Spatio-temporal patterns of hot extremes in China based on complex network analysis

In the face of escalating frequency and severity of hot extreme (HE) events worldwide, understanding their spatio-temporal characteristics and hazard patterns has become crucial. This study employs a complex network (CN) approach, specifically using visibility graph and similarity network analysis, to investigate HEs. According to the HE network, we have successfully identified anomalous years, divided stages of change, selected representative cities, and zoned spatial hazard patterns of HE. Results reveal that 85% of cities in China experienced varying degrees of increasing HEs, with the highest increase observed as 63 times. The HE networks in China exhibit small-world characteristics, allowing the classification of HE changes into 5–8 stages and 10 types. Hefei emerges as the most representative city in this context. Additionally, the hazard of HE in China can be divided into four grades, with a gradual increase from north to south. This study sheds light on the intensifying hot extreme events in China and establishes a connection between CN and HE analysis, offering innovative ideas and methods for studying climate extremes. Graphical abstract