Quantifying Pan-Arctic Freshwater Fluxes: A 20-Year Satellite-Based Daily River Discharge and Runoff Dataset

The Arctic region is undergoing a rapid and intense transformation driven by global climate change. Paradoxically, this coincides with a generalized decline in the density of hydrometric stations, resulting in fragmented spatiotemporal river discharge time series data that are insufficient to capture the complexity of ongoing dynamics. Under these challenging context, Arctic basins play a crucial role in regulating the freshwater budget, influencing ocean circulation and sea ice formation, and acting as a "litmus test" for the hydrological cycle's response to warming.

To address the scarcity of in-situ data, our work aims to provide continuous river discharge and runoff estimates at a daily scale and 0.25° spatial resolution for the entire continental Pan-Arctic region over the period 2003–2022. We employ STREAM model (SaTellite based Runoff Evaluation And Mapping; Camici et al., 2022), a semi-distributed conceptual hydrological model forced exclusively by temperature data and satellite observations, including precipitation, soil moisture, snow cover fraction, and Terrestrial Water Storage (TWS) anomalies from GRACE (Gravity Recovery and Climate Experiment) and its Follow-On mission (GRACE-FO). The integration of gravimetry data represents a key innovation—particularly relevant in the context of the future NGGM-MAGIC (Next-Generation Gravity Mission / Mass-change And Geophysics International Constellation) mission—as it enhances the characterization of hydrological processes in cold regions where TWS changes significantly drive river discharge and runoff variability. The model was first calibrated on 15 "donor" Arctic basins, achieving a median Kling-Gupta Efficiency index (KGE) of 0.80. To cover ungauged areas, we developed a regionalization framework based on aridity-index clustering, extending estimates to the entire Pan-Arctic domain. The resulting dataset was independently validated against 26 gauging stations and benchmarked against existing reanalysis products.

Results demonstrate that the regionalized model faithfully reproduces discharge seasonality and interannual variability over 70% of the Pan-Arctic area. Furthermore, trend analysis reveals statistically significant runoff trends in 18% of the domain, highlighting that the Pan-Arctic does not exhibit a uniform response to climate change, but rather diverse, localized reactions.

This work provides a consistent hydrological baseline based solely on satellite data, filling the gaps left by fragmented in-situ river discharge monitoring networks and offering a robust tool to investigate the interactions between climate change and hydrological extremes in the Pan-Arctic region, a critical climate hotspot.