연구보고서
Diagnostics of the multi-model ensemble seasonal forecast of the APEC Climate Center: Forecast skill and predictability of ENSO and its response
- 저자
- 손수진 박사
- 작성일
- 2017.07.04
- 조회
- 281
- 요약
- 목차
The effect of the amplitude and type of the El Niño-Southern Oscillation (ENSO) on the sea surface temperature (SST) predictability has been assessed through Asia–Pacific Economic Cooperation (APEC) Climate Center (APCC) Tier-1 multi-model ensemble (MME) hindcast experiments. The forecast skill and predictability of the conventional ENSO and its diversity were also examined. The MME predicts the Niño3.4 index well, even at a 4-month lead time. However, capturing the SST pattern during weak ENSO events is difficult. All weak El Niños were found to be El Niño Modoki. Some models fail to reproduce the associated tripole SST pattern over the tropical Pacific. The SST signals of the MME are less persistent than the temperature observed during weak La Niña periods.
The leading modes and residuals of the tropical Pacific SST variability were identified based on the empirical orthogonal function (EOF) analysis of monthly values. The first forced mode and its residual indicate the canonical ENSO-related and ENSO diversity-related variability, respectively. The forecast skill of the conventional ENSO remains significant up to a 6-month lead time. The ability to
predict ENSO diversity is limited and has a shorter lead time. In addition, the variance of ENSO diversity is relatively smaller than that of the canonical ENSO-related variability. The potential predictability of the ENSO diversity was also investigated to reveal the signal factor of the types. The predictability source was determined for the entire tropical Pacific; the predictability appears to be stable at long lead times.
These results have the following broad implications. First, a good ENSO forecast-based only on the Niño3.4 index cannot guarantee a good SST prediction. The strength and type of the ENSO needs to be considered when interpreting SST and other products of dynamical seasonal prediction systems. Second, it is important to know the upper limit of the predictability and actual forecast skill of the ENSO diversity, which helps to improve the model and better understand the ENSO prediction. The results shed light on the challenging problem of ENSO changes and the diversity-related predictability.

