The structural equation modeling constructed for runoff change attribution analysis outperforms traditional methods

Recently, with the increasing global warming and intensity of human activities, runoff has significantly decreased. This has had a severe impact on water resources development and utilization, as well as ecological and environmental construction. To understand hydrological cycles and promote sustainable utilization of water resources, it is crucial to conduct attribution analysis of runoff changes. This study employed traditional models, including empirical statistics, elasticity coefficients, hydrological modeling, and introduced a simple and efficient model – structural equation modeling (SEM) – to quantify the impact of various factors on runoff changes. The findings revealed a notable decreasing trend in natural runoff in the Shiyang River from 1960 to 2020, while precipitation, snowmelt, potential evapotranspiration (ET0), and temperature exhibited increasing trends. The integrated the results from multiple models highlighted that human activities emerged as the primary driver influencing runoff changes. From 1972 to 2020, the factors contributing to runoff reduction were ranked in the order: water withdrawal (50 %) > LUCC (20 %) > ET0 (16 %) > temperature (8 %). Factors contributing to runoff increase were ranked: precipitation = water diversion project (20 %) > snowmelt (6 %). The impact of reservoir regulation on runoff initially exhibited a negative correlation (impact rate of 10 %) and later turned positive (impact rate of 6 %). The SEM model, employed for runoff change attribution analysis, demonstrated a good fit and was consistent with results obtained from the SWAT model. Compared to traditional methods, SEM not only allowed for quantitative analysis of the factors’ impact on runoff but also explored the interactions between factors and the direct and indirect effects of factors on runoff. This study introduces innovative methods for the attribution analysis of runoff changes, offering valuable insights crucial for the ecological sustainable development of water resources in arid and semi-arid regions.