Maliki et al. (2026) Employing artificial intelligence to predict δ¹⁸O and δ²H isotope ratios in precipitation in Iraq under changing climate patterns

Identification

• Journal: Scientific Reports
• Year: 2026
• Date: 2026-01-08
• Authors: Ali Al Maliki, Ali Al-Naji, Ahmed Kadhim Al Lami, Haitham Abdulmohsin Afan, Maryam Bayatvarkeshi, Nadhir Al-Ansari
• DOI: 10.1038/s41598-026-35047-x

Research Groups

• Scientific Research Commission, Baghdad, Iraq
• College of Engineering, Al-Ayen University, Thi-Qar, Iraq
• Electrical Engineering Technical College, Middle Technical University, Baghdad, Iraq
• Department of Physics, College of Science, Al-Nahrain University, Jadriya, Baghdad, Iraq
• Upper Euphrates Center for Sustainable Development Research, University of Anbar, Ramadi, Iraq
• Department of Geography and Environmental Management, University of Waterloo, ON, Canada
• Civil, Environmental and Natural Resources Engineering, Lulea University of Technology, Lulea, Sweden

Short Summary

This study developed a mathematical predictive model for δ¹⁸O and δ²H isotope ratios in precipitation in Iraq using various machine learning techniques, demonstrating that the Random Forest model achieved superior predictive performance with a calibration coefficient (R²) of 0.8983.

Objective

• To develop a mathematical predictive model for rainfall isotopic values (δ¹⁸O and δ²H) using artificial intelligence techniques in Iraq.
• To assess the impact of ambient temperature, elevation, and relative humidity on the measured values of stable isotopes of hydrogen and oxygen (δ²H and δ¹⁸O).

Study Configuration

• Spatial Scale: Iraq, covering 34 meteorological stations distributed across diverse topographic regions.
• Temporal Scale: A 14-year period (2010–2024), with rain samples collected during the rainy season (November to April).

Methodology and Data

• Models used: Support Vector Machine (SVM), Gradient Boosting Regressor (GBR), Artificial Neural Network (ANN), CatBoost, XGBoost, and Random Forest (RF).
• Data sources:
  • Stable isotope data (δ¹⁸O and δ²H) in precipitation: Collected from 34 meteorological stations across Iraq, with an isotopic database obtained from the Water Isotope System for Data Analysis, Visualization, and Electronic Retrieval (WISER) of the International Atomic Energy Agency (IAEA).
  • Meteorological parameters: Precipitation amount, air temperature, relative humidity, and calculated station elevation.
  • Dataset size: 279 original samples, augmented to 27,600 samples for training and testing.

Main Results

• The Random Forest (RF) model achieved the highest predictive accuracy, with an R² value of 0.8983 in the testing set, explaining approximately 90% of the variability in isotopic composition.
• The RF model also recorded the lowest mean absolute error (MAE) of 1.39 per mil and the lowest root mean square error (RMSE) of 3.60 per mil.
• XGBoost and CatBoost also performed well but were outperformed by RF, while SVM exhibited the weakest predictive capability (R² = 0.18, MAE = 6.54 per mil).
• Feature importance analysis by the RF model identified rain amount and air temperature as the most important features for predicting isotopes, followed by relative humidity and elevation.
• Data augmentation, which expanded the dataset and introduced slight random changes to input features, resulted in strong predictive capabilities on the test data (R² = 0.92 for δ¹⁸O and R² = 0.93 for δ²H).

Contributions

• Developed a reliable and robust artificial intelligence model (Random Forest) for reconstructing rainfall isotope signatures (δ¹⁸O and δ²H) from routine meteorological data in arid and semi-arid regions.
• Demonstrated the effectiveness of integrating machine learning, particularly the Random Forest approach, in enhancing the modeling of isotopic signature predictions in environmental studies.
• Provided a powerful tool for reconstructing historical isotopic datasets, supporting climate variability assessment and sustainable water resources management, especially in data-scarce regions.
• Validated the methodology in Iraq, a region characterized by diverse and extreme climatic variability, addressing a gap in the literature regarding the reliability of predictive techniques under such conditions.

Funding

This paper did not receive any funding.

Citation

@article{Maliki2026Employing,
  author = {Maliki, Ali Al and Al-Naji, Ali and Lami, Ahmed Kadhim Al and Afan, Haitham Abdulmohsin and Bayatvarkeshi, Maryam and Al-Ansari, Nadhir},
  title = {Employing artificial intelligence to predict δ¹⁸O and δ²H isotope ratios in precipitation in Iraq under changing climate patterns},
  journal = {Scientific Reports},
  year = {2026},
  doi = {10.1038/s41598-026-35047-x},
  url = {https://doi.org/10.1038/s41598-026-35047-x}
}