Scholarlink

<p>Drought, an extreme climate event with far-reaching impacts, stands as a difficult challenge in our contemporary world. Hence, the objective of this research is to enhance drought assessment accuracy while reducing ensemble uncertainty and to characterize the future trend under different scenarios. Therefore, the novelty of this study is to propose a novel framework for drought assessment. Unlike conventional equal-weight ensemble approaches, this study employs a Stepwise Regression (SR) methodology to choose appropriate models for each location and introduces a “Deviation Based Scheme (DBS)” weighting scheme. In addition, by using ensemble data, we also introduce an innovative standardized index known as the Spatial-wise Feature Selected Multi-model Ensemble Standardized Drought Index (SFSMSDI). The projected data is standardized by the K-Components Gaussian Mixture Model (K-CGMM). Comparative evaluation demonstrates that DBS significantly improves simulation accuracy relative to the Simple Model Average (SMA), as evidenced by reduced Sum of Squared Differences (SSD) and Root Mean Square Error (RMSE) values. Moreover, analysis of Steady State Probabilities (SSPs) under SSP1-2.6 scenarios indicates slight increases in drought persistence at longer time scales. Whereas, under SSP2-4.5, SSPs suggest that transitional emission pathways may amplify drought persistence without extreme intensification. On the other hand, SSPs under the SSP5-8.5 scenario indicate significant moisture deficits over the Tibetan Plateau. This research serves as an urgent call to action for policymakers, highlighting the imperative need for novel mitigation policies to safeguard the sustainability of the Tibetan Plateau’s environment.</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

A New Stepwise Regression Based Multi-Model Ensemble for Accurate Assessment of Future Drought

  • 期刊论文

  • 发表日期：

    2026年3月28日

查看全文
  • Abdul Baseer,
  • Mahrukh Yousaf,
  • Zulfiqar Ali,
  • Zongxue Xu

摘要

Drought, an extreme climate event with far-reaching impacts, stands as a difficult challenge in our contemporary world. Hence, the objective of this research is to enhance drought assessment accuracy while reducing ensemble uncertainty and to characterize the future trend under different scenarios. Therefore, the novelty of this study is to propose a novel framework for drought assessment. Unlike conventional equal-weight ensemble approaches, this study employs a Stepwise Regression (SR) methodology to choose appropriate models for each location and introduces a “Deviation Based Scheme (DBS)” weighting scheme. In addition, by using ensemble data, we also introduce an innovative standardized index known as the Spatial-wise Feature Selected Multi-model Ensemble Standardized Drought Index (SFSMSDI). The projected data is standardized by the K-Components Gaussian Mixture Model (K-CGMM). Comparative evaluation demonstrates that DBS significantly improves simulation accuracy relative to the Simple Model Average (SMA), as evidenced by reduced Sum of Squared Differences (SSD) and Root Mean Square Error (RMSE) values. Moreover, analysis of Steady State Probabilities (SSPs) under SSP1-2.6 scenarios indicates slight increases in drought persistence at longer time scales. Whereas, under SSP2-4.5, SSPs suggest that transitional emission pathways may amplify drought persistence without extreme intensification. On the other hand, SSPs under the SSP5-8.5 scenario indicate significant moisture deficits over the Tibetan Plateau. This research serves as an urgent call to action for policymakers, highlighting the imperative need for novel mitigation policies to safeguard the sustainability of the Tibetan Plateau’s environment.