2.5 billion people around the world rely on groundwater as their source of drinking water. These people reside primarily in rural areas, where extensive groundwater use by both domestic and agricultural users has caused 20 % of the worlds aquifers to be considered overdrafted. Similar trends can be seen in California’s Central Valley, where a majority of the United States fruits, nuts, and vegetables are produced. The Tulare Lake Basin, occupying the lower one-third of the Central Valley, is home to four of the top ten agricultural counties in the United States. Agricultural production in this region, enabled by a Mediterranean climate, has caused an average groundwater table decline of 0.11 m per year (2.3 km3 per year) over the last six decades. Groundwater table decline in the area furthermore causes issues with water quality as younger, potentially anthropogenic contaminant laden water, is drawn into previously un-contaminated wells. Due to having some of the highest poverty rates in the State of California, communities in the Tulare Lake Basin are often unable to deploy expensive methods, such as well water treatment plants or injection wells, of dealing with these water quality and quantity issues. Managed Aquifer Recharge (MAR) has been proposed as a cost-effective method to remedy both of these concerns by intentionally recharging an aquifer to raise water tables, dilute toxic substances or create other environmental benefit. However, the potential for MAR to mobilize contaminants as recharging water interacts with native aquifer conditions is poorly understood. To address this knowledge gap, we investigated how the source of groundwater recharge, geochemical factors, and geospatial variables (e.g. applied fertilizer amounts, annual precipitation totals etc.) are related to six prevalent or emerging contaminants (arsenic, nitrate, uranium, hexavalent chromium, fluoride, perchlorate) found in drinking water supply wells in the Tulare Lake Basin. Identifying the likely source of water in 786 drinking water supply wells involved using well construction data and concentrations of tritium, stable water isotopes, and dissolved noble gas parameters in groundwater, surface water and precipitation samples in the region. The contaminant distribution of each well water source was then compared by the Tukey-Kramer test, while geochemical and geospatial factors were correlated to contaminants by a matrix of Spearman’s rank correlation coefficient. Results indicate that recharge from un-evaporated surface waters (e.g. stream water) pose no excess risk to drinking water from groundwater and that toxic substances are primarily influenced by geochemical factors and nearby urban sites of chemical production. This suggests that the recharge of clean water through conveyance infrastructure or recharge basins can improve groundwater availability without sacrificing its drinkability, as long as both subsurface geochemistry is properly anticipated and urban manufacturing runoff is minimally present in the source area of a well or potential MAR site. Decision makers can use this information, along with existing demographic based recommendations, to more equitably secure water supplies through cost effective surface recharge technologies.

Dependence on groundwater for irrigation and consumptive use has resulted in the widespread depletion of groundwater aquifers across the world. In most of the semi-arid Southwest of the United States, groundwater is increasingly being regulated in efforts to sustainably manage this limited resource. Various Managed Aquifer Recharge (MAR) techniques are used to increase the sustainable management of groundwater resources. Agricultural Managed Aquifer Recharge (Ag-MAR) is a promising form of managed aquifer recharge, where farmland is flooded during the winter using excess surface water resources from runoff in order to recharge the underlying groundwater aquifer. In addition to increasing the security of groundwater resources and improving general drought resilience, Ag-MAR may have additional beneficial outcomes including flood risk reduction, drought preparedness, maintenance of environmental flows in aquatic ecosystems, prevention of saltwater intrusion in coastal aquifer systems, mitigating land subsidence as well as water quality benefits such as flushing salts from the shallow vadose zone. One main concern surrounding Ag-MAR implementation is the potential for increased nitrate (NO3-) leaching from historically cultivated agricultural land. When ingested in high concentrations, NO3- has been linked to methaemoglobinaemia (“blue baby syndrome”), miscarriages, and non-Hodgkin’s lymphoma. In order to evaluate the general viability of Ag-MAR as an appropriate MAR technique that presents minimal risk to agricultural production systems and groundwater contamination, a thorough understanding of nitrogen cycle dynamics under Ag-MAR must be developed. This dissertation focuses on evaluating the hydrologic and biogeochemical processes driving NO3- leaching and nitrogen cycling in the shallow vadose zone, under varying soil textures and Ag-MAR best management practices. First, I present field and laboratory experiments, which examine how different recharge practices affect NO3- leaching, mineralization and denitrification processes in different soil textures. Results show that short-lived, pulsed Ag-MAR flooding events cause NO3- leaching and organic N mineralization, whereby the dominant soil texture of an Ag-MAR site impacts both the timing of NO3- leaching, and the conditions for biogeochemical processes under Ag-MAR. Specifically, reducing time between flooding events for Ag-MAR reduces mineralization/nitrification, which in turn decreases mass of NO3- leached. This has implications for the development of Ag-MAR best management practices (BMPs), suggesting that in a N mineralization dominated system, short flooding frequencies may decrease mineralization/nitrification and NO3- leaching potential of Ag-MAR. Next, I developed a dual-porosity, mobile-immobile zone (MIM), reactive transport model using HP1 (HYDRUS-1D and PHREEQC), simulating NO3- leaching and biogeochemical processes under large water application events using observed datasets from the previous field and laboratory experiments. When comparing this HP1-MIM model to traditional NO3- leaching models, I find that the incorporation of environmental factors, and physical non-equilibrium dynamics improve model performance when estimating cumulative NO3- leached from the shallow vadose zone, and the amount of NO3- in the residual soil profile following water application events. Finally, I use this reactive transport model to perform a multi-scenario analysis, examining NO3- leaching potential, residual NO3- in the soil profile after water application events, and biogeochemical rates during water application events, over a variety of Ag-MAR best management practices (flooding frequency, flooding magnitude) and climate scenarios (dry year, wet year) to determine what Ag-MAR practices minimize NO3- leaching to the groundwater. For this analysis, I estimate the range of NO3- leached under Ag-MAR for two soil textures. Results of this analysis indicate that soil texture has a large influence on the moisture and oxygen regime of the soil, which in turn defines whether the soil system is dominated by denitrification or mineralization during and after Ag-MAR events. This has implications regarding Ag-MAR best management practices, and the general viability of Ag-MAR implementation. Across all soil types, we see the ability of high-magnitude water applications to dilute the mass of NO3- being leached below the shallow vadose zone to low concentrations in the bulk recharge. This indicates that under proper management, NO3- leaching during high magnitude water applications can be minimized.