Houénafa et al. (2026) S-GRHyMoLAP: A stochastic extension of the GRHyMoLAP model with different diffusion mechanisms for probabilistic rainfall-runoff modeling

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-09-18
• Authors: Sianou Ezéckiel Houénafa, Lionel Cédric Gohouede, Romuald Daniel BOY-NGBOGBELE, Melissa Latella, Olatunji Olugoke Johnson, Cenk Sezen
• DOI: 10.1016/j.jhydrol.2026.136449

Research Groups

African Institute for Mathematical Sciences (AIMS), Cape Town, South Africa
Department of Applied Mathematics, Stellenbosch University, Stellenbosch, South Africa
Department of Mathematics, Pan African University Institute for Basic Sciences, Technology and Innovation, Nairobi, Kenya
CMCC Foundation – Euro-Mediterranean Center on Climate Change, Lecce, Italy
Department of Mathematics, University of Manchester, Manchester, United Kingdom
Faculty of Engineering, Ondokuz Mayis University, 55139 Samsun, Turkey

Short Summary

This study introduces a stochastic extension of the GRHyMoLAP model (S-GRHyMoLAP) with different diffusion mechanisms for probabilistic rainfall-runoff modeling. The proposed framework captures how perturbations in hydrological components contribute to flow dynamics and provides more robust uncertainty quantification.

Objective

• Investigate alternative representations of uncertainties in inputs, model structure, parameters, and their potential interactions in watershed hydrology.

Study Configuration

• Spatial Scale: Catchment scale (526 catchments from the CAMELS-FR dataset)
• Temporal Scale: Long-term (evaluation period not specified)

Methodology and Data

• Models used: GRHyMoLAP model, S-GRHyMoLAP stochastic extension
• Data sources: CAMELS-FR dataset, satellite data, observation data

Main Results

• Input-driven diffusion provides the best probabilistic representation among single-variable diffusion formulations.
• Interaction-driven diffusion performs best in 70% of catchments and outperforms heteroscedastic Gaussian residual (HGR) benchmark method for GRHyMoLAP.

Contributions

• The study introduces a new stochastic framework (S-GRHyMoLAP) that captures the effects of perturbations in hydrological components on flow dynamics.
• The results highlight the potential of incorporating interaction-driven diffusion formulations to improve probabilistic streamflow prediction.

Funding

• This research was funded by [projects, programs, and reference codes not specified]

Citation

@article{Houénafa2026SGRHyMoLAP,
  author = {Houénafa, Sianou Ezéckiel and Gohouede, Lionel Cédric and BOY-NGBOGBELE, Romuald Daniel and Latella, Melissa and Johnson, Olatunji Olugoke and Sezen, Cenk},
  title = {S-GRHyMoLAP: A stochastic extension of the GRHyMoLAP model with different diffusion mechanisms for probabilistic rainfall-runoff modeling},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.136449},
  url = {https://doi.org/10.1016/j.jhydrol.2026.136449}
}