Rainfall–runoff modelling in a tropical climate (Citarum basin, West Java, Indonesia): model and climate data intercomparison
Published in Hydrology Research, 2025
Hydrological research in tropical regions has been limited by a lack of high-quality precipitation and evapotranspiration input data, and high-quality observed runoff calibration data. Hence, most hydrological models have been developed for non-tropical regions and calibrated on non-tropical climatic data. To address these challenges, we assessed the performance of 24 rainfall–runoff models in combination with seven precipitation and three evapotranspiration datasets (504 combinations), in three catchments in the Citarum basin, Indonesia, to identify which models and input datasets were best suited to modelling daily mean flow in a volcanic region and tropical climate, and to assess the parameter sensitivity of the most suitable models. We found that model performance was most strongly affected by the choice of input precipitation data and that the most suitable option was different for each of the three study catchments, varying over tens of kilometres. Hydrological model structure played a less significant role, and models developed explicitly for volcanic or tropical regions did not show a clear advantage over those designed for other regions or climates. However, models with two separate, parallel, fast and slow routing pathways had higher predictive skill than those with just one pathway in representing the year-round range of flow behaviours.
Recommended citation: Vesuviano G., Mathias S. A., Rahayu R., Reaney S. M.; Rainfall–runoff modelling in a tropical climate (Citarum basin, West Java, Indonesia): model and climate data intercomparison. Hydrology Research 2025; nh2025021. doi: https://doi.org/10.2166/nh.2025.021