The western North Pacific is the basin with the most frequent tropical cyclone activity in the world. Tropical cyclone activity, including heavy rainfall, is one of the natural phenomena that occurs in the East Asia. There are many institutes and/or organizations that work toward predicting seasonal tropical cyclone activity over the western North Pacific. Most of these institutes and/or organizations predict the number of tropical cyclones and Accumulated Cyclone Energy (ACE) but do not cover seasonal tropical cyclone forecast tracking. Studying the variability of tropical cyclone tracks associated with large circulation is very important for predicting tropical cyclone activity. This study analyzes the characteristics of tropical cyclone activity to predict tropical cyclone track variability over the western North Pacific during the boreal summer (June to August, JJA season).

The observed oceanic, atmospheric, and tropical cyclone best track from the National Center for Environmental Prediction (NCEP), the National Oceanic and Atmospheric Administration (NOAA), the Regional Specialized Meteorological Center-Tokyo (RSMC) and the Asia-Pacific Economic Cooperation Climate Center (APCC) Multi Model Ensemble (MME) from 1983-2005 were used as hindcast data. APCC collects dynamical ensemble seasonal production data from national meteorological and hydrological services in the APEC member economies, and produces and disseminates one-month lead three month mean climate forecasts every month, with four deterministic and one probabilistic forecasts. For this study, the Simple Composite Method (SCM) ensemble method was adopted, which is a simple arithmetic mean of bias corrected prediction, giving equal weight to the predictions from each individual model. The performance skill of the MME is generally higher than that of the constituent models individually. To define the dominant modes of variation associated with the tropical cyclone track, Empirical Orthogonal Function (EOF) analysis was applied. For the significance test, the Student’s t-test statistical method was chosen for this study.

The first EOF mode of the tropical cyclone track pattern explains 13.2% of the total variance and shows positive loading over the tropical western North Pacific. This suggests an above normal frequency in the occurrence of north-westward tracks toward the East China, Korean peninsula, and Japan during the positive phase of this mode. Negative loading was found over the South China Sea and the Philippine Sea. The first mode spatial distribution is very similar to the tropical cyclone track density during El Niño Modoki years from previous studies. The El Niño Modoki Index (EMI), calculated from observations, and the MME are highly correlated with the first EOF mode Principal Component (PC) of the tropical cyclone track, with correlation coefficients of 0.49 and 0.41, respectively. The East Asia Summer Monsoon (EASM) has been considered as one of the important factors affecting atmospheric circulation over the western North Pacific and it is significantly correlated with the EOF first mode PC of the tropical cyclone track. Both the observations and the MME are correlated with the EOF first mode PC, with correlation coefficients of 0.53 and 0.48, respectively.

According to the characteristics of the tropical cyclone track using EOF analysis, in the first mode normalized PC of the tropical cyclone track, the above normal year should be selected when the standard deviation is more than 1 sigma and the below normal year for less than 1 sigma. In an above normal year, the Sea Surface Temperature (SST) anomaly is warmer than the climatology over the Kuroshio extension and Central Pacific ocean, but the SST over the China and Philippine Seas is cooler than the climatology. Cyclonic wind circulation in the low levels is developed near the Philippine Sea. On the other hand, the SST anomaly is warmer than the climatology over the Western Pacific ocean and cooler than the climatology over the Central Pacific ocean. Anticyclonic circulation in the low levels is developed near the Philippine Sea during the below normal year. 5880-gpm can be defined as the boundary line for the Western North Pacific High (WNPH) for two phases. The WNPH during the above normal year is located more toward the east than the climatology. In contrast, the WNPH during a below normal year extends westward. For a below normal year, a tropical cyclone cannot move northward due to the extended WNPH over the east China sea. However, tropical cyclones can move northward in the above normal years.

This study developed a new index, the Tropical Cyclone Track Index (TCTI), which uses multiple linear regression to predict the seasonal tropical cyclone track. TCTI is composed of EMI and EASMI. TCTI indicates a 0.58 correlation coefficient with the first EOF mode PC of the tropical cyclone track. The TCTI shows good performance in reproducing the first EOF mode PC time series for tropical cyclone track analysis.

1. 서론

2. 연구 자료 및 방법

2.1 연구자료

2.2 분석방법

3. 연구내용

3.1 태풍 진로의 시공간적 변동성 분석

3.2 태풍 진로와 관측 자료의 해양･ 대기 변동성 분석

3.3 태풍 진로와 MME 자료의 해양･대기 변동성 분석

3.4 태풍 진로 계절 예측을 위한 지수 개발

4. 결론 및 토론