This study presents an analysis of the observed inter-annual variability and inter-decadal trends in river discharge across northern Canada for 1964–2013. The 42 rivers chosen for this study span a combined gauged area of 5.26 × 10⁶ km² and are selected based on data availability and quality, gauged area and record length. Inter-annual variability in river discharge is greatest for the eastern Arctic Ocean (coefficient of variation, CV = 16 %) due to the Caniapiscau River diversion into the La Grande Rivière system for enhanced hydropower production. Variability is lowest for the study area as a whole (CV = 7 %). Based on the Mann–Kendall test (MKT), no significant (p > 0.05) trend in annual discharge from 1964 to 2013 is observed in the Bering Sea, western Arctic Ocean, western Hudson and James Bay, and Labrador Sea; for northern Canada as a whole, however, a statistically significant (p < 0.05) decline of 102.8 km³ 25 yr⁻¹ in discharge occurs over the first half of the study period followed by a statistically significant (p < 0.05) increase of 208.8 km³ 25 yr⁻¹ in the latter half. Increasing (decreasing) trends in river discharge to the eastern Hudson and James Bay (eastern Arctic Ocean) are largely explained by the Caniapiscau diversion to the La Grande Rivière system. Strong regional variations in seasonal trends of river discharge are observed, with overall winter (summer) flows increasing (decreasing, with the exception of the most recent decade) partly due to flow regulation and storage for enhanced hydropower production along the Hudson and James Bay, the eastern Arctic Ocean and Labrador Sea. Flow regulation also suppresses the natural variability of river discharge, particularly during cold seasons.

Lying in the frontline of the prevailing midlatitude westerlies, British Columbia and southeastern Alaska (BCSAK) often receive copious amounts of precipitation through atmospheric rivers (ARs). This study quantifies the contribution of ARs to annual, seasonal, and extreme precipitation across BCSAK from 1979 to 2012 using a recently developed high‐resolution gridded precipitation data set, a regional AR catalog, and integrated vapor transport fields calculated from a reanalysis data set. On average, ARs contribute 13% of total annual precipitation with the higher contribution along the coastal regions (up to 33%), parts of which are one of the wettest locations on Earth, followed by the Columbia and Rocky Mountains (~9%–15%). The highest contributions occur during September (up to 57%) and October (up to 49%). The contribution of ARs to extreme precipitation attains >90% along the western arc of the Coast Mountains and the coastal regions of BCSAK. ARs act as the main synoptic‐scale mechanism that brings rainfall to the Rocky Mountains in winter. The probability of observing AR‐related precipitation increases over the study period; however, no change occurs in the average AR‐related precipitation amount for most of BCSAK during 1979–2012. This work provides insights on the critical role ARs play on the water resources of northwestern North America and has broader implications on community water supply and management, hydropower operations, and flood risk assessment and mitigation.

Snow‐dominated watersheds are bellwethers of climate change. Hydroclimate projections in such basins often find reductions in annual peak runoff due to decreased snowpack under global warming. British Columbia's Fraser River Basin (FRB) is a large, nival basin with exposure to moisture‐laden atmospheric rivers (ARs) originating in the Pacific Ocean. Landfalling ARs over the region in winter are projected to increase in both strength and frequency in Coupled Model Intercomparison Project Phase 5 (CMIP5) climate models. We investigate future changes in hydrology and annual peak daily streamflow (APF) in the FRB using a hydrologic model driven by a bias‐corrected CMIP5 ensemble. Under RCP8.5, the FRB evolves toward a nival‐pluvial regime featuring an increasing association of extreme rainfall with APF, a doubling in cold season peak discharge, and a decrease in the return period of the largest historical flow, from a 1‐in‐200 year to 1‐in‐50 year event by the late 21^st century.