North China was hit by a devastating drought in the summer of 2015, affecting about 21 million people and 3.4 million hectares crops in seven provinces. The direct economic loss reached up to 11.48 billion RMB. Meanwhile, a super El Ni？o was developing, which resulted in widespread droughts and floods across the globe. With a good prediction of the 2015/16 super El Ni？o, NCEP’s Climate Forecast System version 2 (CFSv2) roughly captured the extreme summer drought over North China. This raises a question of whether the 2015/16 super El Ni？o help the forecasting of the 2015 extreme summer drought. 

During June-July 2015, 24 missions of artificial precipitation were implemented to relieve severe drought in Dalian, North China’s Liaoning province. (Source: news.xinhuanet.com)  

“A strong El Ni？o does not necessarily result in a higher predictability of extreme drought,” Dr. Xing YUAN and his colleagues from Institute of Atmospheric Physics answered this question in their recently published research in Scientific Report, and meanwhile they think that “the occurrence of North China drought depends on whether the low-latitude precursor (e.g., El Ni？o) evolves synergistically with high-latitude precursor (e.g., Eurasian spring snow cover reduction) to trigger a positive summer Eurasian teleconnection (EU) pattern that favors anomalous northerly and air sinking over North China, weakens the East Asian summer monsoon, and thereby reduces the moisture transported from the south. So a successful seasonal forecasting of the North China summer drought relies on whether the model captures the EU pattern.” 

According to Dr. YUAN, almost all CFSv2 ensemble members predicted 2015 strong El Ni？o quite well, but some missed the North China summer drought when they failed to capture the EU circulation pattern. YUAN’s studies suggest that a dynamical-statistical forecasting approach that combines both the low- and high-latitudes precursors enhances the predictability of extreme droughts over North China. And therefore it is more skillful than the dynamical forecasting at long lead.