Exploring the potential processes controlling changes in precipitation–runoff relationships in non-stationary environments

Abstract

The influence of climate change and anthropogenic activities on precipitation–runoff relationships (PRRs) has been widely discussed. Traditional models assuming stationary conditions can lead to inaccurate streamflow predictions. To address this issue, we propose a Driving index for changes in Precipitation–Runoff Relationships (DPRR), identified as key PRR influencers, involving climate forcing, groundwater, vegetation dynamics, and anthropogenic influences. According to the quantitative results of inputting the candidate driving factors into a holistic conceptual model, the possible process explanations for changes in the PRR were deduced. This framework is validated across five sub-basins in the Wei River basin. Moreover, non-stationary hydrological processes were initially detected, and the nonlinear correlations among the factors were assessed. The results show that baseflow emerges as the primary factor positively influencing the PRR (enhancing the PRR) but with high uncertainty. Potential evapotranspiration plays a dominant role in driving negative PRR changes in the sub-basins which are characterized by a semi-arid climate and minor human interference. Vegetation dynamics negatively influence the PRR, with driving levels correlating with the scale of soil and water conservation engineering, displaying lower uncertainty. Anthropogenic influences, represented by impervious surface ratio (ISR), night-time light (NTL), and population density (POP), exhibit varying driving levels, with ISR having the strongest and most direct impact, closely linked to urbanization processes and scales. The temporal dynamics of driving factors computed by dynamic DPRR generally correspond with hydrological regime shifts in non-stationary environments. The study’s findings offer a comprehensive understanding of hydrological processes, enabling informed decision-making for the development of sustainable hydrological models.