Using High Resolution Mass Spectrometry to Evaluate the Impact of Hurricane Florence in 2018 on Water Quality of North Carolina Coastal Aquifers

With over half of North Carolinians reliant on groundwater for their drinking water supply, monitoring and maintaining the water quality of the North Carolina Coastal Plain (NCCP) aquifers is vital. As rising global temperatures result in more frequent and intense storms in the NCCP and elsewhere, we need to understand the impacts of these events on the water quality of coastal aquifers. Hurricane Florence caused record flooding in the NCCP in 2018 and serves as an excellent example of what could take place globally. Pre- and post-hurricane groundwater samples were collected and analyzed by EPA Method 625 targeting 180 regulated organic chemicals and by gas chromatography with high resolution mass spectrometry (GC-HRMS) to perform non-targeted and suspect screening analysis of organic chemicals. The EPA Method 625 resulted in no detections in any samples. The higher sensitivity GC-HRMS method resulted in over 5,000 chemical features in all samples, with post-hurricane samples having a small decrease in unique features for surficial wells and a small increase for confined wells. Processing the GC-HRMS data through both the NIST 17 mass spectral database (M¹) and the US EPA ToxCast database resulted in far fewer tentative identifications, with ~150 and ~20 chemicals identified using 75% and 90% match criteria, respectively. The identified chemicals included aromatic hydrocarbons, pesticides, phthalates, among other classes. The post-flood increases in groundwater chemicals included highly water soluble chemicals such as the herbicide sulfentrazone (log K_OW = 0.99), but also included many less water soluble chemicals such as pyrene (log K_OW = 4.88) and various phthalates (log K_OW range of ~2.5-8). These results indicate that groundwater vulnerability during extreme flood events cannot be predicted based simply on water solubility principles; rather, higher sensitivity and higher resolution analytical methods are needed to generate data for both aquifer vulnerability modeling and human health risk assessments. This work also has important implications globally, as groundwater is the primary source of drinking water for over 1.5 billion people worldwide.