Can runoff modeled at coarse resolution simulate floods at finer resolutions? A case study over the Ohio River Basin

• The VIC model with a kinematic wave routing model reasonably simulates flood statistics and reproduces flood events. • Smaller catchments with dams have large errors, but neglecting dams does not impact model results in larger basins. • VIC-simulated runoff and the subsequent routed floods are insensitive to subdaily model timesteps but are sensitive to spatial resolution. Recent studies on flood-generating mechanisms have advanced understanding of the hydrologic processes that lead to riverine flooding. Flood modeling frameworks can play a role in furthering this research, but they can be computationally intensive. This study tests the ability to simulate floods using the widely used Variable Infiltration Capacity (VIC) land surface model with a kinematic wave routing model over the Ohio River basin, which is flood-prone and topographically variable. Using 200 USGS streamflow gauges, the model estimates the median annual maximum daily flow (AMF) with an average bias of 1.8% across the gauges and the 90th percentile AMF with an average bias of 6.2% for 1979–2022. Errors tend to be larger in flatter regions and in smaller basins with dams, highlighting the role dams play in reducing flood peaks in this basin. Model experiments show that the simulated AMF is not sensitive to the subdaily model timestep, but it is sensitive to the spatial resolution, with larger grid cells resulting in underestimating AMF. Overall, this modeling framework reproduces flooding across a range of land cover, topography, and drainage area, indicating it can be used in future studies to investigate flood generating mechanisms and flood risk estimation.