Alongside mean increases in poleward moisture transport (PMT) to the Arctic, most climate models also project a linear increase in the interannual variability (IAV) with future warming. It is still uncertain to what extent atmospheric rivers (ARs) contribute to both the mean and the IAV increase of PMT. We analyzed large-ensemble climate simulations to 1) explore the link between PMT and ARs in the present-day (PD) and in two warmer climates (+2°C and +3°C compared to pre-industrial global mean temperature), 2) assess the dynamic contribution to changes in future ARs, and 3) analyze the effect of ARs on Arctic climate on interannual timescales. We find that the share of AR-related PMT (ARPMT) to PMT increases from 42% in the PD to 53% in the +3°C climate. The increase in AR-frequency and intensity is almost exclusively caused by significantly higher atmospheric moisture levels, while dynamic changes can regionally amplify or dampen the moisture-induced increase in ARs. The amount of ARs reaching the Arctic in any given region and season strongly depends on the regional jet stream position and speed southwest of this region. Our results indicate that positive ARPMT anomalies are profoundly linked to increased surface air temperature and precipitation, especially in the colder seasons, and have a predominantly negative effect on sea ice. AR events are expected to strongly affect Arctic climate variability in the future, when any AR-induced temperature anomaly occurs in an already warmer Arctic and a larger share of precipitation falls as rain.

The Southern Ocean is responsible for the majority of the global oceanic heat uptake that contributes to global sea level rise. At the same time, ocean temperatures do not change at the same rate in all regions and sea level variability is also affected by changes in salinity. This study investigates ten years of steric height variability (2008 to 2017) in the Southern Ocean (30°S to 70°S) by analysing temperature and salinity variations obtained from the GLORYS-031 model provided by the European Copernicus Marine Environment Monitoring Service (CMEMS). The thermohaline variability is decomposed into thermohaline modes using a functional Principal Component Analysis (fPCA). Thermohaline modes provide a natural basis to decompose the joint temperature-salinity vertical profiles into a sum of vertical modes weighted by their respective principal components (PCs) that can be related to steric height variability. Interannual steric height trends are found to differ significantly between subtropical and subpolar regions, simultaneously with a shift from a thermohaline stratification dominated by the first ‘thermal’ mode in the north to the second ‘saline’ mode in the South. The Polar Front appears as a natural boundary between the two regions, where steric height variations are minimized. Despite higher melt rates and atmospheric temperatures, steric height in Antarctic waters (0-2000 m) has dropped since 2008 due to higher salt content in the surface and upper intermediate layer and partially colder waters, while subtropical waters farther north have mostly risen due to increased heat storage.