The Arctic freshwater cycle is changing rapidly, which will require adequate monitoring of river flows to detect, observe, and understand changes and provide adaptation information. There has, however, been little detail about where the greatest flow changes are projected, and where monitoring therefore may need to be strengthened. In this study, we used a set of recent climate model runs and an advanced macro‐scale hydrological model to analyze how flows across the continental pan‐Arctic are projected to change and where the climate models agree on significant changes. We also developed a method to identify where monitoring stations should be placed to observe these significant changes, and compared this set of suggested locations with the existing network of monitoring stations. Overall, our results reinforce earlier indications of large increases in flow over much of the Arctic, but we also identify some areas where projections agree on significant changes but disagree on the sign of change. For monitoring, central and eastern Siberia, Alaska, and central Canada are hot spots for the highest changes. To take advantage of existing networks, a number of stations across central Canada and western and central Siberia could form a prioritized set. Further development of model representation of high‐latitude hydrology would improve confidence in the areas we identify here. Nevertheless, ongoing observation programs may consider these suggested locations in efforts to improve monitoring of the rapidly changing Arctic freshwater cycle.

Pan‐Arctic riverine dissolved organic carbon (DOC) fluxes represent a major transfer of carbon from land‐to‐ocean, and past scaling estimates have been predominantly derived from the six major Arctic rivers. However, smaller watersheds are constrained to northern high‐latitude regions and, particularly with respect to the Eurasian Arctic, have received little attention. In this study, we evaluated the concentration of DOC and composition of dissolved organic matter (DOM) via optical parameters, biomarkers (lignin phenols), and ultrahigh resolution mass spectrometry in the Northern Dvina River (a midsized high‐latitude constrained river). Elevated DOC, lignin concentrations, and aromatic DOM indicators were observed throughout the year in comparison to the major Arctic rivers with seasonality exhibiting a clear spring freshet and also some years a secondary pulse in the autumn concurrent with the onset of freezing. Chromophoric DOM absorbance at a₃₅₀ was strongly correlated to DOC and lignin across the hydrograph; however, the relationships did not fit previous models derived from the six major Arctic rivers. Updated DOC and lignin fluxes were derived for the pan‐Arctic watershed by scaling from the Northern Dvina resulting in increased DOC and lignin fluxes (50 Tg yr⁻¹ and 216 Gg yr⁻¹, respectively) compared to past estimates. This leads to a reduction in the residence time for terrestrial carbon in the Arctic Ocean (0.5 to 1.8 years). These findings suggest that constrained northern high‐latitude rivers are underrepresented in models of fluxes based from the six largest Arctic rivers with important ramifications for the export and fate of terrestrial carbon in the Arctic Ocean.

Observations show increases in river discharge to the Arctic Ocean especially in winter over the last decades but the physical mechanisms driving these changes are not yet fully understood. We hypothesize that even in the absence of a precipitation increase, permafrost degradation alone can lead to increased annual river runoff. To test this hypothesis we perform 12 millennium-long simulations over an idealized hypothetical watershed (1 km²) using a distributed, physically based water balance model (Water flow and Balance Simulation Model, WaSiM). The model is forced by both a hypothetical warming defined by an air temperature increase of 7.5 ^∘C over 100 years, and a corresponding cooling scenario. To assess model sensitivity we vary soil saturated hydraulic conductivity and lateral subsurface flow configuration. Under the warming scenario, changes in subsurface water transport due to ground temperature changes result in a 7%–14% increase in annual runoff accompanied by a 6%–20% decrease in evapotranspiration. The increase in runoff is most pronounced in winter. Hence, the simulations demonstrate that changes in permafrost characteristics due to climate warming and associated changes in evapotranspiration provide a plausible mechanism for the observed runoff increases in Arctic watersheds. In addition, our experiments show that when lateral subsurface moisture transport is not included, as commonly done in global-scale Earth System Models, the equilibrium water balance in response to the warming or cooling is similar to the water balance in simulations where lateral subsurface transport is included. However, the transient changes in water balance components prior to reaching equilibrium differ greatly between the two. For example, for high saturated hydraulic conductivity only when lateral subsurface transport is considered, a period of decreased runoff occurs immediately after the warming. This period is characterized by a positive change in soil moisture storage caused by the soil moisture deficit developed during prior cooling.