Long-range statistical forecasting of Korean summer precipitation

Abstract

We examined long-range statistical forecasting methods for Korean summer precipitation (KSP). We reviewed existing literature on the East Asian summer monsoon to develop a background on current KSP research and on the relationship of KSP to climate variations. Second, we explored interannual variability of KSP using composite and correlation analyses. We found that circulation anomalies in the spring prior to the monsoon in the tropical northwest Pacific alter sea surface temperatures (SST). These SST anomalies then persist into the following summer, leading to summer circulation anomalies that alter the flow into Korea and the precipitation on the seasonal scale. From this relationship, we developed a seasonal forecasting index. Third, we looked at KSP on the intraseasonal scale, to develop statistical forecast methods with five to thirty day leadtimes. We found that the Korean summer monsoon onset, break and withdrawal are positively correlated to the El NinÌ o / La NinÌ a state. We found that the Madden-Julian Oscillation (MJO), when conditioned with our seasonal index, showed skill in forecasting with lead times out to 20 days. Last, we found that tropical cyclone activity in Korea is impacted by ENSO on the interannual scale, and MJO on the intraseasonal scale.