Open Hydroclimatic Data and Iterative Machine Learning for River Discharge Forecasting in Babahoyo, Ecuador: Benchmarking, Uncertainty, and Operational Limits Without In Situ Validation
This study evaluates the potential and limitations of open/model-derived hydroclimatic data for multi-horizon river discharge forecasting in Babahoyo, Ecuador. A quantitative, applied, retrospective longitudinal design used daily data for 2017–2026. The discharge target is the GloFAS v4 seamless pro...
Guardado en:
| Autores principales: | , , , , , , , , |
|---|---|
| Formato: | Artículo |
| Lenguaje: | inglés |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | Enlace permanente (DOI o Handle) |
| Etiquetas: |
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1881700517490458624 |
|---|---|
| author | Rolando Licapa Redolfo Alexander Haro-Sarango Persi Vera Zelada Denis Javier Aranguri Cayetano Roxana Mabel Sempértegui Rafael Olegario Cabrera-Cabrera Edwar Cieza-Sánchez Martha G Huaman Tanta Diana Carolina Castillo-Martínez |
| author_browse | Alexander Haro-Sarango Denis Javier Aranguri Cayetano Diana Carolina Castillo-Martínez Edwar Cieza-Sánchez Martha G Huaman Tanta Olegario Cabrera-Cabrera Persi Vera Zelada Rolando Licapa Redolfo Roxana Mabel Sempértegui Rafael |
| author_facet | Rolando Licapa Redolfo Alexander Haro-Sarango Persi Vera Zelada Denis Javier Aranguri Cayetano Roxana Mabel Sempértegui Rafael Olegario Cabrera-Cabrera Edwar Cieza-Sánchez Martha G Huaman Tanta Diana Carolina Castillo-Martínez |
| author_sort | Rolando Licapa Redolfo |
| collection | Producción científica |
| description | This study evaluates the potential and limitations of open/model-derived hydroclimatic data for multi-horizon river discharge forecasting in Babahoyo, Ecuador. A quantitative, applied, retrospective longitudinal design used daily data for 2017–2026. The discharge target is the GloFAS v4 seamless product—reanalysis until July 2022, archived operational forecast thereafter; meteorological predictors are ERA5/ERA5-Land/IFS reanalysis, not forecasts. The framework combined leakage-aware feature engineering, temporal validation, rolling-origin backtesting, naïve baselines, machine-learning regression, conformal prediction intervals, and high-flow classification. Performance was strongest at one day, where models reproduced the signal closely (R2 = 0.909), although persistence remained highly competitive. Skill deteriorated at t + 7 and t + 14, where peak timing and magnitude became unreliable. Interval coverage was near-nominal at t + 1 but unreliable at longer horizons. The high-flow classifier identified most q90 cases, yet moderate precision and the absence of gauge validation prevent operational warning claims. Because the target is a 5 km grid simulation of a channel 100–150 m wide, the metrics quantify agreement with GloFAS, not with the physical river, and are reported to three significant digits. Overall, the study is a conservative benchmark for open hydroclimatic data in data-limited tropical floodplains: useful for exploratory monitoring and uncertainty diagnosis, but not a substitute for local hydrometric validation. |
| format | Article |
| id | openalex:W7204073752 |
| institution | UPeU |
| language | English |
| publishDate | 2026 |
| publishDateRange | 2026 |
| publishDateSort | 2026 |
| record_format | dspace |
| spelling | openalex:W7204073752Open Hydroclimatic Data and Iterative Machine Learning for River Discharge Forecasting in Babahoyo, Ecuador: Benchmarking, Uncertainty, and Operational Limits Without In Situ ValidationRolando Licapa RedolfoAlexander Haro-SarangoPersi Vera ZeladaDenis Javier Aranguri CayetanoRoxana Mabel Sempértegui RafaelOlegario Cabrera-CabreraEdwar Cieza-SánchezMartha G Huaman TantaDiana Carolina Castillo-Martínezriver discharge predictionGloFASERA5 reanalysisforecast uncertaintyfeature leakagetropical floodplainBabahoyodata-limited regionsHydrological Forecasting Using AIHydrology and Watershed Management StudiesEnvironmental Monitoring and Data ManagementThis study evaluates the potential and limitations of open/model-derived hydroclimatic data for multi-horizon river discharge forecasting in Babahoyo, Ecuador. A quantitative, applied, retrospective longitudinal design used daily data for 2017–2026. The discharge target is the GloFAS v4 seamless product—reanalysis until July 2022, archived operational forecast thereafter; meteorological predictors are ERA5/ERA5-Land/IFS reanalysis, not forecasts. The framework combined leakage-aware feature engineering, temporal validation, rolling-origin backtesting, naïve baselines, machine-learning regression, conformal prediction intervals, and high-flow classification. Performance was strongest at one day, where models reproduced the signal closely (R2 = 0.909), although persistence remained highly competitive. Skill deteriorated at t + 7 and t + 14, where peak timing and magnitude became unreliable. Interval coverage was near-nominal at t + 1 but unreliable at longer horizons. The high-flow classifier identified most q90 cases, yet moderate precision and the absence of gauge validation prevent operational warning claims. Because the target is a 5 km grid simulation of a channel 100–150 m wide, the metrics quantify agreement with GloFAS, not with the physical river, and are reported to three significant digits. Overall, the study is a conservative benchmark for open hydroclimatic data in data-limited tropical floodplains: useful for exploratory monitoring and uncertainty diagnosis, but not a substitute for local hydrometric validation.Articlehttps://doi.org/10.3390/atmos17080808W7204073752enAtmosphere2073-44332026-08-21 |
| spellingShingle | river discharge prediction GloFAS ERA5 reanalysis forecast uncertainty feature leakage tropical floodplain Babahoyo data-limited regions Hydrological Forecasting Using AI Hydrology and Watershed Management Studies Environmental Monitoring and Data Management Rolando Licapa Redolfo Alexander Haro-Sarango Persi Vera Zelada Denis Javier Aranguri Cayetano Roxana Mabel Sempértegui Rafael Olegario Cabrera-Cabrera Edwar Cieza-Sánchez Martha G Huaman Tanta Diana Carolina Castillo-Martínez Open Hydroclimatic Data and Iterative Machine Learning for River Discharge Forecasting in Babahoyo, Ecuador: Benchmarking, Uncertainty, and Operational Limits Without In Situ Validation |
| title | Open Hydroclimatic Data and Iterative Machine Learning for River Discharge Forecasting in Babahoyo, Ecuador: Benchmarking, Uncertainty, and Operational Limits Without In Situ Validation |
| title_full | Open Hydroclimatic Data and Iterative Machine Learning for River Discharge Forecasting in Babahoyo, Ecuador: Benchmarking, Uncertainty, and Operational Limits Without In Situ Validation |
| title_fullStr | Open Hydroclimatic Data and Iterative Machine Learning for River Discharge Forecasting in Babahoyo, Ecuador: Benchmarking, Uncertainty, and Operational Limits Without In Situ Validation |
| title_full_unstemmed | Open Hydroclimatic Data and Iterative Machine Learning for River Discharge Forecasting in Babahoyo, Ecuador: Benchmarking, Uncertainty, and Operational Limits Without In Situ Validation |
| title_short | Open Hydroclimatic Data and Iterative Machine Learning for River Discharge Forecasting in Babahoyo, Ecuador: Benchmarking, Uncertainty, and Operational Limits Without In Situ Validation |
| title_sort | open hydroclimatic data and iterative machine learning for river discharge forecasting in babahoyo ecuador benchmarking uncertainty and operational limits without in situ validation |
| topic | river discharge prediction GloFAS ERA5 reanalysis forecast uncertainty feature leakage tropical floodplain Babahoyo data-limited regions Hydrological Forecasting Using AI Hydrology and Watershed Management Studies Environmental Monitoring and Data Management |
| topic_facet | river discharge prediction GloFAS ERA5 reanalysis forecast uncertainty feature leakage tropical floodplain Babahoyo data-limited regions Hydrological Forecasting Using AI Hydrology and Watershed Management Studies Environmental Monitoring and Data Management |
| url | https://doi.org/10.3390/atmos17080808 |
| work_keys_str_mv | AT rolandolicaparedolfo openhydroclimaticdataanditerativemachinelearningforriverdischargeforecastinginbabahoyoecuadorbenchmarkinguncertaintyandoperationallimitswithoutinsituvalidation AT alexanderharosarango openhydroclimaticdataanditerativemachinelearningforriverdischargeforecastinginbabahoyoecuadorbenchmarkinguncertaintyandoperationallimitswithoutinsituvalidation AT persiverazelada openhydroclimaticdataanditerativemachinelearningforriverdischargeforecastinginbabahoyoecuadorbenchmarkinguncertaintyandoperationallimitswithoutinsituvalidation AT denisjavieraranguricayetano openhydroclimaticdataanditerativemachinelearningforriverdischargeforecastinginbabahoyoecuadorbenchmarkinguncertaintyandoperationallimitswithoutinsituvalidation AT roxanamabelsemperteguirafael openhydroclimaticdataanditerativemachinelearningforriverdischargeforecastinginbabahoyoecuadorbenchmarkinguncertaintyandoperationallimitswithoutinsituvalidation AT olegariocabreracabrera openhydroclimaticdataanditerativemachinelearningforriverdischargeforecastinginbabahoyoecuadorbenchmarkinguncertaintyandoperationallimitswithoutinsituvalidation AT edwarciezasanchez openhydroclimaticdataanditerativemachinelearningforriverdischargeforecastinginbabahoyoecuadorbenchmarkinguncertaintyandoperationallimitswithoutinsituvalidation AT marthaghuamantanta openhydroclimaticdataanditerativemachinelearningforriverdischargeforecastinginbabahoyoecuadorbenchmarkinguncertaintyandoperationallimitswithoutinsituvalidation AT dianacarolinacastillomartinez openhydroclimaticdataanditerativemachinelearningforriverdischargeforecastinginbabahoyoecuadorbenchmarkinguncertaintyandoperationallimitswithoutinsituvalidation |