Experimental Forecast with a New Version of the LDEO Model

contributed by Dake Chen, Stephen E. Zebiak, and Mark A. Cane

Lamont-Doherty Earth Observatory of Columbia University, Palisades, New York

For more than a decade, the LDEO model (Cane et al., 1986; Zebiak and Cane, 1987) has played an important role in our understanding and prediction of ENSO. However, the predictive skill of the original Lamont model (LDEO1) is severely limited by its unbalanced initialization scheme, its sole dependence on wind data and its large systematic biases. In the last few years, we have made considerable improvements in model initialization and data assimilation, resulting in LDEO2 and LDEO3 (Chen et al., 1995, 1998, 1999). Most recently, we have effectively reduced the systematic biases of the Lamont model with a simple statistical correction based on the regression between the leading EOFs of the model errors and the leading MEOFs of the model states (Chen et al., 2000). It is now more straightforward to assimilate data for model initialization because of much reduced model-data incompatibility. The bias-corrected model not only performs better in retrospective forecasting, but also exhibits a more realistic internal variability.

Here we present the current forecasts of this new version of the LDEO model (LDEO4). Figure 1 shows model predicted SST and wind stress anomalies in the tropical Pacific for the next three seasons. These are ensemble averages of the forecasts started from 1999-2000 winter season (DJF), with observed SST, wind and sea level data assimilated. Two sets of forecasts were made in the middle of each month with two different kinds of sea level data (tide gauge and Topex altimeter) for initialization. Thus the ensembles shown in Figure 1 are based on six individual forecasts. A closer look at the forecast integrations for NINO3 is provided in Figure 2, which shows individual 12 month forecasts beginning from 1-month-apart initial conditions from January 1998 to February 2000. Both sets of forecasts are quite self-consistent (not much spread), and the one using Topex data seems to produce initial conditions closer to those observed. The LDEO model is predicting a gradual warming from the present cold conditions to above-normal SST in the next three seasons. Anomalies above 0.5 C will appear in the eastern Pacific cold tongue region through spring and summer, and then in the eastern and central equatorial Pacific in fall.

References:

Cane, M. A., S. E. Zebiak and S. C. Dolan, 1986: Experimental forecasts of El Niño, Nature, 321, 827-832.

Chen, D., S. E. Zebiak, A. J. Busalacchi and M. A. Cane, 1995: An improved procedure for El Niño forecasting: implications for predictability. Science, 269, 1699-1702.

Chen, D., M. A. Cane, and S. E. Zebiak, 1998: The impact of sea level data assimilation on the Lamont model prediction of the 1997/98 El Niño, Geophys. Res. Lett., 25, 2837-2840.

Chen, D., M. A. Cane, and S. E. Zebiak, 1999: The impact of NSCAT winds on predicting the 1997/98 El Niño: A case study with the Lamont model. J. Geophys. Res., 104, 11321-11327.

Chen, D., M. A. Cane, S. E. Zebiak, Rafael Canizares and A. Kaplan, 2000, Bias correction of an ocean-atmosphere coupled model, Geophys. Res. Lett., in press.

Zebiak, S. E. and M. A. Cane, 1987: A model El Niño-Southern Oscillation. Mon. Wea. Rev., 115, 2262-2278.