Submarine groundwater discharge (SGD) is a spatially pervasive phenomenon that adds freshwater, nutrients, dissolved metals, bacteria, and other constituents to the coastal ocean. This dissertation investigated SGD-related inputs of nutrients and fecal indicator bacteria (Escherichia coli and Enterococcus sp.) to coastal waters in two Hawaiian locations, the north shore of Kaua'i and the Kona coast of Hawai'i. Concentrations of caffeine, which has been used previously as a wastewater tracer, were measured in groundwater and surface water on the north shore of Kaua'i. Both study areas have relatively light levels of urban and agricultural development, and maintaining good water quality is essential for their tourism-based economies, coral reefs, fisheries, and traditional way of life. Radium (Ra), an element with naturally elevated concentrations in coastal groundwater, was used as an SGD tracer and a mass-balance approach was used to quantify SGD. On the north shore of Kaua'i, agriculture was associated with higher nitrate + nitrite concentrations in the fresh SGD component, while phosphate and silica appeared to be controlled by geological differences in aquifer substrate. High ammonium concentrations in the fresh SGD component at one site may have been caused by a leaky cesspool. In Kona, no relation between urban development or agriculture and groundwater nutrient concentrations was observed, but bare lava rock was associated with higher nitrate + nitrite and silica concentrations in fresh SGD. Sites closer to golf courses also had higher nitrate + nitrite concentrations. Conservative estimates of total SGD on the north shore of Kaua'i ranged from 1.3 to 7.8 L per meter per minute, or up to 10% of Hanalei River discharge, and SGD contributed significant nitrate + nitrite inputs to Hanalei Bay. Estimates of SGD in Kona ranged from 5 to 1200 L per meter per minute, with between 10 and 100% of the brackish SGD comprised by the fresh SGD component. SGD-related water and nutrient fluxes on the Kona Coast -- where no rivers and streams are present -- were large compared to those reported for other sites worldwide. Caffeine concentrations in environmental waters on the north shore of Kaua'i ranged from 0-88 ng/L, on the low end of what has been reported for other locations. Metribuzin, an herbicide, was also detected at concentrations from 4-11 ng/L in five groundwater and surface water samples. A sensitivity analysis of Ra-based methods of estimating water ages and coastal mixing rates revealed that water ages shorter than 3 d cannot be estimated with confidence using Ra-based methods, even if the only uncertainty considered is analytical error. In conclusion, this dissertation provides new data about SGD and related inputs of nutrients and bacteria to Hawaiian coastal waters, suggests that even low levels of development may influence nutrient concentrations in coastal groundwater, presents the first caffeine concentrations measured in environmental water samples collected in a tropical setting, and explores the limits of applicability of Ra-based methods of estimating water ages and coastal mixing rates, providing guidance for researchers conducting Ra-based SGD studies in the future.

The focus of this dissertation is submarine groundwater discharge (SGD), the direct flow of groundwater from the seabed to the sea, and onsite wastewater treatment systems in coastal California. The research focuses primarily on a single coastal community in central California, Stinson Beach, where conventional onsite treatment systems, or septic systems, are used exclusively for wastewater disposal. The overarching goal of the work has been to quantify the magnitude and timing of SGD at the site and to provide insight into how onsite wastewater treatment at Stinson Beach affects local groundwater quality and, via SGD, surface water quality, all with the broader goal of informing and guiding future development along the California coast. The dissertation includes four research chapters, each focused on one or more important aspects of the issue of SGD and coastal septic systems at Stinson Beach. The first research chapter entitled "Submarine discharge of nutrient-enriched fresh groundwater at Stinson Beach, California is enhanced during neap tides" describes a 14-day study investigating the influence of fortnightly spring-neap tidal variability on submarine discharge of fresh and saline groundwaters at the site. Fresh, shallow groundwater at the site was observed to contain high concentrations of nutrients including dissolved inorganic nitrogen, soluble reactive phosphate, and silicate, as well as human fecal bacteria. A groundwater-derived freshening and nutrification of the surf zone was observed at neap tide, and was followed by a 4-day increase in chlorophyll a concentrations in the surf zone. Analytical models and a fresh water budget in the surf zone were used to estimate the saline and fresh SGD at both neap and spring tides. Fresh SGD at the site was estimated at between 1.2 and 4.7 L min-1 m-1 shoreline during neap tides compared to 0.1 and 0.5 L min-1 m-1 during spring tides. This compares to saline SGD estimates of 15.9 and 22.0 L min-1 m-1 during neap and spring tides, respectively. Despite the smaller total discharge of groundwater during neap compared to spring tides, the larger fresh discharge component during neap tides raised surf zone silicate, DIN, and SRP by 14%, 35%, and 27%, respectively, relative to spring tides. The observed fortnightly 'pulse' of fresh groundwater-derived nutrients into the surf zone was consistent with seaward hydraulic gradients across the fresh part of the beach aquifer, which varied due to aquifer overheight, or the mounding of groundwater due to variable infiltration of salt water during tides, adjacent the beach face. Darcy-Dupuit estimates of seaward fresh groundwater flow in this area agreed well with the fresh discharge results of the mass balance.vvThe second research chapter of the dissertation entitled "Submarine groundwater discharge to a high-energy surf zone at Stinson Beach, California, estimated using radium isotopes" describes and compares results from a pair of two-week long experiments conducted in the dry season (July 2006) and wet season (March 2007) to examine tidal, wave-driven and seasonal variability of SGD at the site using natural radium tracers. Tide stage, tide range, breaker height and season each explained a significant degree of radium variability in the surf zone. A mass balance of excess radium in the surf zone was used to estimate SGD and associated nutrient fluxes during each season, confirming larger discharge rates during the wet season. Our results indicate median groundwater discharge rates of 6 to 8 L min-1 m-1 in July 2006 and 38 to 43 L min-1 m-1 in March 2007. SGD from 200 m of Stinson Beach in March 2007 was shown to contribute a flux of phosphate and dissolved inorganic nitrogen approximately equal to that associated with all local creeks and streams within 6 km of the study site at that time. The third research chapter "Fresh submarine groundwater discharge from a coastal aquifer forced by the Mediterranean climate of central California" is a numerical investigation of groundwater flow at the land-sea interface forced by precipitation and evapotranspiration typical of the Mediterranean climate of coastal California. A numerical groundwater model was developed using the variable density groundwater flow code SEAWAT-2000 to examine the influence of seasonally variable recharge conditions typical of coastal California on the magnitude and timing of fresh submarine groundwater discharge from a generic coastal aquifer with a constant head (non-tidal) ocean boundary. Model dimensions and hydrogeologic characteristics were chosen based on a combination of observations from field studies at Stinson Beach, California, and published numerical investigations of coastal groundwater flow. Average monthly recharge was calculated from historical precipitation records and potential evapotranspiration rates calculated from climatological observations made near the field site. Calculated recharge was approximately sinusoidal across the year, with positive recharge rates dominated by precipitation during the rainy winter and negative recharge rates dominated by evapotranspiration during the hot, precipitation-free summer. Rates of fresh discharge from the model aquifer to the ocean exhibited similar temporal characteristics for two modeled scenarios, a first including a constant head fresh landward boundary condition and a second including a constant flux fresh landward boundary condition. Discharge in both models peaked in January during the period of maximum precipitation and recharge, and declined until reaching a minimum in September, two months after the minimum recharge period in July. Minimum simulated discharge rates for two simulated scenarios were 17% and 18% lower in September than the maximum simulated discharges in winter. Monthly mean discharge from Lagunitas Creek, a creek near Stinson Beach, reached maximum and minimum values in February and September, respectively. The exponential decline in creek discharge was fast compared to the decline in modeled SGD, however, suggesting that fresh SGD and associated nutrient fluxes may play a particularly important role in coastal ecosystems in early summer when surface water discharge has nearly reached a minimum but discharge of substantial quantities of fresh groundwater is still substantial. The final research chapter "Nitrogen, fecal indicator bacteria, and coliphage attenuation and flux from a septic leach field to the coastal ocean" describes a two-year field study to measure the flux and attenuation of nitrogen, fecal indicator bacteria, and bacteriophage in groundwater adjacent to a large coastal septic system in Central California. The study was carried out at Stinson Beach Park, Golden Gate National Recreation Area, sixteen kilometers northwest of the San Francisco Golden Gate Bridge. Long-term measurements of septic effluent quality and volumetric discharge to the leach field, synoptic DC resistivity profiling of the saltwater/freshwater interface, continuous measurements of hydraulic head in the coastal aquifer, and the installation and subsequent monitoring of a dense array of multi-level monitoring wells adjacent to the leach field for chemical and microbiological constituents were carried out. Our results indicate a nitrogen- and inorganic carbon-rich plume of septic effluent flowing from the leach field through the beach to the subterranean estuary, or the mixing zone of fresh and saline groundwaters. Attenuation of E. coli and coliphage was complete within the vadose zone and the first few meters of transport. Enterococci were detected throughout the well network during one sampling event during which no attenuation was observed, and no attenuation of total nitrogen was observed along the flowpath during the experiment. Median estimates of total nitrogen fluxing toward the ocean downgradient from the leach field ranged from 1.6 to 70.6 moles day-1, depending on season and transect location. Except for enterococcus, the behavior of nitrogen and microbial pollutants in the field was consistent with results from laboratory experiments, which demonstrated low denitrification potential in slurry tests, but fast fecal indicator bacteria and virus attenuation rates in saturated column experiments. Comparisons of total nitrogen flux to the subterranean estuary in this study agree well with SGD-associated nutrient flux estimates from prior studies at the site, suggesting that septic systems at the site are a persistent source of nitrogen to the subterranean estuary and may at times also be a source of enterococci. Denitrification potentials measured at the site suggest a possible role for in-situ remediation strategies to optimize nutrient removal in the beach aquifer.

Sustainable management of water resources is quickly increasing in importance on a global scale. An important piece of the puzzle is the characterization of marine water and determining its importance to geochemical budgets. To do this, submarine groundwater discharges must be carefully studied. Comprehensively exploring the subject, Submarine G

Submarine groundwater discharge (SGD) is a spatially pervasive phenomenon that adds freshwater, nutrients, dissolved metals, bacteria, and other constituents to the coastal ocean. This dissertation investigated SGD-related inputs of nutrients and fecal indicator bacteria (Escherichia coli and Enterococcus sp.) to coastal waters in two Hawaiian locations, the north shore of Kaua'i and the Kona coast of Hawai'i. Concentrations of caffeine, which has been used previously as a wastewater tracer, were measured in groundwater and surface water on the north shore of Kaua'i. Both study areas have relatively light levels of urban and agricultural development, and maintaining good water quality is essential for their tourism-based economies, coral reefs, fisheries, and traditional way of life. Radium (Ra), an element with naturally elevated concentrations in coastal groundwater, was used as an SGD tracer and a mass-balance approach was used to quantify SGD. On the north shore of Kaua'i, agriculture was associated with higher nitrate + nitrite concentrations in the fresh SGD component, while phosphate and silica appeared to be controlled by geological differences in aquifer substrate. High ammonium concentrations in the fresh SGD component at one site may have been caused by a leaky cesspool. In Kona, no relation between urban development or agriculture and groundwater nutrient concentrations was observed, but bare lava rock was associated with higher nitrate + nitrite and silica concentrations in fresh SGD. Sites closer to golf courses also had higher nitrate + nitrite concentrations. Conservative estimates of total SGD on the north shore of Kaua'i ranged from 1.3 to 7.8 L per meter per minute, or up to 10% of Hanalei River discharge, and SGD contributed significant nitrate + nitrite inputs to Hanalei Bay. Estimates of SGD in Kona ranged from 5 to 1200 L per meter per minute, with between 10 and 100% of the brackish SGD comprised by the fresh SGD component. SGD-related water and nutrient fluxes on the Kona Coast -- where no rivers and streams are present -- were large compared to those reported for other sites worldwide. Caffeine concentrations in environmental waters on the north shore of Kaua'i ranged from 0-88 ng/L, on the low end of what has been reported for other locations. Metribuzin, an herbicide, was also detected at concentrations from 4-11 ng/L in five groundwater and surface water samples. A sensitivity analysis of Ra-based methods of estimating water ages and coastal mixing rates revealed that water ages shorter than 3 d cannot be estimated with confidence using Ra-based methods, even if the only uncertainty considered is analytical error. In conclusion, this dissertation provides new data about SGD and related inputs of nutrients and bacteria to Hawaiian coastal waters, suggests that even low levels of development may influence nutrient concentrations in coastal groundwater, presents the first caffeine concentrations measured in environmental water samples collected in a tropical setting, and explores the limits of applicability of Ra-based methods of estimating water ages and coastal mixing rates, providing guidance for researchers conducting Ra-based SGD studies in the future.

Recent studies have provided evidence that submarine groundwater discharge is a signifi cant source of water and dissolved nutrients to the coastal ocean. The chemical implications of these studies are especially important because, relative to surface water, groundwater is typically enriched in many compounds including nitrogen bearing nutrients. Therefore by affecting the supply and relative proportions of essential nutrients, direct groundwater discharge has the potential to influence phytoplankton populations and estuarine ecosystems as a whole. Another potential submarine discharge that may occur in the shallow restricted waters of the Texas coast is leakage of oil-field brine. Such leakage also has important ecological implications similar in some ways to groundwater discharge. The studies in this dissertation concern improving the methods and techniques used in measuring submarine discharges. Using the Texas Coastal Bend as a study area I have: 1) conducted a detailed evaluation of submarine discharges to Nueces Bay and 2) compared indications of submarine discharge between Nueces, Baffin, and Copano Bays. These investigations use a combination of geochemical and geophysical techniques. The geochemical methods are based primarily on measurements of naturally-occurring dissolved Ra isotopes in samples of bay, river, ocean, and groundwater. The geophysical methods employ electrical resistivity profiling to look for evidence of groundwater movement within the bay bottom sediments. Results show that dissolved radium concentrations within Nueces Bay are among the highest observed in coastal estuaries. Geochemical analysis and geophysical surveys indicate that both groundwater and leakage of oil-field brine are potential submarine inputs. Samples from Nueces, Copano, and Baffin Bays show that the seasonal increase in dissolved 226Ra activity for Nueces Bay is substantially larger than that of the other two bays. This increase is not readily explained by either evaporation or riverine supply. These results clearly suggest that the Ra supply to Nueces Bay is unusually large. For Nueces Bay, the most relevant differences between the three bays that might account for this are 1) the proportionally larger salt marsh and 2) the higher density of petroleum wells and pipelines. Though submarine groundwater discharge is not to be ruled out, leakage of oil-field brine is strongly indicated.