Asymmetry in the Prediction Skill of Spring Droughts and Pluvials Over Northeast Asia

Abstract

Skillful seasonal prediction of extreme precipitation is crucial to mitigate socioeconomic damage, but current climate forecast models often failed to predict seasonal droughts and pluvials. This study aims to evaluate the prediction skill of springtime (March through May) precipitation over Northeast Asia (NEA) in one lead month. Over 1992–2010, the role of tropospheric jet streams in predicting NEA pluivals and droughts is examined using fully-coupled climate forecast model (CFM) simulations and atmospheric global circulation model (AGCM) simulations forced by observed SST anomalies. Results show a relatively better prediction skill of spring pluvials than that of spring droughts from both CFMs and AGCMs, indicating asymmetry in the seasonal prediction skill of extreme precipitation over NEA. A relatively low prediction skill of spring precipitation deficits in fully-coupled climate forecast models is caused by the warm (cold) biases of sea surface temperature anomalies (SSTA) over the northern Pacific Ocean in droughts (pluvials). This result suggests that the reduced SSTA biases over the northern Pacific Ocean are necessary for a systematic prediction skill of spring droughts and pluvials over NEA.