This study investigated the predictability of summertime (June through August, JJA) temperature in South Korea. On interannual time scale, the JJA temperature variations in South Korea were, to a large extent, associated with the large-scale, low-level circulation variability in East Asia. The yearly variability of large-scale JJA circulation was then quantified by area averaged 850hPa geopotential height, termed as Korea-Japan index (KJI, 30~45N, 120~150E) and western North Pacific subtropical high index (WNPSHI, 15~25N, 110~140E). These indices were linked with remote forcing of tropical origin. The KJI was coupled with a decaying El Niño a month in advance, while the WNPSHI was influenced by sea surface temperature (SST) anomaly in the western North Pacific and a developing El Niño one to three months ahead. In addition, the JJA temperature over South Korea affected directly by tropical SST anomaly in the western North Pacific in May. Based on these teleconnection patterns, a multivariate regression model using the SST surrogates for the KJI and WNPSH as explanatory variables and a simple empirical model with an area averaged May SST were established to reconstruct the JJA temperature in South Korea. Both of the empirical models predicted reasonably well the JJA temperature with correlation coefficients of 0.57 for the multivariate regression model and 0.53 for the simple empirical model. These statistical models were also used to predict monthly temperature and, likewise, the predictability differed little. In contrast, however, if these models were applied to a suit of multi-model ensemble (MME) seasonal forecasts, the multivariate regression model outperformed the simple empirical model at all lead times ranging from one to six months. The notably different predictability arguably arose from the higher fidelity of the simulated tropical SST in the Central and Eastern Pacific than in the Western Pacific. In addition, the multivariate model outstripped the one-month lead MME prediction for summertime temperature, suggesting that the reproducibility of tropical SST, despite the regional dependency, can pave a promising way for a statistical model to predict JJA temperature in South Korea.

1. 서론

2. 자료

2.1 관측 자료

2.1.1 한반도 기온 자료

2.1.2 전지구 자료

2.2 전지구 기후모델 계절예측 자료

3. 결과

3.1 우리나라의 하계 기온

3.1.1 우리나라 하계 기온의 기후학적 특성

3.1.2 우리나라 하계 기온의 경년 변동성

3.2 우리나라 하계 기온과 동아시아 대규모 순환

3.3 열대-아열대 원격상관을 이용한 우리나라 하계 기온 예측

3.4 APCC MME 예측 자료를 이용한 우리나라 하계 기온 예측

4. 요약 및 결론

REFERANCES