Over the last two decades, studies of the chemistry of the coastal ocean have found high concentrations of natural tracers that can only be explained by salty groundwater discharging from the seafloor. Based on the tracer concentrations, the volume of this discharge is very large (exceeding discharge from rivers), but it was not clear where such a large volume of groundwater could be discharging, undetected for decades. One explanation was that the tracer could be supplied by slow (m/yr) discharge of saline groundwater spread over the very large area of the continental shelf. The specific goal of this project was to see if we could detect this type of groundwater discharge relatively farr offshore. We installed a new wellfield 5-15 km offshore near Charleston, SC, to collect temperature data and wrote numerical models to analyze this data. Results revealed groundwater flow that varied in space and time, ranging from downward flow at rates of 5 m/yr to upward flow reaching rates of 20 m/yr for a short time. These results support slow background flow of varying rates (-5 to 5 m/yr) punctuated by periods of significant (15-20 m/yr) upward flow that coincide with strong ocean upwelling events, in which cold seawater from the margin of the continental shelf travels up onto the continental shelf. The links between ocean upwelling and groundwater discharge are still under investigation, but we hypothesize that the marine conditions that allow upwelling also allow upward flow of groundwater. These pulses of groundwater discharge may also provide a mechanism for previously-unexplained seasonal variations in tracer concentrations. Together, these results support the existence of very large exchanges of saline groundwater across the continental shelf and suggest a previously-unknown mechanism for groundwater flow, associated with marine upwelling events. Geochemical analyses of subseafloor groundwater confirmed that discharging groundwater is a significant source of nutrients, carbon, and other dissolved constitutuents to the the ocean. Our work suggests that textbook diagrams of the hydrologic cycle will need to be redrawn in coming years to include new, large flows from continental shelves to the ocean. Ocean budgets will similarly need to be recalculated to include large inputs of carbon, nutrients, and other constituents that are delivered to the ocean via this flow. The project also had significant broader educational impacts. The project trained three PhD students (one female), two M.S. students (both female), 11 undergraduates (two African-American, 8 female) and one high school student. The project also established the Girls Go for I.T. summer camp outreach program, which is designed to for middle-school-aged girls to become familiar with coding concepts and to encourage them to consider STEM careers. Since 2014, the camp has reached a diverse group of approximately 150 girls. Our demographics in 2018 were 50% African American, 35% Caucasian, 15% Asian or other. This camp also (1) trained a local middle school teacher, who went on to found an IT club at her middle school (2) provided material for an after-school coding club at a local middle school with a very high fraction of underserved students (15 boys and girls), and (3) was adapted to provide a summer coding experience for 4th-6th graders at a summer full-day program (25 boys and girls). Last Modified: 08/16/2018 Submitted by: Alicia M Wilson