Scholarlink

<p>Occurrence of natural disasters is significantly increased in recent years. However, drought is one of the persistent threats to life on Earth. Therefore, several studies in previous years worked for the management of drought. However, precipitation is a key determinant of drought, and global climate models (GCMs) are widely used for drought projections. Multi-model ensemble (MME) approaches aim to improve projection accuracy by integrating multiple GCMs, but existing methods often fail to capture non-linear relationships between observed and modeled precipitation data. This study proposes a novel MME approach, maximal information coefficient exponential transformation (<i>MICET</i>), to address non-linear dependencies and discrepancies between observed and simulated precipitation. Additionally, a new drought index, standardized maximized informed (SMI), is introduced, leveraging <i>MICET</i>-aggregated data for improved drought characterization. The study employs 18 GCMs historical simulated precipitation data of the coupled model intercomparison project phase 6 (CMIP6) spanning 1961–2014. <i>MICET</i> is evaluated against the traditional simple model averaging (SMA) approach using multiple performance criteria. The newly developed SMI index is used to assess drought characteristics across different temporal scales. <i>MICET</i> consistently outperformed SMA in all evaluation metrics, demonstrating its ability to effectively aggregate GCM outputs. The analysis of drought characteristics using the SMI index revealed an insignificant increase in drought trends for short time scales but a significant upward trend for longer time scales. Moreover, the results indicate that no drought (<i>ND</i>) conditions are the most probable state, with extreme wet and dry conditions being less frequent. The proposed <i>MICET</i> approach offers a robust solution for improving precipitation-based drought projections by addressing complex relationships in MME models. The newly introduced SMI index provides a more reliable representation of drought trends, emphasizing the increasing severity of long-term droughts due to climate change.&#xa0;</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

Development of a Novel Multi-Model Ensemble Weighting Scheme for Improved Drought Assessment

  • 期刊论文

  • 发表日期：

    2025年4月21日

查看全文
  • Mahrukh Yousaf,
  • Ali Iqbal,
  • Sadia Qamar,
  • Muhammad Shakeel,
  • Maryam Ilyas,
  • Zulfiqar Ali,
  • Rizwan Niaz

摘要

Occurrence of natural disasters is significantly increased in recent years. However, drought is one of the persistent threats to life on Earth. Therefore, several studies in previous years worked for the management of drought. However, precipitation is a key determinant of drought, and global climate models (GCMs) are widely used for drought projections. Multi-model ensemble (MME) approaches aim to improve projection accuracy by integrating multiple GCMs, but existing methods often fail to capture non-linear relationships between observed and modeled precipitation data. This study proposes a novel MME approach, maximal information coefficient exponential transformation (MICET), to address non-linear dependencies and discrepancies between observed and simulated precipitation. Additionally, a new drought index, standardized maximized informed (SMI), is introduced, leveraging MICET-aggregated data for improved drought characterization. The study employs 18 GCMs historical simulated precipitation data of the coupled model intercomparison project phase 6 (CMIP6) spanning 1961–2014. MICET is evaluated against the traditional simple model averaging (SMA) approach using multiple performance criteria. The newly developed SMI index is used to assess drought characteristics across different temporal scales. MICET consistently outperformed SMA in all evaluation metrics, demonstrating its ability to effectively aggregate GCM outputs. The analysis of drought characteristics using the SMI index revealed an insignificant increase in drought trends for short time scales but a significant upward trend for longer time scales. Moreover, the results indicate that no drought (ND) conditions are the most probable state, with extreme wet and dry conditions being less frequent. The proposed MICET approach offers a robust solution for improving precipitation-based drought projections by addressing complex relationships in MME models. The newly introduced SMI index provides a more reliable representation of drought trends, emphasizing the increasing severity of long-term droughts due to climate change.