High spatial and temporal resolution models are essential for understanding future climate impacts and developing effective climate resilience plans. However, existing regional and global river models often lack the resolution needed to accurately capture local conditions. This study uses a series of high-resolution models, including the Regional Arctic System Model, mizuRoute, and the river basin model, to analyze Arctic and sub-Arctic Alaskan hydrology. We compare a historical baseline (1991–2020) with six midcentury (2035–64) futures: two pseudo–global warming scenarios based on historical meteorology and four direct dynamically downscaled global climate models. The six futures reveal significant uncertainty in future annual discharge and peak flows, although a widespread increase in discharge during April (+63%) and October (+31%) is consistently shown across models. Projected increases in rain and shifting weather patterns lead to a transition from snow to rain in spring and autumn, reducing the fraction of snowmelt contributing to river discharge. Rising evapotranspiration moderates discharge changes, particularly in autumn, by offsetting precipitation increases. Average summer river temperatures are projected to increase by approximately 1.5°C, doubling the number of river segments that experience 18°C days, a critical threshold for salmon survival, and intensifying the heat flux to the ocean adding an average of 3.3 × 10¹² MJ yr⁻¹. These changes in the hydrologic cycle could profoundly impact riverine and oceanic ecosystems, posing substantial challenges to communities reliant on these environments.