Scholarlink

<p>Drought is a complex natural hazard; therefore, its management poses one of the biggest challenges in the twenty-first century. Multi-Model ensemble of Global Climate Models (GCMs) has important role in future drought assessment. This study proposes a novel framework for predicting twenty-first century drought using multiple simulated data of precipitation of GCMs – Mutual Information-Based Regional Integrated Standardized Multicriteria Decision for Drought Assessment (MIRISMDA). The framework of MIRSMDA consists of four phases: The first phase employs Multi-Criteria Group Decision Making Analysis based on Normalized Joint Mutual Information Maximization (NJMIM) scores to select a subset from a pool of multiple GCMs. The second phase aggregates the selected GCMs. The third phase uses an optimum MME. The fourth phase derive a new drought index – Mutual Information-Based Regional Integrated Standardized Multicriteria Decision for Drought Assessment (MIRISMDA). Five multi-model ensemble methods are evaluated using the Kling-Gupta Efficiency (KGE) metric, with the best ensemble weights subsequently applied to aggregate future scenarios of the selected GCMs. For the choice of appropriate probability model, BICs minimum value reveals the superiority of KCGMD. Additionally, the Bias-Corrected Eigenvector Forecast Ensemble (BCEVFE) are found to be better in capturing observed precipitation trend. For predicting future drought classes under MIRISMDA index, we used Steady States Probability of Markov Chain. To apply the proposed framework and index, we used historical precipitation records spanning from 1950 to 2014, gathered from 28 meteorological stations across the Punjab province of Pakistan. For future projections, we incorporated simulation data from 22 Global Climate Models (GCMs) included in the Coupled Model Intercomparison Project Phase 6 (CMIP6), covering both past and anticipated climate conditions. The evaluation of future drought patterns, based on steady-state probabilities derived from the proposed index, indicates an increasing likelihood of drought events in the study area. This tendency is more evident under high-emission scenarios and at extended time scales. These findings highlight the significant influence climate change may have on future drought occurrences in the region.</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 statistical framework for future drought assessment – mutual information-based regional integrated standardized multicriteria decision for drought assessment

  • 期刊论文

  • 发表日期：

    2025年5月16日

查看全文
  • Muhammad Shakeel,
  • Zulfiqar Ali

摘要

Drought is a complex natural hazard; therefore, its management poses one of the biggest challenges in the twenty-first century. Multi-Model ensemble of Global Climate Models (GCMs) has important role in future drought assessment. This study proposes a novel framework for predicting twenty-first century drought using multiple simulated data of precipitation of GCMs – Mutual Information-Based Regional Integrated Standardized Multicriteria Decision for Drought Assessment (MIRISMDA). The framework of MIRSMDA consists of four phases: The first phase employs Multi-Criteria Group Decision Making Analysis based on Normalized Joint Mutual Information Maximization (NJMIM) scores to select a subset from a pool of multiple GCMs. The second phase aggregates the selected GCMs. The third phase uses an optimum MME. The fourth phase derive a new drought index – Mutual Information-Based Regional Integrated Standardized Multicriteria Decision for Drought Assessment (MIRISMDA). Five multi-model ensemble methods are evaluated using the Kling-Gupta Efficiency (KGE) metric, with the best ensemble weights subsequently applied to aggregate future scenarios of the selected GCMs. For the choice of appropriate probability model, BICs minimum value reveals the superiority of KCGMD. Additionally, the Bias-Corrected Eigenvector Forecast Ensemble (BCEVFE) are found to be better in capturing observed precipitation trend. For predicting future drought classes under MIRISMDA index, we used Steady States Probability of Markov Chain. To apply the proposed framework and index, we used historical precipitation records spanning from 1950 to 2014, gathered from 28 meteorological stations across the Punjab province of Pakistan. For future projections, we incorporated simulation data from 22 Global Climate Models (GCMs) included in the Coupled Model Intercomparison Project Phase 6 (CMIP6), covering both past and anticipated climate conditions. The evaluation of future drought patterns, based on steady-state probabilities derived from the proposed index, indicates an increasing likelihood of drought events in the study area. This tendency is more evident under high-emission scenarios and at extended time scales. These findings highlight the significant influence climate change may have on future drought occurrences in the region.