A Gauge-Based Monthly Natural Discharge Reconstruction Dataset for the Eurasian Arctic, 1952–2025

Abstract. River discharge provides one of the clearest measures of how Arctic freshwater systems respond to climate change, yet long-term and continuous monitoring remains challenging. Here we present a gauge-based monthly natural discharge reconstruction dataset for 94 gauging stations in the Eurasian Arctic from 1952 to 2025. The dataset was generated using a basin-scale framework driven by precipitation and air temperature and aided by snow information. Terrestrial water storage (TWS) serves as the key link in the framework. Historical TWS was first reconstructed using an improved state-update empirical model constrained by TWS observations from the Gravity Recovery and Climate Experiment (GRACE), and monthly natural discharge was then estimated from reconstructed TWS using a refined runoff–TWS relationship calibrated against gauge discharge observations. The reconstructed monthly natural discharge agreed well with observations at minimally regulated gauges and during pre-regulation periods at regulated gauges, with median monthly Kling–Gupta efficiency (KGE) values exceeding 0.88. When monthly values were aggregated to annual discharge, the reconstruction still captured observed variability at minimally regulated gauges, with a median annual KGE of 0.73. The dataset generally agreed better with gauge observations than discharge estimates derived from existing GRACE-like TWS products and an existing gridded runoff reconstruction product. Example applications illustrate spatially heterogeneous trends in annual discharge and seasonal allocation over the past seven decades and show how the dataset can be used to assess the effects of human regulation on river discharge. This dataset provides a long-term natural-discharge baseline for characterizing historical discharge variability and assessing the roles of climate variability and human regulation in poorly monitored Arctic basins. The reconstructed monthly natural discharge dataset is available at https://doi.org/10.5281/zenodo.21157816 (Liu et al., 2026).