Scholarlink

<p>Accurate projection and characterization of future droughts are critical for developing sustainable policies. However, the complex structure of rainfall patterns often compromises computational accuracy in modeling future drought characteristics. This study proposes a novel framework for analyzing future drought characterization using 22 global climate model (GCM) simulations from the Coupled Model Intercomparison Project Phase 6. The framework introduces two key components: a weighted ensemble procedure, termed the global climate model fusion ensemble (GCMF-Ensemble), and a new drought index, the Multimodal Simulated Standardized Drought Index (MSSDI). The GCMF-Ensemble methodology integrates two weighting schemes—Record Score Weights and Disparity in Uniqueness Assessment—to enhance ensemble performance. For MSSDI, aggregated monthly precipitation data are standardized using mixture probability models. The proposed framework is applied to 94 grid points across Pakistan, utilizing historical data from 1950 to 2014. To evaluate the efficacy of the GCMF-Ensemble, its error metrics are compared with three existing techniques: equal weighted average ensemble (EWAE), Bayesian model averaging ensemble (BMAE), and mutual information-based ensemble (MIBE). The GCMF-Ensemble achieves superior performance, with minimum RMSE and MAE values of (7.93, 5.73), outperforming EWAE (8.54, 6.41), BMAE (9.83, 7.15), and MIBE (8.93, 6.80). For future drought characterization under MSSDI, three shared socioeconomic pathways (SSPs)—SSP1-2.6, SSP2-4.5, and SSP5-8.5—are analyzed for the period 2015–2100. Steady-state probabilities are computed to assess long-term drought behavior across multiple timescales under these scenarios. The results demonstrate that the GCMF-Ensemble methodology is both effective and versatile, significantly improving the performance of multi-model ensembles for drought characterization.</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

Enhancing drought projections in Pakistan using a weighted ensemble of precipitation and drought indices

  • 期刊论文

  • 发表日期：

    2025年7月1日

查看全文
  • Hussnain Abbas,
  • Zulfiqar Ali

摘要

Accurate projection and characterization of future droughts are critical for developing sustainable policies. However, the complex structure of rainfall patterns often compromises computational accuracy in modeling future drought characteristics. This study proposes a novel framework for analyzing future drought characterization using 22 global climate model (GCM) simulations from the Coupled Model Intercomparison Project Phase 6. The framework introduces two key components: a weighted ensemble procedure, termed the global climate model fusion ensemble (GCMF-Ensemble), and a new drought index, the Multimodal Simulated Standardized Drought Index (MSSDI). The GCMF-Ensemble methodology integrates two weighting schemes—Record Score Weights and Disparity in Uniqueness Assessment—to enhance ensemble performance. For MSSDI, aggregated monthly precipitation data are standardized using mixture probability models. The proposed framework is applied to 94 grid points across Pakistan, utilizing historical data from 1950 to 2014. To evaluate the efficacy of the GCMF-Ensemble, its error metrics are compared with three existing techniques: equal weighted average ensemble (EWAE), Bayesian model averaging ensemble (BMAE), and mutual information-based ensemble (MIBE). The GCMF-Ensemble achieves superior performance, with minimum RMSE and MAE values of (7.93, 5.73), outperforming EWAE (8.54, 6.41), BMAE (9.83, 7.15), and MIBE (8.93, 6.80). For future drought characterization under MSSDI, three shared socioeconomic pathways (SSPs)—SSP1-2.6, SSP2-4.5, and SSP5-8.5—are analyzed for the period 2015–2100. Steady-state probabilities are computed to assess long-term drought behavior across multiple timescales under these scenarios. The results demonstrate that the GCMF-Ensemble methodology is both effective and versatile, significantly improving the performance of multi-model ensembles for drought characterization.