Predictive skill of statistical and dynamical climate models in SST forecasts during the 1997-98 El Nino episode and the 1998 La Nina onset

Critical reviews of forecasts of ENSO conditions, based on a set of 15 dyna mical and statistical models, are given for the 1997-98 El Nino event and t he initial stages of the 1998-99 La Nina. While many of the models forecast ed some degree of warming one to two seasons prior to the onset of the El N ino in boreal spring of 1997, none predicted its strength until the event w as already becoming very strong in late spring. Neither the dynamical nor t he statistical models, as groups, performed significantly better than the o ther during this episode. The best performing statistical models and dynami cal models forecast SST anomalies of about +1 degrees C (vs 2.5 degrees-3 d egrees observed) in the Nino 3.4 region prior to any observed positive anom alies. The most comprehensive dynamical models performed better than the si mple dynamical models. Once the El Nino had developed in mid-1997, a larger set of models was able to forecast its peak in late 1997 and dissipation a nd reversal to cold conditions in late spring/early summer 1998. Overall, h owever, skill for these recent two years does not appear greater than that found over an earlier (1982-93) period. In both cases, median model correla tion skill averaged over lead times of one to three seasons is near or just above 0.6. Because ENSO extremes usually develop in boreal spring or early summer and persist through the following winter, forecasting impact tendencies in extr atropical North America for winter (when impacts are most pronounced) at 5 months of lead time is not difficult, requiring only good observations of t he summer ENSO state and knowledge of the winter teleconnections. Because o f the strength of the 1997-98 El Nino acid the consequent skill of 5-month lead forecasts of U.S. winter 1997-98 impacts, the success of these forecas ts was noticed to an unprecedented extent by the general public. However, f orecasting impacts in austral winter that occur simultaneously with the ini tial appearance of an ENSO extreme (e.g., in Chile, Uruguay, Kiribati, Ecua dor, and Pent) require forecasting the boreal spring/summer onset of ENSO e vents themselves at several months of lead time. This latter task is formid able, as evidenced by the fact that formal announcements of an El Nino did not occur until May, leaving little time for users in the above regions to prepare. Verbal summaries of ENSO forecasts issued to users worldwide during the 199 7-98 El Nino event contained ambiguities. To address the needs for forecast s to be expressed verbally for nontechnical users and also to be precise en ough for meaningful utility and verification, a simple numerically based ve rbal classification system for describing ENSO-related forecasts is present ed.