Extreme runoff variation and non-stationary frequency analysis based on external forcing and internal variability decomposition

In recent years, extreme runoff has been affected by increasing climate change, which causes non-stationary behaviors in extreme runoff series. Climate change is driven by external forcing and internal variability. However, the role of these two factors in runoff variability remains unclear. Taking the historical period as the baseline, this study employs four Single-Model Initial-Condition Large Ensembles (SMILEs) to investigate future changes in extreme runoff represented by annual maximum 1-day runoff (AM1R) over China and to evaluate the impacts of external forcing and internal variability on these changes. A decomposition-based non-stationary frequency analysis method is proposed to estimate the frequency changes of extreme runoff events, which incorporates components of runoff influenced by external forcing and internal variability. Two shared socioeconomic pathways (i.e., SSP2-4.5 and SSP5-8.5) are selected for the future. The results show that the catchments with increased AM1R are more than those with decreased AM1R under SSP-2.4.5 and SSP5-8.5 scenarios for all SMILEs, with the catchments showing decreased AM1R mainly in Qinghai-Tibet Plateau and northeastern China. The impact of external forcing on runoff is stronger than that of internal variability at more than 35% and 62% of catchments for all SMILEs under SSP2-4.5 and SSP5-8.5 scenarios, respectively. The catchments with significant trends of AM1R are mainly in the eastern Qinghai-Tibet Plateau under the SSP2-4.5 scenario, while those are mainly in Qinghai-Tibet Plateau and southwestern China under the SSP5-8.5 scenario. For changes in the frequency of extreme runoff events, corresponding to the 50-yr return level of AM1R in the historical period, the return period is projected to become shorter in at least 66% of catchments for all SMILEs under the two scenarios. The study indicates that extreme runoff events are likely to become more frequent in the future, which is important for the flood prevention policy.