Large Floods on the Ohio River Inferred from Slackwater Deposits

Riverine flooding causes millions of dollars of damage annually, but existing estimates of flood hazard are often limited by the short timescale of recorded data. Paleoflood hydrology is a valuable tool used to extend the systematic record and reduce uncertainty in flood frequency analysis results. Slackwater deposits are traditionally used in paleoflood hydrology as flood event markers, but this approach has typically been limited to arid environments. However, slackwater deposits were recently found along the Tennessee River, demonstrating the potential for robust reconstructions of flood discharges in alluvial settings in the temperate eastern United States. In this study we used a combination of hydraulic modeling, field sampling, and geochemical analyses to identify and distinguish slackwater deposits along the lower Ohio River. A HEC-RAS model of the Ohio River was used to simulate river stages at a given discharge, evaluate inundation extent during past floods, and locate sites where slackwater deposits from large floods may be found. We found evidence of two large flood events in the form of distinct sedimentary layers buried in an alcove 12 m above the bankfull elevation of the river. They likely correspond to the 1937 flood of record on the lower Ohio River and a large flood around AD 1650 identified in multiple studies along the Tennessee River. We used X-ray diffraction and particle size analysis to confirm that the mineralogy and grain size distribution of the flood unit matches that of sediments transported by the Ohio River load, and to distinguish it from the sediment that was produced by in-situ weathering of local bedrock. This work will allow us to better constrain flood hazard on the Ohio River by reducing uncertainty in flood frequency analysis estimates and allows for to the application of slackwater deposits in other temperate river corridors.