Bias Correction of CHIRPS Precipitation for Hydrological Modeling and Return Period Analysis in the Data Scarce Zapote River Basin, Costa Rica
Salvato in:
| Autori: | , |
|---|---|
| Data di pubblicazione: | 2026 |
| Descrizione: | Estimating extreme precipitation and flood response remains challenging in tropical basins with sparse and discontinuous rain-gauge records. This study applied linearregression- based bias correction to CHIRPS data to reconstruct a continuous series of annual maximum 24-h precipitation series for the Zapote River Basin, Costa Rica over the 1993-2023 period. Two non-overlapping rain-gauge records (1984–1994 and 2003–2022) were used, with 70% of the observations allocated to regression fitting and 30% to independent validation. Performance was evaluated using NSE, PBIAS, RMSE, and RSR. The reconstructed series was fitted to a Gumbel distribution, assessed using the Kolmogorov–Smirnov test, and used to derive IDF curves and design hyetographs for 25-, 50-, and 100-year return periods. The main findings were as follows: (1) NSE values were 0.90 and 0.50, PBIAS values were −2.0% and −13.7%, RSR values were 0.30 and 0.60, and RMSE values were 10.0 and 16.4 mm for stations 69679 and 69538, respectively; (2) design precipitation increased from 107 to 130 mm; and (3) HEC-HMS produced uncalibrated peak discharges ranging from 410.6 to 534.1 m³/s, while Iber identified potential overbank flooding and increasing water depths in the lower basin. Because streamflow observations were unavailable, these outputs represent physically plausible design scenarios rather than validated predictions. The proposed workflow supports preliminary flood hazard assessment in data-scarce tropical basins. |
| Stato: | Kérwá |
| Istituzione: | Universidad de Costa Rica |
| Repositorio: | Kérwá |
| Lingua: | Inglés |
| OAI Identifier: | oai:kerwa.ucr.ac.cr:10669/105137 |
| Accesso online: | https://hdl.handle.net/10669/105137 https://doi.org/10.1016/j.hydrch.2026.100051 |
| Keyword: | Bias-corrected CHIRPS Flow IDF Curve Precipitation |