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.

Selecting a reliable global climate model as the driving forcing in simulations with dynamic downscaling is critical for obtaining a reliable regional ocean climate. With respect to their accuracy in providing physical quantities and long-term trends, we quantify the performances of 17 models from the Coupled Model Inter-comparison Project Phase 6 (CMIP6) over the North Pacific (NP) and Northwest Pacific (NWP) oceans for 1979–2014. Based on normalized evaluation measures, each model’s performance for a physical quantity is mainly quantified by the performance score (PS), which ranges from 0 to 100. Overall, the CMIP6 models reasonably reproduce the physical quantities of the driving variables and the warming ocean heat content and temperature trends. However, their performances significantly depend on the variables and region analyzed. The EC-Earth-Veg and CNRM-CM6-1 models show the best performances for the NP and NWP oceans, respectively, with the highest PS values of 85.89 and 76.97, respectively. The EC-Earth3 model series are less sensitive to the driving variables in the NP ocean, as reflected in their PS. The model performance is significantly dependent on the driving variables in the NWP ocean. Nevertheless, providing a better physical quantity does not correlate with a better performance for trend. However, MRI-ESM2-0 model shows a high performance for the physical quantity in the NWP ocean with warming trends similar to references, and it could thus be used as an appropriate driving forcing in dynamic downscaling of this ocean. This study provides objective information for studies involving dynamic downscaling of the NP and NWP oceans.