Revisiting spatio-temporal variation in compound dry-hot extreme events in China with newly developed nonstationary index

Abstract

Compound dry-hot extreme events (CDHEs), as the most typical compound extreme events, bring more harm to human society than single extreme events. Traditional indicators based on stationary assumptions of hydrometeorological variables for CDHEs detection may no longer be valid due to anthropogenic and climate change impacts. The nonstationary hydrometeorological series has been extensively studied but rarely considered in identifying CDHEs. Therefore, this paper develops a nonstationary compound dry-hot index (NCDHI) with climate index and anthropogenic factors as covariates, to revisit CDHEs in China from 1961 to 2020 using the Generalized Additive Models for Location, Scale, and Shape (GAMLSS) model. The results show that the nonstationary model is better than the traditional stationary model in fitting precipitation and temperature series. Validation using typical disaster events and losses data reveals a higher correlation between the NCDHI and actual disaster losses, confirming the good applicability of the NCDHI in China. Areas affected by CDHEs of varying severity have increased in China during the study period. Meanwhile, the severity of CDHEs has also been exacerbated, with more severe in the central and eastern regions. Furthermore, CDHEs in the western regions, though less intense, occur more frequently. The proposed NCDHI can capture the characteristics of CDHEs in China, which provides a new idea for constructing a compound dry-hot index that can effectively adapt to environmental changes. The index can further improve the scientific understanding of compound extreme events' temporal and spatial patterns and provide a scientific basis for regional risk management and disaster prevention and mitigation.