Scholarlink

<p>Hydro-meteorological variables are essential drivers of the water cycle, and understanding their dynamics is crucial for effective water resources management and disaster mitigation. Conventional trend analysis methods typically assume sequence normality and independent, lacking flexibility to efficiently analyze non-stationary and nonlinear hydrological series. To overcome these limitations, this study proposes a Grey Wolf Optimization-enhanced Complete Ensemble Empirical Mode Decomposition with Adaptive Noise method for integrated trend-periodic analysis. Validation using synthetic datasets demonstrates superior performance over the traditional methods, achieving accuracy higher than 85% for sequences with complex autocorrelation and heterogeneous distributions, while identifying periodicities undetectable by other methods. Applied to the Yangtze River Basin, the method reveals (1) warming trend in annual mean temperature, (2) spatially heterogeneous precipitation changes i.e., increases in northwest and northeast, decreases in central and southwestern regions. (3) declining runoff trends in the mainstem and most tributaries, with significant periodicities at 2–3-year and 11-year intervals. With the trend-periodic analysis, the runoff extreme can be forecasted. This method achieves an accuracy rate of 79.31%, surpassing the wavelet decomposition method by 20.67% points (58.64%). Extrapolation indicates potential low-value runoff extremes during 2025–2027, suggesting sustained high frequency or intensity of severe droughts over the coming decade.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Grey Wolf optimization enhanced adaptive decomposition for trend periodic analysis of nonstationary and nonlinear hyrologic series

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月7日

查看全文
  • Jinbei Li,
  • Wei Ding,
  • Hao Wang

摘要

Hydro-meteorological variables are essential drivers of the water cycle, and understanding their dynamics is crucial for effective water resources management and disaster mitigation. Conventional trend analysis methods typically assume sequence normality and independent, lacking flexibility to efficiently analyze non-stationary and nonlinear hydrological series. To overcome these limitations, this study proposes a Grey Wolf Optimization-enhanced Complete Ensemble Empirical Mode Decomposition with Adaptive Noise method for integrated trend-periodic analysis. Validation using synthetic datasets demonstrates superior performance over the traditional methods, achieving accuracy higher than 85% for sequences with complex autocorrelation and heterogeneous distributions, while identifying periodicities undetectable by other methods. Applied to the Yangtze River Basin, the method reveals (1) warming trend in annual mean temperature, (2) spatially heterogeneous precipitation changes i.e., increases in northwest and northeast, decreases in central and southwestern regions. (3) declining runoff trends in the mainstem and most tributaries, with significant periodicities at 2–3-year and 11-year intervals. With the trend-periodic analysis, the runoff extreme can be forecasted. This method achieves an accuracy rate of 79.31%, surpassing the wavelet decomposition method by 20.67% points (58.64%). Extrapolation indicates potential low-value runoff extremes during 2025–2027, suggesting sustained high frequency or intensity of severe droughts over the coming decade.