Yarahmadi et al. (2026) Machine learning-enhanced game theory composite drought index for reliable drought monitoring and forecasting in cold Mediterranean mountain climates

Identification

• Journal: The Science of The Total Environment
• Year: 2026
• Date: 2026-09-18
• Authors: Mozhgan Yarahmadi, Ali Haghizadeh, Leila Ghasemi
• DOI: 10.1016/j.scitotenv.2026.182290

Research Groups

Department of Watershed Management, Faculty of Natural Resources, University of Tehran, Iran
Department of Watershed Management, Faculty of Natural Resources, Lorestan University, Iran

Short Summary

This study developed a game theory-based composite drought index (GTDI) for reliable drought monitoring and forecasting in cold Mediterranean mountain climates. The GTDI was found to be strongly correlated with the Standardized Precipitation-Evapotranspiration Index (SPEI), indicating meteorological drivers as the main influence.

Objective

• Investigate the development of a spatially variable weighted drought index using game theory for reliable drought monitoring and forecasting in cold Mediterranean mountain climates.

Study Configuration

• Spatial Scale: Lorestan Province, Iran
• Temporal Scale: 1969–2020 (51 years)

Methodology and Data

• Models used: Game Theory, Random Forest
• Data sources: ERA5-Land reanalysis data for precipitation, temperature, and soil moisture

Main Results

• The GTDI was found to be strongly correlated with SPEI (R = 0.97, p < 0.001), indicating meteorological drivers as the main influence.
• Four seasonal drought patterns were identified: severe summer drought (32.3%), moderate winter drought (32.2%), mild autumn drought (18.0%), and spring non-drought conditions (17.2%).
• A Random Forest forecast for 2021–2025 gave high accuracy (R2 = 0.9114, RMSE = 0.2876, MAE = 0.2008).

Contributions

• The study provides a reliable composite drought index for monitoring and forecasting droughts in cold Mediterranean mountain climates.
• The GTDI was found to be useful for identifying seasonal drought patterns and predicting future drought conditions.

Funding

• This research was funded by the University of Tehran (Grant number: [not specified])
• This research was also supported by the Lorestan University Research Fund (Grant number: [not specified])

Citation

@article{Yarahmadi2026Machine,
  author = {Yarahmadi, Mozhgan and Haghizadeh, Ali and Ghasemi, Leila},
  title = {Machine learning-enhanced game theory composite drought index for reliable drought monitoring and forecasting in cold Mediterranean mountain climates},
  journal = {The Science of The Total Environment},
  year = {2026},
  doi = {10.1016/j.scitotenv.2026.182290},
  url = {https://doi.org/10.1016/j.scitotenv.2026.182290}
}