Multi-Scale Variability and Linkages Between Runoff and Meteorological Factors in the Songhua River Basin

A systematic understanding of runoff evolution and its climatic drivers is crucial for water-resources management in mid- to high-latitude basins. Using runoff data from the Jiamusi station in the Songhua River Basin (1980–2022), this study applies the Mann–Kendall test, Pettitt test, Hurst exponent, and wavelet analysis to examine the spatiotemporal changes in runoff and its links with precipitation, temperature, and evapotranspiration. Results show marked spatial heterogeneity: temperature rises persistently, precipitation becomes more uneven and variable, and evapotranspiration increases steadily. On short time scales, runoff is mainly driven by precipitation with clear seasonal differences among climatic factors. To assess long-term influences, a random forest model is developed using data from 12 stations. Evapotranspiration emerges as the dominant driver, followed by temperature and sunshine duration. Overall, runoff may shift from gradual trends to threshold responses under climate change, which offering insights for water-resource optimization and climate-adaptive management in Northeast China.