Quantitative relationship between SOI and observed precipitation in southern Korea and Japan by nonparametric approaches

Several studies have tried to link southern oscillation index (SOI) and observed precipitation in Korea and Japan. However, so far no clear relationship between the SOI and the precipitation in the area has been found. In the present study, categorized SOI are used to reveal quantitative and statistically significant influence on monthly precipitation at Busan in Korea and at Fukuoka in Japan. Monthly precipitation data were transformed to cubic root and nonexceedance probability time series. Correlation between the categorized SOI and transformed precipitation was calculated by Kendall's τ and Pearson's r. The results show that significant correlation is at hand when using both approaches. Significant correlation at the 1% level was obtained with lag time of 4 months under the 'Strong La Niña' category (SOI>2) at both stations. The results show that southern Korea and southern parts of Japan are located in the same or very close influence range of SOI.