SST Predictions with a Global Coupled GCM

contributed by Zhengxin Zhu, Bohua Huang, David G. DeWitt, J. Shukla, and Edwin K. Schneider

Center for Ocean-Land-Atmosphere Studies (COLA), Calverton, Maryland

A coupled ocean-atmosphere system has been developed at COLA for making seasonal to interannual forecasts of tropical Pacific SST (Schneider et al., 1998a). The system incorporates subsurface ocean measurements in the initial conditions. Predictions are made with a coupled ocean-atmosphere GCM with no flux corrections or anomaly coupling. The strategy of anomaly initial conditions is employed to reduce the shock of insertion of real ocean data into an imperfect model.

The atmospheric component of the coupled model is the latest version of the COLA atmospheric GCM, which is an global spectral model with a state of the art suite of physical parameterizations, as described by DeWitt and Schneider (1998). The resolution is triangular 30 with 18 levels in the vertical. The deep convection parameterization is the relaxed Arakawa-Schubert scheme (Moorthi and Suarez, 1992), and the diagnostic cloud-radiation interaction scheme follows Kiehl et al. (1994,1996).

The ocean component of the coupled model is the GFDL MOM1 ocean model (Pacanowski et al., 1993), with an almost global domain (0E to 360E, 70S to 65N). The latitudinal resolution is 0.5 degree between 10S and 10N, changing to 1.5 degree beyond 20N and 20S. The zonal resolution is 1.5 degree, and there are 20 levels in the vertical. A sunlight penetration scheme is added to the ocean model, which is important for realistic simulation of the annual cycle of the tropical ocean (Schneider and Zhu, 1998). The tropical resolution is the same as used in the COLA anomaly coupled forecast system (Kirtman et al., 1997).

The ocean data assimilation (ODA) system has been described in detail by Huang and Kinter (1997). It uses the variational optimal interpolation method following Derber and Rosati (1989). The ocean model used in ODA is identical to that used in the coupled model. The period of the analysis is from January 1986 to October 1998.

The coupled system does not apply any flux corrections or statistical corrections during its coupled runs. However, an anomaly treatment of the ocean and atmosphere initial conditions is used to eliminate initial shock. The procedure is as follows: First we run the coupled model for many years starting from an ocean state from the ODA to determine the model climatology. The ocean initial condition is then formed by adding the ODA anomalies to the model climatology. This anomaly treatment includes all the prognostic variables. The atmospheric model is spun up for one month before coupling, using SST that is the sum of model climatology and the observed SST anomalies. Thus, the ocean and atmosphere are in an approximately balanced condition when starting a coupled prediction run.

To document the skill of the forecast system, a series of 48 hindcasts was conducted, which start from the end of January, April, June and September of each year from 1986 to 1997. Fig. 1(a) shows the anomaly correlation of the predicted and observed NINO3 SST. The correlation is above 0.6 up to 6 months and above 0.5 up to 12 months of lead time. Fig. 1(b) is the point correlation of SST at 6 months lead time from the hindcasts. It can be seen that the higher correlations (>0.5) are in the equatorial Pacific, with a broader area in the eastern Pacific.

Fig. 2 shows the tropical Pacific SST anomalies taken from the latest forecast, which initial condition is the end of October, 1998. The forecast shows the cold anomaly to be at peak next spring, with an extreme of more than -3 degrees in the central Pacific. The cold anomaly then gradually weakens in the summer of 1999.

Acknowledgments: This work was supported under NOAA grant NA26-GP0149 and NA46-GP0217 and NSF grant ATM-93-21354.

References:

Derber, J. and A. Rosati, 1989: A global oceanic data assimilation system, J. Phys. Oceanogr., 19, 1333-1347.

DeWitt, D. G. and E. K. Schneider, 1998: On the processes determining the annual cycle of Equatorial sea surface temperature: A coupled general circulation model perspective. Mon. Wea. Rev., accepted.

Huang, B. and J. L. Kinter III, 1997: A global analysis for 1986-1992 Cola Rep. 35, 39pp.

Kiehl, J., J. Hack, B. Briegleb, 1994: The simulated Earth radiation budget of the National Center for Atmospheric Research community climate model CCM2 and comparisons with the Earth Radiation Budget Experiment (ERBE). J. Geophys. Res., 99, 20815-20827.

Kiehl, J., B. Boville, B. Briegleb, J. Hack, P. Rasch, and D. Williamson, 1996: Description of the NCAR Community Climate Model (CCM3). NCAR Technical Note NCAR/TN-420+STR.

Kirtman, B., J. Shukla, B. Huang, Z. Zhu, and E. Schneider, 1997: Multiseasonal predictions with a coupled tropical ocean global atmosphere system. Mon. Wea.Rev., 125, 789-808.

Moorthi, S. and M. Suarez, 1992: Relaxed Arakawa-Schubert: A parameterization of moist convection for general circulation models. Mon. Wea. Rev., 120, 978-1002.

Pacanowski, R., K. Dixon, and A. Rosati, 1993: The GFDL modular ocean model users guide, version 1.0. GFDL Ocean Tech. Rep. No. 2.

Schneider, E. K. and Z. Zhu, 1998: Sensitivity of the simulated annual cycle of SST in the equatorial Pacific to sunlight penetration. J. Climate, 11, 1932-1950.

Schneider, E. K., B. Huang, Z. Zhu, D. Dewitt, J. Kinter III, B. Kirtman, and J. Shukla, 1998a: Ocean data assimilation, initialization, and predictions of ENSO with a coupled GCM. Mon. Wea. Rev., accepted.

Schneider, E. K., Z. Zhu, B. Huang, D. DeWitt, and J. Shukla, 1998b: SST predictions with a global coupled GCM. Experimental Long-Lead Forecast Bulletin, Vol. 7, No. 2, 6-9.

Fig. 1. Hindcast verification of coupled prediction, updated from that shown in Fig. 1 of the June, 1998 ELLFB (Schneider et al., 1998b). a) Anomaly correlation of NINO3 SSTA from the total 48 hindcasts. b) Point correlation of the hindcasts at 6 months lead.

Fig. 2. Forecast of tropical Pacific SST anomalies. The initial condition is October 31, 1998.