Monthly runoff and sediment fluxes in a typical arid and semiarid river basin: Trends, driving factors, and attribution analysis

Abstract River runoff and sediment fluxes exhibited pronounced temporal variability driven by both climatic and anthropogenic factors. This study investigated runoff and sediment dynamics in the Ulan Mulun River basin using statistical and signal decomposition methods to identify multiscale trends, dominant drivers, and attribution mechanisms. The results showed that annual runoff and sediment load exhibited declining but statistically nonsignificant trends from 1997 to 2020, with rates of −0.5 × 10 6 m 3 /yr and −5.9 × 10 4 t/yr, respectively. At the monthly scale, time series of runoff and sediment fluxes, along with associated environmental variables, were decomposed into six intrinsic mode functions (IMFs) and a residual component. The first three IMFs, with periods of 4.5, 8.5, and 14.6 months, indicate that short‐timescale oscillations dominate the temporal variability. Scale‐dependent and phase‐shifted responses were observed between runoff–sediment fluxes and environmental factors. Multiple wavelet coherence (MWC) revealed that precipitation was the primary driver of runoff, while runoff was the dominant control on sediment load. Over the annual scale, hydrological time series were segmented into three distinct periods: baseline period (P0: 1997–2005), first change period (P1: 2006–2015), and second change period (P2: 2016–2020). Attribution analyses consistently showed that human activities dominated the reduction in both runoff and sediment load during P1. In contrast, during P2, increasing precipitation became the main driver of increased runoff, while sediment load remained dominated by human activities, with their influence intensifying over time. This stage‐specific shift reveals a structural decoupling of sediment transport from natural hydrological regimes under intensive ecological restoration. Understanding the dynamic variations of runoff and sediment fluxes and their scale‐dependent relationships with environmental factors is crucial for assessing landscape stability and managing water and sediment resources in arid and semiarid river basins.