GROUNDWATER INPUTS AND WATER QUALITY DURING A RECORD-SETTING FLOOD ON THE MERAMEC RIVER, EAST-CENTRAL MISSOURI

Floods are among the most frequent and ubiquitous natural disasters, causing loss of life and billions of dollars of property damage on an annual basis. Due to the rapid onset and unpredictable nature of flood events, few studies have offered high-resolution analyses of flow components and water quality during extreme floods. This study examines the hydrologic and geochemical responses of the Meramec River in east-central Missouri as well as two of its largest tributaries, the Big River and Bourbeuse River, during a record-setting flood in April to May of 2017. Over the course of 8.1 d, large amounts of precipitation (26.0 cm) fell on the basin leading to peak discharges of 4,785 m³/s on the Meramec River, 1,710 m³/s on the Big River, and 1,226 m³/s on the Bourbeuse River, representing 79-, 89-, and 171-fold increases from baseflow conditions, respectively. Peak discharges along portions of the Meramec River and Big River are the largest on record. We deployed a continuous water quality monitoring device in the Meramec River prior to the flood to record temperature, specific conductivity (SpC), Cl^-, pH, and turbidity. We also made point measurements of the same parameters and collected water samples at the Meramec River (n = 25), Big River (n = 10), and Bourbeuse River (n = 10) to measure nutrients, major and minor elements, and water isotopes. Expectedly, major ions decreased (by up to 98%) following peak discharge. However, Fe, Al, and PO₄^3- increased (by up to 330%), likely due to sediment suspension. We used our SpC data to quantify the relative contribution of pre-event (baseflow) and event (recent rainfall) water using two-component hydrograph separations. Integrating over the entire flood, baseflow contributed to 41%, 27%, and 28% of the total flow for the Meramec River, Big River, and Bourbeuse River, respectively, while baseflow inputs were correspondingly 30%, 16%, and 20% at peak discharge. These values are surprisingly high given the severity and magnitude of the flooding. Indeed, the 2017 flood exhibited baseflow contributions similar to smaller floods that generally have >70% lower peak flows. These large baseflow inputs during the 2017 flood are likely partially the consequence of rainfall delivery over a longer time period (i.e., 8.1 d), but this unexpected result for a record-setting flood requires further research.