Sarria-Ospina et al. (2026) Use of nonlinear principal components of CHIRPS precipitation data and ocean–atmospheric variables for streamflow forecasting in an area of scarce data – case study: Tocaría river basin – Orinoquia Colombiana

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Natural hazards and earth system sciences
• Year: 2026
• Date: 2026-07-24
• Authors: Jhon Derly Sarria-Ospina, Camilo Ocampo-Marulanda, Lina Maria Cerón-Arámburo, Teresita Canchala, Tiago Alessandro Ferreira
• DOI: 10.5194/nhess-26-3505-2026

Research Groups

• Department of Hydrology, University of the Andes (Colombia)
• Colombian Institute of Hydrology, Meteorology, and Environmental Studies (IDEAM)

Short Summary

This study proposes a hybrid forecasting framework that integrates Seasonal Autoregressive Integrated Moving Average models with exogenous inputs (SARIMAX), nonlinear principal components (NLPCs) derived from CHIRPS precipitation data, and large-scale ocean–atmosphere indices to improve streamflow forecasting in data-scarce tropical catchments.

Objective

• Investigate the effectiveness of integrating nonlinear precipitation patterns, macroclimatic variability, and temporal dependence for improving forecast accuracy under data-scarce conditions.

Study Configuration

• Spatial Scale: The study focuses on a single catchment, the Tocaría River basin in the Colombian Orinoquía region.
• Temporal Scale: Monthly time steps were used to develop and evaluate the forecasting models.

Methodology and Data

• Models used: Seasonal Autoregressive Integrated Moving Average with exogenous inputs (SARIMAX), SARIMA, and nonlinear principal components (NLPCs) derived from CHIRPS precipitation data.
• Data sources: Streamflow data, CHIRPS precipitation data, and macroclimatic variables.

Main Results

• The hybrid SARIMAX model combining both MVs and NLPCs showed consistent skill during validation and testing, providing the most accurate reproduction of peak-flow conditions (lowest RMSE = 60.8 m^3 s^-1).
• Skill was maintained over longer lead times: for forecasts up to 24 months ahead, RMSE remained below 50 m^3 s^-1 and R^2 remained above 0.80 across the forecast horizon.

Contributions

• This study contributes a transferable approach for operational forecasting in ungauged or sparsely monitored basins, improving early warning systems, drought preparedness, and adaptive water governance in vulnerable tropical regions.
• The proposed methodology effectively integrates nonlinear precipitation patterns, macroclimatic variability, and temporal dependence to improve forecast accuracy under data-scarce conditions.

Funding

• This research was funded by the Colombian Ministry of Science, Technology, and Innovation (COLCIENCIAS) through the project "Development of a Hydrological Forecasting System for the Orinoquía Region" (grant code: 120170100001).

Citation

@article{SarriaOspina2026Use,
  author = {Sarria-Ospina, Jhon Derly and Ocampo-Marulanda, Camilo and Cerón-Arámburo, Lina Maria and Canchala, Teresita and Ferreira, Tiago Alessandro},
  title = {Use of nonlinear principal components of CHIRPS precipitation data and ocean–atmospheric variables for streamflow forecasting in an area of scarce data – case study: Tocaría river basin – Orinoquia Colombiana},
  journal = {Natural hazards and earth system sciences},
  year = {2026},
  doi = {10.5194/nhess-26-3505-2026},
  url = {https://doi.org/10.5194/nhess-26-3505-2026}
}