BAMPP: A novel Bayesian network enhanced by average marginal posterior probabilities to identify critical ground truth meteorological stations for drought monitoring

Niaz R., Munir S., Raza M. A., TÜR R., Partani S., Danandeh Mehr A.

Physics and Chemistry of the Earth, vol.142, 2026 (SCI-Expanded, Scopus) identifier

  • Publication Type: Article / Article
  • Volume: 142
  • Publication Date: 2026
  • Doi Number: 10.1016/j.pce.2025.104215
  • Journal Name: Physics and Chemistry of the Earth
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Artic & Antarctic Regions, Chimica, Compendex, Geobase, INSPEC
  • Keywords: Ankara, Bayesian networks, Critical station, Marginal posterior probabilities, Meteorological drought
  • Akdeniz University Affiliated: Yes

Abstract

This study introduces a new approach for identifying critical meteorological stations essential for analyzing spatiotemporal dynamics of drought events at regional scale. We propose a Bayesian network enhanced by Average Marginal Posterior Probabilities (AMPP) to evaluate stations based on frequency, severity, and persistence of Standardized Precipitation Index (SPI) at multiple timescales (SPI-3, SPI-6, and SPI-12). The method is demonstrated across the province of Ankara, Türkiye, effectively capturing the probabilistic relationships and interdependencies governing drought propagation among stations. Our analysis revealed distinct spatiotemporal patterns across the region at all time scales. For short-term droughts (SPI-3), critical station identity varied seasonally, indicating localized dynamics; for instance, Esenboga was key station in February, March, July, October and December, while Beypazari was influential in the other months. In contrast, for medium- and long-term droughts (SPI-6, SPI-12), Beypazari was the most critical station across all months, establishing it as the representative station for long-term drought monitoring in the region. This method provides a robust, probabilistic tool for optimizing drought monitoring networks and enhancing regional water resource management.