Characteristics and subseasonal prediction of four types of cold waves in China
摘要
Cold waves are the most frequent winter disaster in China. Using Rotated Empirical Orthogonal Function (REOF) analysis, we categorize winter cold waves in China during 1983–2022 into four types based on their occurrence areas: East China (EC), Northwest China (NW), North China (NC), and Northeast China (NE). Each type exhibits distinct spatial–temporal characteristics and atmospheric circulation patterns. Prior to each cold wave outbreak, a baroclinic structure forms in the middle and lower troposphere, with low-pressure systems in central regions. As the cold wave develops, an anomalous high-pressure system moves from the northwest, replacing the low-pressure system but with varying intensity among the four regions. Notably, NW-type cold waves display negative sea level pressure anomalies over the polar region, linked to a weak positive Arctic Oscillation (AO) phase. Whereas, the EC, NC, and NE types of cold waves generally occur during a negative AO phase. The size and intensity of the Siberian High and Aleutian Low also vary among them, with the EC-type cold waves having the strongest values. Subseasonal prediction skills for these types of cold waves are assessed using the hindcasts of NUIST CFS1.1 model, which shows good prediction skill one pentad in advance but struggles beyond two pentads. At 2-pentad lead, the model starts to underestimate atmospheric circulation anomalies—particularly the Siberian High and northerly winds—resulting in insufficient cold air from north and weaker cold advection, leading to underestimated cold waves.