海温异常对中国汛期降水的影响及预测应用

Review of the Influence and Application of SST Anomaly to Flood Season Precipitation in China

以厄尔尼诺-南方涛动(ENSO)循环和其他关键区海温对东亚季风关键系统影响机理的科学认知与应用为线索,回顾我国汛期降水业务发布预测的技巧。按照三类雨型划分,1981—2020年每10年的雨型预测正确率分别为50%、60%、50%、70%;按照四类雨型划分,1981—2020年每10年的雨型预测正确率分别为30%、30%、40%、50%,即我国汛期旱涝空间分布型的预测准确率明显提高。筛选预测准确率偏低且有显著洪涝发生年用于复盘,重点分析当年的主要预测依据和预测偏差较大的原因。结果表明:对海温影响东亚夏季风系统机理的有限认知影响很大,包括ENSO循环不同位相的影响、ENSO影响的不对称性、ENSO空间型的变化、印度洋等海域海温异常的影响均起重要作用,提出多因子多时间尺度协同作用理论、客观化预测方法、精细化监测预测影响评估一体化系统建设等有助于提高精准化预测能力和精细化服务水平。

The spatial distribution of precipitation anomalies during flood season and characteristics of drought and flood disasters in China are directly affected by the speed and stagnation of the East Asian summer monsoon(EASM).EASM is significantly affected by external forcing such as sea surface temperature,land surface processes,ice and snow cover,and internal dynamic anomalies of atmospheric circulation.The sea surface temperature(SST)anomaly and its evolution have always been important factors for predicting precipitation during the flood season,considering lead time and the strength of precipitation prediction in flood season.Based on the scientific understanding and application of the mechanism of El N in~o-southern oscillation(ENSO)cycle and other Ocean SST on the key factors of EASM,the prediction skill of flood season precipitation is reviewed.According to a prediction evaluation spanning over 40 years of historical records,the prediction accuracy for different types of rainfall pattern,the prediction accuracy of rain types in 1981-1990,1991-2000,2001-2010,and 2011-2020 is 50%/30%,60%/30%,50%/40%,and 70%/50%,respectively.In other words,the prediction of the primary rainfall patterns during the flood season in China is closer to the observation,and the accuracy of predicting spatial distribution patterns of drought and flood has significantly improved.This improvement can be attributed to the in-depth understanding of the impact of SST on EASM activities and enhancements made to dynamic climate models.In the history of flood season prediction,there have been both successful and unsuccessful cases.The years with low prediction accuracy and significant flooding events are as follows:1983,1991,1999,2003,and 2014.The primary basis for prediction is analyzed,revealing that the limited understanding of the mechanism of SST affecting the EASM had a great impact on the skill of precipitation predictions during the flood season.Among these factors,the influence of different phases of the ENSO cycle,the asymmetry of ENSO’s influence,the change in ENSO spatial patterns,and the influence of other local seas,such as the Indian Ocean SST anomaly,all play important roles.The importance of multi-factor and multi-scale synergy theory and application,as well as the technical support of the objectification method for prediction,are emphasized in summarizing causes for low prediction skill cases.Finally,some suggestions for improving future flood season precipitation predictions are put forward,and it is emphasized that the development of a multi-factor and multi-time scale synergistic theory,an objective climate prediction method,and an integrated system for monitoring,predictions and impact assessment will significantly enhance predictions and provide services for flood season precipitation.