A Deep‐Learning Ensemble Method to Detect Atmospheric Rivers and Its Application to Projected Changes in Precipitation Regime

This study aims to detect atmospheric rivers (ARs) around the world by developing a deep‐learning ensemble method using AR catalogs of the ClimateNet data set. The ensemble method, based on 20 semantic segmentation algorithms, notably reduces the bias of the testing data set, with its intersection over union score being 1.7%–10.1% higher than that of individual algorithms. This method is then applied to the Coupled Model Intercomparison Project Phase 6 (CMIP6) datasets to quantify AR frequency and its related precipitation in the historical period (1985–2014) and future period (2070–2099) under the Shared Socioeconomic Pathways 5–8.5 warming scenario. The six key regions, which are distributed in different continents of the globe and greatly influenced by ARs, are particularly highlighted. The results show that CMIP6 multi‐model mean with the deep‐learning ensemble method reasonably reproduces the observed AR frequency. In most key regions, both heavy precipitation (90–99 percentile) and extremely heavy precipitation (>99 percentile) are projected to increase in a warming climate mainly due to the increased AR‐related precipitation. The AR contributions to future heavy and extremely heavy precipitation increase range from 145.1% to 280.5% and from 36.2% to 213.5%, respectively, indicating that ARs should be taken into account to better understand the future extreme precipitation changes. Plain Language Summary Atmospheric rivers (ARs) play a key role in determining the global water cycle and regional precipitation in midlatitudes. Identifying their occurrences and possible changes is therefore critical to understanding the hydroclimate and its change. However, there exist different AR detection tools applying different methods or using different thresholds. The disparities between the outputs of these methods are large and often difficult to reconcile. This study first develops a deep‐learning ensemble method to identify ARs and then applies it to study the AR frequency and its related precipitation in the state‐of‐the‐art climate model simulations for historical and future climates. The results show that AR frequency is reasonably well reproduced by the multi‐model mean. More importantly, it is projected to increase in the future at both the global and regional scales. The six key regions, in which precipitations are robustly influenced by ARs, show that both heavy and extremely heavy precipitation are projected to increase, primarily due to the enhanced