Ullah et al. (2026) Regional Temperature Trends and Future Warming Risks in Pakistan Using Bias‐Corrected CMIP6 Ensembles and Machine Learning Techniques

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: International Journal of Climatology
• Year: 2026
• Date: 2026-08-02
• Authors: Hamd Ullah, Firdos Khan, Majid Khan, Muhammad Abbas
• DOI: 10.1002/joc.70497

Research Groups

Not specified

Short Summary

This study enhances regional temperature projections for Pakistan by integrating CMIP6 models with Fuzzy C-Means clustering, statistical bias correction, and machine learning to provide localized climate insights.

Objective

• To improve the accuracy of regional temperature projections in Pakistan to support climate adaptation, water management, and disaster response.

Study Configuration

• Spatial Scale: National (Pakistan), subdivided into six climatically similar regions using Fuzzy C-Means clustering.
• Temporal Scale: Future projections divided into three periods: 2015–2044, 2045–2074, and 2075–2100.

Methodology and Data

• Models used: CMIP6 (Coupled Model Intercomparison Project Phase 6), Fuzzy C-Means clustering, Sen's slope estimator, and Random Forest regression (RFR).
• Data sources: CMIP6 model outputs and observed temperature data for bias correction.

Main Results

• Trend Analysis: Sen's slope identified steady warming trends, while Random Forest regression more effectively captured nonlinear temperature changes.
• Regional Impact: Significant warming is projected for central and southern Pakistan, with the most pronounced increases occurring in lowland and southern areas.
• Scenario Comparison: The SSP5-8.5 scenario predicts accelerated warming threatening public health and food security, whereas the SSP2-4.5 scenario suggests partial stabilization toward the end of the century.
• Error Reduction: Statistical bias correction successfully reduced systematic deviations between modeled and observed data.

Contributions

• Fills a critical gap in localized climate research for Pakistan by providing a high-resolution regional framework that integrates machine learning and clustering to improve the precision of future temperature projections.

Funding

Not specified

Citation

@article{Ullah2026Regional,
  author = {Ullah, Hamd and Khan, Firdos and Khan, Majid and Abbas, Muhammad},
  title = {Regional Temperature Trends and Future Warming Risks in Pakistan Using Bias‐Corrected <scp>CMIP6</scp> Ensembles and Machine Learning Techniques},
  journal = {International Journal of Climatology},
  year = {2026},
  doi = {10.1002/joc.70497},
  url = {https://doi.org/10.1002/joc.70497}
}