Spatiotemporal evolution characteristics and attribution analysis of baseflow in the source region of the Yellow River

Baseflow refers to the portion of streamflow that is primarily sustained by groundwater and other delayed flow components. It constitutes a critical part of river discharge and plays an essential role in maintaining streamflow continuity during dry seasons, regulating hydrological processes, and supporting the health of riverine ecosystems. In this study, we focus on the headwaters of the Yellow River—one of the most important components of the so-called “Asian Water Tower”—which are characterized by high elevations, complex hydrogeological conditions, seasonal permafrost, and a fragile ecological environment. By applying baseflow separation techniques, the Mann–Kendall non-parametric trend test, baseflow signatures, and Pearson correlation analysis, a multidimensional analytical framework is constructed to comprehensively assess the temporal dynamics of baseflow over a long-term period. The key findings are as follows: (1) Baseflow volume in the source region of the Yellow River has generally exhibited an increasing trend over the study period. Compared to the reference period of 1985–1999, baseflow during 2000–2012 increased at the Huangheyan and Jimai stations, while it decreased at the Maqu, Jungong, and Tangnaihai stations. From 2013 to 2020, however, all five stations recorded notable increases in baseflow volume, with growth rates exceeding 14% at all stations. All five stations exhibited increasing trends in annual baseflow. Among them, the Jimai station showed a statistically significant upward trend, while the increasing trends at the other four stations were not statistically significant. Tangnaihai showed the highest annual increase (2.3 m³/s), whereas the smallest increase was observed at Huangheyan (0.15 m³/s), indicating spatial heterogeneity in baseflow responses across the basin. Seasonally, baseflow increased significantly at Jimai in spring, autumn, and winter, and in autumn at Huangheyan, whereas no significant seasonal increasing trends were observed at Maqu, Jungong, or Tangnaihai. (2) The magnitude baseflow signatures exhibited an overall increasing trend, although not statistically significant (p > 0.05). Regarding dynamic baseflow signatures, the highest baseflow index values were observed at the Jungong and Tangnaihai stations, with flood-season and non-flood-season indices of 0.49 and 0.46, and 0.61 and 0.60, respectively. These elevated values are attributed to favorable vegetation cover and the presence of seasonal permafrost, which enhance soil infiltration capacity and thereby promote baseflow generation. In contrast, the Huangheyan station recorded the lowest baseflow index values for both flood and non-flood seasons, likely due to the widespread presence of continuous permafrost, which constrains baseflow processes. Moreover, both the slope of the baseflow duration curve and the seasonal ratio exhibited a spatially increasing trend from northwest to southeast, with the highest values recorded at the Maqu station, reaching 2.41 and 4.21, respectively. (3) The response of baseflow signatures to environmental factors showed significant variability. Magnitude baseflow signatures were positively correlated with precipitation and soil moisture, while exhibiting the strongest negative correlation with potential evapotranspiration. In addition, the selection of monitoring stations, the length of the time series, and the temporal variability of precipitation all contributed to uncertainties in the estimated values of the baseflow index and the seasonal ratio. Due to the influence of anthropogenic factors such as drainage engineering and overgrazing, the source region of the Yellow River has experienced substantial grassland degradation and wetland shrinkage. These environmental changes have led to a marked decline in soil infiltration capacity, thereby weakening the continuity and stability of baseflow. The ability of baseflow to mitigate and delay drought impacts has been considerably reduced. This study contributes to a deeper understanding of hydrological response characteristics in the source region of the Yellow River. It provides a scientific basis for developing effective strategies to address climate change and for optimizing local water resource allocation. Furthermore, the findings have important implications for integrated water resources management and the promotion of sustainable development across the entire Yellow River Basin.