Evaluation of the Tropical Intraseasonal Oscillation in a Coupled Climate Model

Document Sample
Evaluation of the Tropical Intraseasonal Oscillation in a Coupled Climate Model Powered By Docstoc
					NOAA 32nd Climate Diagnostics & Prediction Workshop               22-26 October 2007




        Evaluation of the tropical intraseasonal
         oscillation in a coupled climate model

     Suhee Park, Young-Hwa Byun, Han-Cheol Lim, and Won-Tae Kwon
        National Institute of Meteorological Research (METRI) /KMA, Korea


                                 Contents
          1.   Introduction
          2.   Development of METRI CGCM
          3.   Intraseasonal oscillations in model simulations
          4.   The effect of air-sea coupling
          5.   Summary
                           Introduction (1/2)
 Background

     The development of the METRI CGCM
  -   The METRI coupled GCM is developed for the seasonal prediction (2007)

     East Asian Summer Monsoon (EASM)
  -   An accurate prediction of the EASM is important for economy, water management and human
      life in East Asian region.
  -   It has complex space and time structures, which covers both subtropics and mid-latitudes and
      from subseasonal to interdecadal time scale.

     Deficiencies in the skill of dynamical seasonal prediction
  -   The limited predictability of the EASM may be due to the fact that the contribution
      from the external variability over the region is relatively weak and comparable to
      that from internal variability (e.g., Stern and Miyakoda 1995; Goswami et al. 2006).
  -   Goswami et al. (2006) proposed that the internal interannual variability of the
      monsoon annual cycle is primarily due to interaction between the monsoon annual
      cycle and the summer intraseasonal oscillations.

     Therefore, the ability of the dynamical model for the seasonal
      prediction to represent the intraseasonal oscillation (ISO) is
      important.
                               Introduction (2/2)
 Previous studies about simulation of the ISO

      AGCM
   -   Poor simulation of the ISO is a generic problem in GCMs. Typically, model ISOs are too
       weak and propagate too fast (Slingo et al. 1996)
      CGCM
   -   Current state-of-the-art coulped GCMs still have significant problems and display a wide
       range of skill in simulating the tropical intraseasonal variability. The total intraseasonal (2–
       128 day) variance of precipitation is too weak in most of the models. (Lin et al. 2006)

      Sensitivity to model ISOs
   -   Factors expected to be important for ISO simulations: model physics, model resolution
       and air–sea coupling.
   -   Slingo et al. (1996) found that convection schemes with convective available potential
       energy (CAPE) type closure tend to produce more realistic MJO signals. Improvements
       of ISO simulations were also found by adding moisture triggers to the deep convection
       schemes (e.g., Tokioka et al. 1988; Wang and Schlesinger 1999; Lee et al. 2003), or by
       including convective downdrafts and rain evaporation (Maloney and Hartmann 2001).
   -   ISO simulation was found to be improved when using higher horizontal resolution (e.g.,
       Kuma 1994) and/or vertical resolution (Inness et al. 2001).
   -   Coupling to the ocean has been found by many studies to improve the MJO signals (e.g.,
       Waliser et al. 1999; Inness and Slingo 2003, Sperber et al. 2005)
                            Objectives
 The purpose of this study is to evaluate the intraseasonal
  oscillation in a coupled climate model

      Comparison of the simulated ISO by coupled model with those of
       uncoupled model

       Analysis of the effect of air-sea coupling on the ISO
       Development of Coupled Climate Model (1/2)
   METRI CGCM

   Component models: AGCM and OGCM
  The atmospheric component model                                  The ocean component model

  YOURS-GSM                                                        MOM3 (Pacanowski and Griffies 1998)
        Hor. Res. : T62 (about 2˚ x 2˚)                                   Hor. Res. : Tropics 1/3˚, Extratropics 1˚
        Ver. Res. : L28                                                   Ver. Res. : L40
                                                                          domain : 75˚S - 65˚N, 0-360˚E
        PBL : nonlocal scheme (Hong and Pan
        1996)                                                             Ver. mixing: nonlocal K-profile (Larger et
        Land model : two layer soil model (Mart and                       al. 1994)
        Pan 1984)                                                         Hor. Mixing:
        Cumulus convection :                                                - tracers; isoneutral method (Gent and
        -Relaxed Arakawa-Schubert (Moorthi and Suarez                     Mcwilliams 1990)
        1992)                                                               - momentum; nonlinear scheme
                                                                          (Smagorinsky 1963)
                                                                          * Explicit free surface, Partial cell

 YOnsei University Research model System (YOURS) Global Spectral Model (GSM)
  Development of Coupled Climate Model (2/2)
 METRI CGCM

 Coupling Strategy

                                           SST
 Atmospheric model run                                           Ocean model run

      YUORS-GSM                                                      MOM3
  Integration time: 24 hours                                 Integration time: 24 hours
                                           FLUX


                               Zonal/meridional Momentum flux
                                Short/long wave radiation flux
                                         Precipitation
                                  Latent/sensible heat flux


  Initial data:
      I.C. for the atmosphere : NCEP/DOE Reanalysis 2 data
                                     (R2, Kanamitsu et al. 2002)
      I.C. for the ocean : Global Ocean Data Assimilation System data
                                  (GODAS, Dehringer et al. 2005)
                                Experiments
 Experimental Design

 Simulation period
     May-June-July-August-September 1997-2004 (5 months, 8 years)

 Two runs are designed to investigate the characteristics of coupled model.
         Experiments                                 Descriptions
          AGCM                Two-tier SMIP type run (observed SSTs)
                              * YOURS GSM
          CGCM                One-tier SMIP type run
                              * YOURS GSM + MOM3
                   * SMIP: Seasonal Prediction Model Intercomparison Project

 Observed data
 Precipitation :
 CMAP: monthly data
 GPCP Satellite-Derived (IR) GPI Daily Rainfall Estimates : daily data
 Atmospheric variables : NCEP/DOE Reanalysis (R2) data
                             Seasonal mean climate
 Performance – Seasonal mean climate (JJA)
         AGCM is forced by OISST
                   OBS                            AGCM                            CGCM




Prec




VP200




U850




 AGCM : strong convection => overestimated precipitation with strong U850 and VP200
 CGCM : Systematic errors are reduced.
                      Intraseasonal Variability
   Performance – Intraseasonal variability
  Standard deviation of 20-70-day filtered variables
                OBS                      AGCM           CGCM




                SST
Prec




VP200




U850
              Intraseasonal Variability
 Strong ISO case: May 1 ~ June 30, 2002   *20-70-day filtered data

       Prec                U850                     VP200
                   Intraseasonal Variability
 Wavenumber-frequency spectra of precipitation
  Zonal spectra computed over                Meridional spectra computed over
   equatorial (2.5°S–2.5°N, 0–360°E)           the Western Pacific (10°S–37.5°N,
   time–longitude data                         100°E-150°E) time-latitude data



                                       OBS




                                       AGCM




                                       CGCM
                       Intraseasonal Variability
 Complex EOF Composite to capture the evolution of the ISO
 CEOF analysis is performed in order to find the dominant propagating mode
     20-70-day bandpass filtered 200 hPa velocity potential fileds
Composite maps
     Separated by 30 deg over 360 deg phase range corresponding to periods of the
    dominants mode. (from phase-1 to phase-13)
                 OBS                      AGCM                        CGCM



  Phase 1
    0°



  Phase 4
    90°



  Phase 7
   180°



  Phase 10
    270°
                  Intraseasonal Variability
 Complex EOF composite one cycle : Precip. & V850
            OBS                AGCM              CGCM




 Phase 1
   0°




 Phase 4
   90°




 Phase 7
  180°




 Phase 10
   270°
                                Air-sea Coupling
   Lag Correlation of precipitation with surface variables
                                                               (Woolnough et al. 2000)
 Observation



                                                                                    Shortwave
     SST                                                                             radiation
                                                                                       flux




     Latent                                                                          Zonal
      Heat                                                                           Wind
      flux                                                                           stress




 Observations show a coherent relationship between convection and surface variables. Warm
 SSTs lead convection by 5-8 days. These result from increased SW radiation and reduced
 evaporation by weaker winds during suppressed phase of the ISO.
                            Air-sea Coupling
 The ISO and Coupling with the ocean : the Air-sea
  interaction in the ISO over the Indian Ocean and the Western Pacific
                                   Suppressed
                                 convection phase


     Stable condition                          Easterly wind anomaly
     No cloud                                 Reduced Mean state westerly wind

    Downward SW flux increase                  Evaporation decrease
    SW flux to ocean increase                  LH flux to ocean increase



                                 SST increase

                                    Enhanced
                                 convection phase

 It plays an important role in the evolution of the ISO.
                              Air-sea Coupling
    Lag Correlation
    AGCM



                                                                          Shortwave
     SST                                                                   radiation
                                                                             flux




     Latent                                                                Zonal
      Heat                                                                 Wind
      flux                                                                 stress




 AGCM simulations show a poor relationship between convection and SST.
                              Air-sea Coupling
    Lag Correlation
    CGCM



                                                                              Shortwave
     SST                                                                       radiation
                                                                                 flux




     Latent                                                                    Zonal
      Heat                                                                     Wind
      flux                                                                     stress




 CGCM simulations show a coherent relationship between convection and SST.
                                   Summary
 In the CGCM, simulated seasonal mean climate is more similar to observation than
  that in the AGCM.

 The evolution of convective activity over the Indian Ocean and the eastward
  propagation from the Indian Ocean to the western Pacific is clearly better in the
  CGCM than in the AGCM.

 It is found that the coupled model improve the ISO simulation and it seems to be due
  to the effect of the air-sea interaction. Suppressed convections make SSTs warmer
  and warm SSTs initiate enhanced convection.

 It is noted that, although the relation between convection and SST in the CGCM
  experiment is similar to observation, the magnitude of correlation is lower than
  observation, about half. It seems to be results of the unrealistic latent heat flux
  anomalies, which are related not only with the error in wind anomalies, but also with
  the error in the low level mean state wind.
Thank you very much!
                                Seasonal mean climate
  The effect of air-sea interaction on seasonal climate
                                       CGCM minus AGCM
                                                                             Air-sea interaction in CGCM

                             SST
      AGCM minus OBS                                                               Overestimated
                                                                                    convection

                              Prec                                                    Less fluxes
                                                                                       to ocean

                             VP200                                                 Decreasing SST

                                                                                      Reducing
                              U850
                                                                                     precipitation


 Differences between two model simulations can be attributed to the effect of air-sea interaction in CGCM.
 - Overestimated convection from AGCM leads to decreasing flux to ocean  less flux to ocean results in decreasing SST
 cold SSTs over tropical Indian Ocean and Pacific region reduce precipitation  more realistic precipitation
                         Interannual Variability
 Performance – Interannual variability : ENSO
 El Nino (1997,2002,2004) minus La Nina (1998,1999)               Prec


                        SST


                                                  OBS



    OBS




                                                  AGCM




  CGCM

                                                  CGCM




 Interannual variability in CGCM is comparable to that in AGCM forced with observed SST!
             Intraseasonal Variability
 [CASE] May 1 ~ 30, 2002
       SST                  SWF          LHF
                           Air-sea Coupling
 The Air-sea interaction in the ISO simulated by CGCM

                                  Suppressed
                                convection phase


    Stable condition                        Easterly wind anomaly
    No cloud                                Reduced Mean state westerly wind

   Downward SW flux increase                Evaporation decrease
   SW flux to ocean increase                 LH flux to ocean increase


                                SST increase

                                   Enhanced
                                convection phase

				
DOCUMENT INFO
Shared By:
Stats:
views:22
posted:7/1/2008
language:English
pages:23