Detection and attribution of human influence on seasonal extreme precipitation in northern Hemisphere

Abstract

Fingerprinting analysis have detected the impact of human activities on annual precipitation extremes. Using simulations of selected climate models of the Coupled Model Intercomparison Project Phase 6 (CMIP6), this study has comprehensively demonstrated human influence on seasonal precipitation extremes of Northern Hemisphere Land (NHL) over different spatial scales from 1950 to 2014. By assessing the impacts of various climatic forcings on the maximum 1-day (Rx1day) and 5-day (Rx5day) precipitation indices, our results show that greenhouse gas (GHG) forcings predominantly drive the increase in observed Rx1day across most of the NHL in all four seasons, with more pronounced effects in fall and winter than in spring and summer. Furthermore, low-risk regions tend to experience greater GHG-induced Rx1day than high-risk regions in all seasons. Anthropogenic aerosol (AER) forcings significantly weaken Rx1day, particularly during winter in regions like India and southern China. Change point analysis reveals a rapid increase in Rx1day under GHG forcings and a slower decrease under AER forcing since the 1980s. However, the abrupt change in Rx1day under both anthropogenic and natural external (ALL) forcings generally began a decade later, around the 1990s. Using optimal fingerprinting techniques, we demonstrate for the first time that discernible anthropogenic forcings (ANT) impact at least one season in over 80% of CMIP6 subregions, with more than 60% of the contributions being attributable to ANT forcings. The number of subregions with detected ANT forcings is twice as high in winter compared to summer. Although seasonal natural (NAT) forcings are undetectable in one-signal analysis, they are detected in similar regions in both two- and three-signal analysis, suggesting that observed changes should be attributed to both anthropogenic and natural forcings.