Future Changes in Global Atmospheric Rivers Projected by CMIP6 Models

Understanding the present and future features of atmospheric rivers (ARs) is critical for effective disaster prevention and mitigation efforts. This study comprehensively assesses the performance of ARs in Phase 6 of the Coupled Model Intercomparison Project (CMIP6) models on both seasonal and interannual timescales within the historical period and investigates the future projection of ARs under different emission scenarios on a global scale. The multi-model mean results obtained using the PanLu detection algorithm consistently exhibit agreement with the observational AR climatology and capture interannual fluctuations as well as the relationships with large-scale drivers. The future projections reveal increased AR frequency, intensity, duration, and spatial extent and decreased landfall intervals with regional variations and seasonal fluctuations. Besides, the AR frequency increase will accelerate around the middle of the century, attributed to a non-linear rise in surface temperature. Furthermore, mid-latitude ARs are gradually shifting toward higher latitudes in both hemispheres under SSP585, with Greenland experiencing a substantial increase in AR frequency and AR-induced precipitation. The hydrological implications arising from more frequent ARs are manifested more prominently in AR-induced heavy precipitation (HP), with regions historically characterized by lower AR occurrence also receiving a higher percentage of precipitation from ARs. At last, an incremental decomposition highlights the dominant role of thermal effects and relatively limited contributions from dynamical effects in AR changes. Besides, the interplay between regionally divergent temperature amplification results in different dynamically driven AR responses across the globe. Understanding atmospheric rivers (ARs) is crucial for disaster prevention and mitigation. This study evaluates the performance of CMIP6 simulation on historical ARs on seasonal and interannual timescales, assesses their future projection under different emission scenarios, and examines their hydrological implications. The seasonal and interannual AR patterns in models are consistent with the observations. Future projections reveal that ARs will become more frequent, intense, and longer-lasting while the time between landfall AR events will be shorter, with the magnitude varying by region and season. Interestingly, AR frequency is expected to increase even faster after the middle of the century due to a non-linear rise in Earth surface temperature. The study also shows that mid-latitude ARs are shifting toward higher latitudes, bringing more precipitation to the polar regions. The changes in AR characteristics can lead to increased risks of heavy precipitation, especially in regions where ARs are relatively not frequent in the past time. The study also shows that the changes in AR characteristics are primarily contributed by the increase in atmospheric moisture. The contribution from the changes in wind patterns is limited and varies across different regions as a result of regionally divergent temperature amplification due to global warming. CMIP6 simulates atmospheric rivers (ARs) well in seasonal and interannual timescales globally ARs will play a more significant role in the hydrological cycle as global warming alters various AR characteristics Thermal effects dominate AR changes globally, while dynamical effects associated with temperature rise vary regionally