A Multi-Dimensional Characterization of Cold Waves Affecting China and Their Variability

Cold waves are intense extratropical weather events characterized by rapid near-surface cooling. Their variability is commonly described in terms of frequency or intensity, but other intrinsic attributes of their life cycles remain poorly understood. To provide a more comprehensive portrait, we developed an objective event-based detection framework that characterizes cold waves from multiple dimensions, including but not limited to frequency, track, magnitude, intensity, duration, and spatial extent. We applied the framework to cold waves affecting China to investigate their climatological characteristics and long-term variability. During 1940–2021, a total of 1737 cold-wave events are identified, corresponding to a mean frequency of 21.2 yr⁻¹. Their seasonal cycle and spatial distribution broadly align with previous knowledge, confirming the framework’s reliability, while revealing several new features. Two preferred track belts are identified: a west–east belt extending from Xinjiang Uygur Autonomous Region to North China and a meridional belt from North China to South China. It revises the traditional view of central Siberia as source regions for cold waves affecting China and identifies three additional candidate regions over Mongolia, eastern China, and the northeastern Indian Peninsula. The multidimensional characteristics show clear seasonal contrasts. Winter events are moderately intense but persist the longest and affect the largest regions, whereas autumn events are the most intense, and spring events are the most frequent but shorter-lived and more localized. Temporal analysis reveals pronounced decadal variability and an overall decline in frequency, duration, spatial extent, and intensity, accompanied by a slight increase in magnitude, with substantial regional and seasonal differences. An intriguing decadal correspondence is found between a weak East Asian winter monsoon and larger cold-wave magnitude, higher intensity, and shorter duration. These results demonstrate the value of describing and understanding cold-wave activity multidimensionally and highlight our framework’s ability to fulfill this need.