Scholarlink

<p>Accurate computation of daily reference evapotranspiration (ET0) is fundamental for ensuring the efficient and sustainable governance of water supplies. Consequently, this investigation aims to examine the efficacy of an Artificial Neural Network (ANN) architecture, optimized by two cutting-edge hybrid metaheuristic frameworks—the Innovative Gunner (AIG) and the Black Widow Spider Optimization (BWO)—for ET0 prediction. We assessed the AIG-ANN and BWO-ANN frameworks in estimating daily ET0 across four representative climatic regions in Iran: temperate-humid, cold-mountainous, hot-arid, and hot-humid. These novel hybrid methodologies were benchmarked against established paradigms, namely the WANN and FA-ANN models. Climatic inputs, comprising sunshine duration, maximal and minimal air temperatures (Tmax and Tmin), relative humidity, and air flow velocity, were sourced from four distinct meteorological stations spanning the decade from 2012 to 2022. Data partitioning allocated 2012–2019 for model calibration and 2019–2022 for validation, utilizing nine distinct hybrid input combinations. Model performance assessment relied on established statistical measures: the Coefficient of Correlation, Root Mean Square Error (RMSE), Mean Absolute Error, Normalized RMSERMSE (NRMSE), and the Nash–Sutcliffe Efficiency Index. The findings confirmed that all employed models exhibited superior predictive capability under hybrid input configurations. Notably, the statistical metrics confirmed that the AIG-ANN configuration yielded the highest predictive precision for ET0 across all four evaluated sites, signifying a substantial performance enhancement over conventional estimation techniques. This study substantiates that leveraging advanced metaheuristic enhancements, such as AIG, provides a highly accurate and robust mechanism for refining ET0 estimation, thereby directly facilitating improved regional water stewardship, optimized land management, and the formulation of resilient climate change adaptation strategies.</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

Estimating evaporation in diverse iranian climates using hybrid metaheuristic algorithms and ANN

  • 期刊论文

  • 发表日期：

    2025年12月29日

查看全文
  • H. Babaali,
  • R. Chamanpira,
  • R. Dehghani

摘要

Accurate computation of daily reference evapotranspiration (ET0) is fundamental for ensuring the efficient and sustainable governance of water supplies. Consequently, this investigation aims to examine the efficacy of an Artificial Neural Network (ANN) architecture, optimized by two cutting-edge hybrid metaheuristic frameworks—the Innovative Gunner (AIG) and the Black Widow Spider Optimization (BWO)—for ET0 prediction. We assessed the AIG-ANN and BWO-ANN frameworks in estimating daily ET0 across four representative climatic regions in Iran: temperate-humid, cold-mountainous, hot-arid, and hot-humid. These novel hybrid methodologies were benchmarked against established paradigms, namely the WANN and FA-ANN models. Climatic inputs, comprising sunshine duration, maximal and minimal air temperatures (Tmax and Tmin), relative humidity, and air flow velocity, were sourced from four distinct meteorological stations spanning the decade from 2012 to 2022. Data partitioning allocated 2012–2019 for model calibration and 2019–2022 for validation, utilizing nine distinct hybrid input combinations. Model performance assessment relied on established statistical measures: the Coefficient of Correlation, Root Mean Square Error (RMSE), Mean Absolute Error, Normalized RMSERMSE (NRMSE), and the Nash–Sutcliffe Efficiency Index. The findings confirmed that all employed models exhibited superior predictive capability under hybrid input configurations. Notably, the statistical metrics confirmed that the AIG-ANN configuration yielded the highest predictive precision for ET0 across all four evaluated sites, signifying a substantial performance enhancement over conventional estimation techniques. This study substantiates that leveraging advanced metaheuristic enhancements, such as AIG, provides a highly accurate and robust mechanism for refining ET0 estimation, thereby directly facilitating improved regional water stewardship, optimized land management, and the formulation of resilient climate change adaptation strategies.