Scholarlink

<p>Climate change has intensified drought frequency and severity worldwide. As a slow-onset natural hazard with wide-ranging impacts, drought monitoring remains challenging due to the uneven distribution of meteorological stations and strong climatic variability. This study proposes a Regional Standardized Precipitation Drought Index (RSPDI) for regional drought assessment by integrating precipitation dynamics from multiple stations using Bayesian Network (BN) theory. Representative stations were identified through marginal posterior probabilities, and the combined precipitation series was standardized using a K-Component Gaussian Mixture Model (K-CGMM) to provide a flexible alternative to traditional univariate probability models. The BN model was simulated with 200,000 MCMC iterations to ensure precision in probabilistic station selection. Application to two regions in Pakistan showed strong agreement between RSPDI and SPI, with correlations ranging from 0.319 to 0.756 (1-month) and 0.181–0.824 (3-month) time scales. The Astor station exhibited the highest Average Joint Dependency Probability (0.622), confirming the regional representativeness of the proposed method. The RSPDI demonstrates enhanced spatial coherence and robust probabilistic consistency, providing a statistically grounded and transferable framework for regional drought monitoring and early-warning applications under uneven station networks.</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

Regional drought assessment using multi-site probabilistically integrated precipitation by Bayesian network

  • 期刊论文

  • 发表日期：

    2026年2月25日

查看全文
  • Farman Ali,
  • Jing-Cheng Han,
  • Dong Su,
  • Zulfiqar Ali,
  • Yang Zhou,
  • Yuefei Huang,
  • Muhammad Ahmad,
  • Shafeeq Ur Rahman

摘要

Climate change has intensified drought frequency and severity worldwide. As a slow-onset natural hazard with wide-ranging impacts, drought monitoring remains challenging due to the uneven distribution of meteorological stations and strong climatic variability. This study proposes a Regional Standardized Precipitation Drought Index (RSPDI) for regional drought assessment by integrating precipitation dynamics from multiple stations using Bayesian Network (BN) theory. Representative stations were identified through marginal posterior probabilities, and the combined precipitation series was standardized using a K-Component Gaussian Mixture Model (K-CGMM) to provide a flexible alternative to traditional univariate probability models. The BN model was simulated with 200,000 MCMC iterations to ensure precision in probabilistic station selection. Application to two regions in Pakistan showed strong agreement between RSPDI and SPI, with correlations ranging from 0.319 to 0.756 (1-month) and 0.181–0.824 (3-month) time scales. The Astor station exhibited the highest Average Joint Dependency Probability (0.622), confirming the regional representativeness of the proposed method. The RSPDI demonstrates enhanced spatial coherence and robust probabilistic consistency, providing a statistically grounded and transferable framework for regional drought monitoring and early-warning applications under uneven station networks.