Benchmarking Conceptual Rainfall–Runoff Models in Peninsular India: Simulated Streamflow and Evaluation Metrics (CAMELS-IND, MARRMoT)

This dataset accompanies the manuscript “Benchmarking and Selecting Optimal Hydrological Models for Large-Sample Applications Considering Complexity and Uncertainty” (Mangukiya et al., under review, Water Resources Research). It provides simulated streamflow outputs and evaluation metrics for 47 conceptual rainfall–runoff models applied across 159 watersheds in Peninsular India. Model simulations were generated using the Modular Assessment of Rainfall–Runoff Models Toolbox (MARRMoT) (Trotter & Knoben, 2023), which implements 47 established hydrological models within a unified and consistent framework. Model forcing data and catchment attributes were derived from the CAMELS-IND dataset (Mangukiya et al., 2025), which provides hydrometeorological time series and physiographic information for catchments in Peninsular India. The archive contains daily simulated streamflow time series for all models and watersheds, along with a comprehensive set of performance metrics computed during the evaluation period. While the associated manuscript primarily reports Nash–Sutcliffe Efficiency (NSE) and percent bias in high-flow volumes (FHV), this dataset additionally includes correlation coefficient, root mean square error (RMSE), Kling–Gupta Efficiency (KGE), percentage bias (PBIAS), and flow-duration-curve-based errors. Contents: simulated_streamflow: Daily simulated streamflow time series for all models and watersheds evaluation_metrics: Model performance metrics computed for each watershed for the validation period best_performing_model.csv: Summary of best-performing models per watershed based on selected evaluation criteria All simulations were produced using a consistent single-parameter-set calibration framework, as described in the associated manuscript. Users are encouraged to interpret the results within the context of the adopted calibration and benchmarking approach. This dataset is intended to support transparency, reproducibility, and future comparative studies on large-sample hydrological modeling, model benchmarking, and structural uncertainty. Related datasets and software: CAMELS-IND : Mangukiya, N. K., et al. (2025). CAMELS-IND: hydrometeorological time series and catchment attributes for 472 catchments in Peninsular India (v2.2). Zenodo. https://doi.org/10.5281/zenodo.11118408 MARRMoT : Trotter, L., & Knoben, W. J. M. (2023). MARRMoT (v2.1.2). Zenodo. https://doi.org/10.5281/zenodo.2482541 Related publications: Trotter, L., Knoben, W. J. M., Fowler, K. J. A., Saft, M., & Peel, M. C. (2022). Modular Assessment of Rainfall–Runoff Models Toolbox (MARRMoT) v2.1. Geoscientific Model Development, 15, 6359–6369. https://doi.org/10.5194/gmd-15-6359-2022 Mangukiya, N. K., et al. (2025). CAMELS-IND: hydrometeorological time series and catchment attributes for 228 catchments in Peninsular India. Earth System Science Data, 17, 461–491. https://doi.org/10.5194/essd-17-461-2025