Scholarlink

<p>Climate change has intensified global meteorological hazardous events, and drought is the most serious. This study provides a comprehensive spatio-temporal analysis of drought dynamics in Pakistan during 2000–2023. The study is innovative in developing a Composite Drought Index (CDI) by integrating multiple remote sensing-based drought indices such as the Vegetation Condition Index, Temperature Condition Index, Vegetation Health Index, Standardized Precipitation Index, and Palmer Drought Severity Index. The satellite imageries data were acquired from the Google Earth engine clouding computing platform, while crop yields data was obtained from Agriculture Marketing Information Service, Pakistan. Unlike previous studies that rely on single indicators, this study employs a unique two-tiered analytical framework. First, a CDI was developed using Principal Component Analysis to map drought-prone regions. Second, the Kernel Regularized Least Squares (KRLS) machine learning approach was applied to quantify the impact of CDI on major crop yields. The results showed a prominent increase in severe drought episodes in Balochistan, Punjab, and Sindh. Also, the results reveal that extreme drought was notably high across the Cholistan and Thal deserts, Balochistan, and Tharparkar regions. The findings of KRLS demonstrate that increased drought severity captured by the CDI, significantly reduces agricultural productivity across all models. Specifically, for every unit increase in drought severity, yields decline by approximately 23.76 units for sugarcane, 14.78 units for maize, 5.09 units for rice, and 4.26 units for wheat. This research is of great significance for government and policymakers in formulating mitigation strategies to prevent green land conversion and achieve sustainable development goals.</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

Spatio-temporal analysis of climate-driven drought dynamics in Pakistan using geo-spatial and machine learning approaches

  • 期刊论文

  • 发表日期：

    2026年3月5日

查看全文
  • Riaz Ul Hissan,
  • Nusrat Parveen,
  • Bilal Hussain,
  • Shafaqat Ali,
  • Abdul Ghafoor,
  • Magdalena Radulescu

摘要

Climate change has intensified global meteorological hazardous events, and drought is the most serious. This study provides a comprehensive spatio-temporal analysis of drought dynamics in Pakistan during 2000–2023. The study is innovative in developing a Composite Drought Index (CDI) by integrating multiple remote sensing-based drought indices such as the Vegetation Condition Index, Temperature Condition Index, Vegetation Health Index, Standardized Precipitation Index, and Palmer Drought Severity Index. The satellite imageries data were acquired from the Google Earth engine clouding computing platform, while crop yields data was obtained from Agriculture Marketing Information Service, Pakistan. Unlike previous studies that rely on single indicators, this study employs a unique two-tiered analytical framework. First, a CDI was developed using Principal Component Analysis to map drought-prone regions. Second, the Kernel Regularized Least Squares (KRLS) machine learning approach was applied to quantify the impact of CDI on major crop yields. The results showed a prominent increase in severe drought episodes in Balochistan, Punjab, and Sindh. Also, the results reveal that extreme drought was notably high across the Cholistan and Thal deserts, Balochistan, and Tharparkar regions. The findings of KRLS demonstrate that increased drought severity captured by the CDI, significantly reduces agricultural productivity across all models. Specifically, for every unit increase in drought severity, yields decline by approximately 23.76 units for sugarcane, 14.78 units for maize, 5.09 units for rice, and 4.26 units for wheat. This research is of great significance for government and policymakers in formulating mitigation strategies to prevent green land conversion and achieve sustainable development goals.