Xinanjiang-Based Interval Forecasting Model for Daily Streamflow Considering Climate Change Impacts

Growing streamflow uncertainty, which is especially evident in high and cold regions, is one indication of climate change impacts on the hydrological cycle. In this work, an interval forecasting model that couples a snowmelt module and an uncertainty module is established based on the Xinanjiang model. The proposed model can consider climate change impacts by quantifying streamflow variations in the form of interval forecasts. Its performance was assessed by applying it in the headwater region of the Yellow River Basin. Interval forecasts and uncertainty analyses were conducted, and results show that the model can accurately describe the daily streamflow process in the study area. Unlike the deterministic forecasting model, the interval forecasting model effectively addresses shortcomings in forecasting high-flow scenarios. Furthermore, uncertainty analyses indicate that the model parameter K (the ratio of potential evapotranspiration to pan evaporation) plays a crucial role in water balance computations; the model parameter B (exponent of the distribution of the soil tension water capacity curve) exhibits sensitivity, suggesting challenges in attaining complete soil saturation across the entire basin. The insensitivity of the snowmelt module parameters implies that the proportion of snowmelt streamflow is relatively low in the annual streamflow and remains stable. The results can provide theoretical references for assessing the uncertainty of streamflow variations and for reservoir regulation in the Yellow River Basin.