Contrasting Interannual Prediction between January and February Temperature in Southern China in the NCEP Climate Forecast System

Abstract

AbstractThis study investigates the prediction of southern China surface air temperature (SAT) in January and February using hindcast and forecast dataset from the second version of the National Centers for Environmental Prediction Climate Forecast System, version 2 (NCEP CFSv2), for the period of 1983–2017. The observed January and February SAT in southern China is teleconnected with the Euro-Atlantic dipole (EAD) and the North Atlantic Oscillation (NAO), respectively. The February SAT is also teleconnected with El Niño–Southern Oscillation (ENSO) via the bridge with the Philippine Sea anticyclone. The CFSv2 better predicts southern China SAT in February than January, where the temporal correlation coefficients between the observed and predicted regional-mean SAT in February and January are +0.81 and +0.27 (+0.32 and +0.04), respectively, for the one-month (two month) ahead prediction. The better prediction in February coincides with 1) accurate responses of the Eurasian circulation and the Philippine Sea anticyclone to the NAO and the ENSO, respectively, and 2) a strong ENSO–NAO linkage. The poorer prediction in January is related to a stronger linkage of the predicted January SAT with the NAO rather than the EAD, as well as a weak ENSO–EAD linkage. These results advance our understanding of the subseasonal prediction of the winter temperature in southern China.