Tepetidis et al. (2026) From deterministic to stochastic: A hybrid framework coupling machine learning with Bluecat for probabilistic satellite-based precipitation estimation over Greece

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2026
• Date: 2026-09-28
• Authors: Nikos Tepetidis, Theano Iliopoulou, Panayiotis Dimitriadis, Ioannis Benekos, Alberto Montanari, Demetris Koutsoyiannis
• DOI: 10.1016/j.ejrh.2026.103904

Research Groups

• Department of Water Resources and Environmental Engineering, School of Civil Engineering, National Technical University of Athens, Heroon Polytechneiou 9, Zographou, 15780, Greece
• Laboratory of Risk Management and Resilience, Hellenic Institute of Transport, Centre of Research and Technology Hellas, Ethnarchou Makariou 34, Ilioupoli, 16341, Greece
• Department of Civil, Chemical, Environmental and Materials Engineering, University of Bologna, Via del Risorgimento 2, Bologna, 40136, Italy
• Department of Geography, School of Environment, Geography and Applied Economics, Harokopio University of Athens, El. Venizelou 70, Athens, 17676, Greece

Short Summary

This study presents a hybrid machine learning-stochastic framework for post-processing satellite-derived precipitation using ground-based observations in Greece. The proposed framework improves the accuracy of satellite precipitation estimates and provides reliable uncertainty quantification.

Objective

• Develop a hybrid machine learning-stochastic framework for merging satellite and ground-based precipitation data.
• Improve the accuracy of satellite precipitation estimates over Greece.

Study Configuration

• Spatial Scale: High-resolution (0.1° × 0.1°) spatial coverage across Greece.
• Temporal Scale: Daily temporal resolution, with a dataset spanning from June 2000 to September 2021.

Methodology and Data

• Models used: Linear Regression (LR), Random Forest (RF), eXtreme Gradient Boosting (XGBoost), Categorical Boosting (CatBoost).
• Data sources: IMERG precipitation data, gauge observations from various stations across Greece.

Main Results

• The proposed framework improves the accuracy of satellite precipitation estimates over Greece.
• The estimated output distribution achieves Nash-Sutcliffe efficiency (NSE) and Kling-Gupta efficiency (KGE) values of 0.99 and 0.96, respectively.
• The framework provides reliable uncertainty quantification.

Contributions

• This study presents a new methodological framework for merging satellite and ground-based precipitation data in Greece.
• The proposed framework improves the accuracy of satellite precipitation estimates over Greece.
• The framework provides reliable uncertainty quantification.

Funding

• This research was funded by [project name], [reference code].

Citation

@article{Tepetidis2026From,
  author = {Tepetidis, Nikos and Iliopoulou, Theano and Dimitriadis, Panayiotis and Benekos, Ioannis and Montanari, Alberto and Koutsoyiannis, Demetris},
  title = {From deterministic to stochastic: A hybrid framework coupling machine learning with Bluecat for probabilistic satellite-based precipitation estimation over Greece},
  journal = {Journal of Hydrology Regional Studies},
  year = {2026},
  doi = {10.1016/j.ejrh.2026.103904},
  url = {https://doi.org/10.1016/j.ejrh.2026.103904}
}