Climate prediction of summer extreme precipitation frequency in the Yangtze River valley based on sea surface temperature in the southern Indian Ocean and ice concentration in the Beaufort Sea

Three statistical prediction models for the summer extreme precipitation frequency (EPF) in the middle and lower reaches of the Yangtze River valley (MLYRV) based on the winter sea surface temperature in the southern Indian Ocean (SIO‐SST; Scheme‐SST), the spring sea‐ice concentration in the Beaufort Sea (Scheme‐SIC), and both predictors (Scheme‐SS), are established by using the year‐to‐year increment method. The winter SIO‐SST may affect the SST anomaly in the east of Australia via a teleconnection pattern. The east of Australia SST signal in winter continues until the following summer and then affects the summer EPF in the MLYRV via modulation of the meridional Hadley circulation in the western North Pacific. The positive Beaufort SIC anomaly may result in a negative summer SST anomaly by the increased surface albedo. In response, the atmospheric circulation presents a dipole anomaly distribution in the Beaufort Sea and the Barents Sea. The Arctic dipole anomaly may bring about frequent extreme precipitation in the MLYRV by adjusting the position of the East Asian westerly jet and Eurasian teleconnection pattern. The prediction skill of the summer EPF for Scheme‐SS is higher than that of Scheme‐SST and Scheme‐SIC in the cross‐validation test during 1962–2017 and the independent hindcast during 1992–2017. Scheme‐SS shows a higher prediction skill of the summer EPF than that of Scheme‐SST and Scheme‐SIC, with a time correlation coefficient of 0.62, accounting for 39% of the total variance, of which 35% is the winter SIO‐SST and 4% the spring Beaufort SIC. Scheme‐SS not only shows a rather high predictive ability for the anomalous summer EPF but can also reproduce the increasing trend of extremely heavy precipitation. The primary focus of this article is to build the climate prediction model for the summer extreme precipitation frequency (EPF) in the middle and lower reaches of the Yangtze River valley based on the preceding Indian Ocean SST and Arctic SIC using the year‐to‐year increment method. The prediction model can not only show the rather high predictive ability for the interannual variability of summer EPF but also reproduce the increasing trend of extreme precipitation. Figure shows Predicted (turquoise line) and observed (red line) JJA EPF DY (a) and EPF (b) in the cross‐validation tests for the period 1962–2017.