          Te: The Temperature of subsurface water
               entrained into the mixed layer
     Te Parameterization scheme

1. Cane-Zebiak scheme

     Te  Tsub  (1   )To
          
                                   
          Td 1 tanh b1 (h  h)  tanh b1 h ,
          
                                                h0
  Tsub   
                                   
          Td 2 tanh b2 (h  h)  tanh b2 h ,
          
                                                h0
          

2. Empirical parameterization scheme
             SST anomaly model
T '        T              T '      T              T '
      u '     (u  u ' )       v'     (v  v ' )
 t         x              x        y              y

         
        ( w  w ) M (  w  w )  wM (  w)
                   '               '
                                                    
                                             (Te  T )
                                                H
                                 (Te  T ' )
                                          '
        ( w  w ) M ( w  w )
                  '            '
                                              T '
                                     H
         h                        2 v
             h  ( H hT )                (Te  T ' )
                          '                     '

         H                     H (H  H 2 )

          Te: The Temperature of subsurface water
               entrained into the mixed layer
An empirical model for the Temperature of
water Entrained (Te) into the mixed layer
   Historical data: simulated & observed
         Forced ocean model run => currents, pressure fields ( Sea level (SL) )


• Inverse modeling of Te
   Obs. SST fields etc. => SST anomaly model => Te
• Statistical relationships based on history: SLTe
    SVD-based analysis of covariance between SL and Te

   Given SL from ocean model => Te => SST calculation

  Nothing could be better than this procedure in simulating SSTA !!
   Data Sets and Construction of Te models

• SST (Reynolds et al.):            1950-1999
• Dynamical ocean model run forced by NCEP: 1960-1999
              => mean fields; anomaly fields: sea level (SL) etc

  Inverse modeling of SSTA equation =>                         Te
________________________________________________________________________

• Te model construction: SL  Te
  Seasonally invariant ( 1 model for all months);   1963-1996
                                          1963-1979          1980-1996
 Intermediate coupled model
                 (ICM):
  IOM + statistical anomaly model
Atmos: A SVD-based wind stress model:

            Observed SSTAs

            Wind stress anomalies (τ)
            from ECHAM4.5 24
            member ensemble mean

Ocean: IOM with Te
Hybrid coupled model (HCM):
           IOM + an AGCM
Atmosphere: The European Center + The Max
            Planck Institute for Meteorology (MPI)
            Atmospheric GCM (ECHAM4.5)
                     T42; 19 layers

Ocean: IOM with Te
    ICM                        HCM                          AGCM
(Intermediate coupled model)   (Hybrid coupled model)



                                 AGCM (ECHAM4.5)


                                                                         SST anomaly model

                                                                s
                                                           ,w
                                                      ,v   s
                                                     us

                                                                    SL        Te model
                               Intermediate ocean model
   Realizations of atmos stochastic
  forcing (SF) in the coupled models

  Wind stress response: Signal    +   Noise (SF)


                          SST


HCM:   AGCM + IOM with Te
        SF: implicitly included
ICM:    statistical + IOM with Te
       (1) Signal part in wind stress anomalies (No SF)
       (2) Both parts: SF explicitly taken into account
                         τ         τ
          Construction of signal & SF
    Wind stress ( τ ) : Signal   +   noise (stochastic forcing)

     obs. SSTAs
-




      AGCM (ECHAM4.5)
      24 ensemble mean                     HCM (ECHAM4.5 + IOM)


                           Residual

                                               A first order autocorrelation
                                                          model (AR1)

                  τ                    τ
                   Signal +              SF     => Used in ICM
     The ICM experiments
τSignal   τSignal + τSF
                   Te model : 63-96;
                     63-79; 80-96
                    τ model : 63-96;
                     63-79; 80-96



                           Te63-79, Te80-96
The   HCMAGCM                    experiments

   AGCM (ECHAM4.5)


                                          SST anomaly model

                               ,ws
                              vs
                       u s,
                                     SL        Te   model
 Intermediate ocean model




                                               Te63-79 , Te80-96
     Coupled modeling experiments on the
     effects of Te & SF on ENSO properties

                          ICM without HCM   ICM with
                          SF                SF

Te(63-79)   Amplitude:
            Period:
            Propagation      ?        ?       ?
            :

Te(80-96)   Amplitude:
            Period:
            Propagation     ?         ?       ?
            :
Nino3 SSTA power sprectra
AGCM (ECHAM4.5) + an IOM with Te
 Effects of SF & Te on ENSO properties
                           ICM without HCM                      ICM with
                           SF                                   SF
            Amplitude:   west of 140W       west of 14 0W       west of 14 0W
Te(63-79)
            Period:      regular (2yr)      irregular (3-5yr)   irregular (3-5yr)
            Propagation: Westward           Westward            Westward



Te(80-96)   Amplitude:      east of 140W      east of 140W east of 140W
            Period:         regular (5yr)   irregular (3-7yr) irregular (3-7yr)
            Propagation:   Eastward         Eastward          Eastward
              Oceanic processes involved
    T    '
                       T              T '      T              T '
               u '       (u  u ' )       v'     (v  v ' )
     t                x              x        y              y

                   
                  ( w  w ' ) M (  w  w ' )  wM (  w)           (T H T )
                                                                        e


                                           (Te  T ' )
                                                     '
                  ( w  w ) M ( w  w )
                               '             '
                                                        T '
                                               H
                   h                        2 v
                       h  ( H hT )                (Te  T ' )
                                    '                     '

                   H                     H (H  H 2 )


              Off-equatorial       Te
                                                     SST
                                         Hori adv.       Vert adv.
mixed layer
                                        Te
              thermolcine
                 Summary
• Demonstrate a new factor determining
                    the El Nino properties

• Different modulating effects of SF & Te:
   SF        Irregularity of ENSO amplitude
                      & oscillation periods
   Te        Phase propagation
    Implication for observations
 Explain       El Nino Property shift in the late 1970s

Pre-shift period: 2-3 yr oscillation; westward propagation
              of SSTA; weak SSTA in far east. Equ. Pacific

Post-shift period: 4-5 yr oscillation; eastward propagation
             of SSTA; strong SSTA in far east. Equ. Pacific
   Implications for coupled modeling

Common biases with OGCM-based Coupled Models:
     Quas-biennial (~2 yr) oscillation;
     westward propagation of SSTA on equ.;
     weak SSTA in far eastern equatorial Pacific


Improvements: better Te simulations
Thank you !!!

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:53
posted:4/8/2013
language:English
pages:41