Scholarlink

<p>This study investigates the impacts of climate extremes shifting precipitation patterns and rising temperatures on agricultural productivity and rangeland ecosystems in Pakistan. Using data from 23 Coupled Model Intercomparison Project Phase 6 (CMIP6) global climate models (GCMs), the research projects future climate scenarios under two socioeconomic pathways, SSP2-4.5 and SSP5-8.5, from 2015 to 2100. Advanced machine learning (ML) techniques, including Long Short-Term Memory (LSTM) networks, Gradient Boosting, Artificial Neural Networks (ANN), Support Vector Machines (SVM), and Linear Regression, were employed to analyze historical (1980–2014) and future climate trends. The results show that LSTM outperforms other models in predicting temperature and precipitation extremes, achieving higher <i>R</i><sup>2</sup> values and lower prediction errors (MSE/RMSE). Under the SSP5-8.5 scenario, severe vulnerabilities are projected for Pakistan, with intensified heatwaves, erratic rainfall, and prolonged droughts reducing crop yields and rangeland productivity. These climate extremes exacerbate food insecurity and disrupt pastoral livelihoods, making adaptive strategies, such as climate-resilient crops and sustainable water management, critical for mitigating these risks. The study highlights the urgent need for targeted policies, including early warning systems and farmer education, to strengthen Pakistan’s capacity to cope with climate impacts, ensuring food security and ecosystem resilience.</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

Machine learning assessment of CMIP6 projected maximum temperature and precipitation impacts on crop yields and rangeland productivity in Pakistan

  • 期刊论文

  • 发表日期：

    2025年8月7日

查看全文
  • Wilayat Shah,
  • Junfei Chen

摘要

This study investigates the impacts of climate extremes shifting precipitation patterns and rising temperatures on agricultural productivity and rangeland ecosystems in Pakistan. Using data from 23 Coupled Model Intercomparison Project Phase 6 (CMIP6) global climate models (GCMs), the research projects future climate scenarios under two socioeconomic pathways, SSP2-4.5 and SSP5-8.5, from 2015 to 2100. Advanced machine learning (ML) techniques, including Long Short-Term Memory (LSTM) networks, Gradient Boosting, Artificial Neural Networks (ANN), Support Vector Machines (SVM), and Linear Regression, were employed to analyze historical (1980–2014) and future climate trends. The results show that LSTM outperforms other models in predicting temperature and precipitation extremes, achieving higher R2 values and lower prediction errors (MSE/RMSE). Under the SSP5-8.5 scenario, severe vulnerabilities are projected for Pakistan, with intensified heatwaves, erratic rainfall, and prolonged droughts reducing crop yields and rangeland productivity. These climate extremes exacerbate food insecurity and disrupt pastoral livelihoods, making adaptive strategies, such as climate-resilient crops and sustainable water management, critical for mitigating these risks. The study highlights the urgent need for targeted policies, including early warning systems and farmer education, to strengthen Pakistan’s capacity to cope with climate impacts, ensuring food security and ecosystem resilience.