Quantifying the contributions of multi-source uncertainty to multi-time scale indicators in hydrological modeling under climate change

Abstract

Study region: The upper Heihe River basin (UHRB) in Northwest China and the upper Yalongjiang River basin (UYRB) in Southwest China. Study focus: Based on the monthly-scale improved WASMOD model, we projected the future hydrological changes over the study area in three periods: near future (NF, 2026–2050), medium future (MF, 2051–2075), and far future (FF, 2076–2100), and evaluated the uncertainty in the future hydrological processes under climate change by using the time-series analysis of variance (ANOVA), focusing on the contributions of six source uncertainty to three hydrological indicators annually and at 25-, 50-, and 70-year time scales. Additionally, we investigated the variations in uncertainty contributions over time scales. New hydrological insights for the region: The results show that the variation in the future multi-year average flows is +7.5%, +4.3%, and +0.4% for the UHRB and +1.7%, -2.2%, and -6.3% for the UYRB in the NF, MF, and FF periods, respectively. The uncertainty evaluation indicates the following: (i) the contribution of internal variability is relatively high annually and decreases over time, reaching a stable level when the time scale exceeds 30 years. (ii) The effects from general circulation models (GCMs) and future scenarios (Shared Socioeconomic Pathway emission scenarios (SSPs)) should be prioritized in predicting future mean or high flows in both watersheds, especially at longer time scales. (iii) Hydrological models with different potential evaporation module (HM-PEs) typically have a greater effect on low flows than on mean and high flows, and the uncertainty associated with the calibration strategies (objective functions (OFs)) should be considered at longer time scales. Therefore, the potential impacts of these uncertainty sources must be carefully considered in policy-making for future water resources management.

Keywords

• Climate change
• Hydrological modeling
• Uncertainty quantification
• Time scale