资源环境科技发展态势分析平台

Substantial efforts have been made in recent decades to understand the characteristics and variations of summer hot extremes and the underlying physical processes involved. However, the seasonal prediction of summer hot extremes remains challenging. The populous middle and lower reaches of the Yangtze River valley (MLYR) in China are severely influenced by hot extremes during summer. This study presents seasonal predictions of summer hot extremes over the MLYR for the period 1979-2016 based on three preceding predictors that are closely linked to hot extremes over this region: spring soil moisture over the southeastern Indochina Peninsula (SIP); spring sea surface temperature (SST) over the western tropical Pacific (WTP); and the difference of Nino 3.4 SST data in May and the previous December. The soil moisture deficit over the SIP and warm SST over the WTP in spring, as well as the difference of Nino 3.4 SST data in May and the previous December, tend to result in positive geopotential height anomalies over the MLYR, which may favor hot extremes by enhancing downward solar radiation, subsidence warming and local soil moisture-temperature coupling associated with precipitation reduction. Using these three predictors, we demonstrate with cross validation that the temporal variations of hot extremes over the MLYR can be skillfully predicted for the study period (i.e., 1979-2016), while biases exist in the magnitude. Hindcast experiments for 2012-2016 show that high prediction skill can be achieved for the spatial patterns of hot extremes, with pattern correlation coefficients of 0.83-0.99. Our findings are expected to facilitate the practical prediction of hot extremes over the MLYR.