Tracing Precipitation Sources of Extreme Flooding During Hurricane Florence Using Oxygen and Hydrogen Isotope Ratios and Back-Trajectory Modeling

Between September 12 and 17, Hurricane Florence produced nearly three feet of rain in some parts of the coastal Carolinas and caused catastrophic flooding as far inland as Durham, North Carolina. To better understand the local variability in precipitation source throughout the storm, we use a combination of stable isotope analysis of hydrogen and oxygen isotope ratios and back-trajectory analyses produced using the HYSPLIT model. We collected precipitation samples every two hours in Durham and when precipitation totals allowed in Hillsborough, North Carolina, from September 15 through 17 as tropical storm conditions impacted the Piedmont. δ¹⁸O and δD values varied considerably throughout the storm, ranging from -7.59‰ to -0.20‰ and -45.70‰ to 3.40‰, respectively. The stable isotope ratios for both locations demonstrate a pattern of initially lower values (ranging from -4.94‰ to -3.64‰ and -29.58‰ to -20.83‰) and an initial peak (as high as-0.58‰ and 5.78‰). They then exhibit a moderate drop and a second peak, followed by an abrupt shift to markedly lower values (-7.59‰ and -45.70‰ in Hillsborough and -7.50‰ and -46.05‰ in Durham). Both datasets also demonstrate a return to higher values towards the end of the event. HYSPLIT backward trajectories indicate that low-level (1000m AMSL) air masses that led to the precipitation shift from a more cyclonic path (linked with the cyclonic movement of Florence) to an overall southeasterly and eventually southerly path over the course of the storm as moisture sources shifted from Florence itself to the funneling of moisture directly from the southeast and south as Florence stalled due to an upper-level high pressure ridge. These results demonstrate the relationship between the stable isotope composition of Florence's rainfall and the large-scale atmospheric features that drove that rainfall as evidenced by the back-trajectory model. Examination of this relationship adds nuance to our understanding of the relationship between stable isotope signatures of precipitation and the conditions that produced that precipitation and can aid our interpretation of stable isotope records captured by climate proxies such as tree rings.