Search

Key characteristics of California groundwater systems related to aquifer vulnerability, sustainability, recharge locations and mechanisms, and anthropogenic impact on recharge are revealed in a spatial geostatistical analysis of a unique data set of tritium, noble gases and other isotopic analyses unprecedented in size at nearly 4000 samples. The correlation length of key groundwater residence time parameters varies between tens of kilometers ((3)H; age) to the order of a hundred kilometers ((4)Heter; (14)C; (3)Hetrit). The correlation length of parameters related to climate, topography and atmospheric processes is on the order of several hundred kilometers (recharge temperature; δ(18)O). Young groundwater ages that highlight regional recharge areas are located in the eastern San Joaquin Valley, in the southern Santa Clara Valley Basin, in the upper LA basin and along unlined canals carrying Colorado River water, showing that much of the recent recharge in central and southern California is dominated by river recharge and managed aquifer recharge. Modern groundwater is found in wells with the top open intervals below 60 m depth in the southeastern San Joaquin Valley, Santa Clara Valley and Los Angeles basin, as the result of intensive pumping and/or managed aquifer recharge operations., (Copyright © 2016 Elsevier Ltd. All rights reserved.)

Data from 11,000 public supply wells in 87 study areas were used to assess the quality of nearly all of the groundwater used for public supply in California. Two metrics were developed for quantifying groundwater quality: area with high concentrations (km(2) or proportion) and equivalent-population relying upon groundwater with high concentrations (number of people or proportion). Concentrations are considered high if they are above a human-health benchmark. When expressed as proportions, the metrics are area-weighted and population-weighted detection frequencies. On a statewide-scale, about 20% of the groundwater used for public supply has high concentrations for one or more constituents (23% by area and 18% by equivalent-population). On the basis of both area and equivalent-population, trace elements are more prevalent at high concentrations than either nitrate or organic compounds at the statewide-scale, in eight of nine hydrogeologic provinces, and in about three-quarters of the study areas. At a statewide-scale, nitrate is more prevalent than organic compounds based on area, but not on the basis of equivalent-population. The approach developed for this paper, unlike many studies, recognizes the importance of appropriately weighting information when changing scales, and is broadly applicable to other areas.

Aquifer vulnerability models were developed to map groundwater nitrate concentration at domestic and public-supply well depths in the Central Valley, California. We compared three modeling methods for ability to predict nitrate concentration >4 mg/L: logistic regression (LR), random forest classification (RFC), and random forest regression (RFR). All three models indicated processes of nitrogen fertilizer input at the land surface, transmission through coarse-textured, well-drained soils, and transport in the aquifer to the well screen. The total percent correct predictions were similar among the three models (69-82%), but RFR had greater sensitivity (84% for shallow wells and 51% for deep wells). The results suggest that RFR can better identify areas with high nitrate concentration but that LR and RFC may better describe bulk conditions in the aquifer. A unique aspect of the modeling approach was inclusion of outputs from previous, physically based hydrologic and textural models as predictor variables, which were important to the models. Vertical water fluxes in the aquifer and percent coarse material above the well screen were ranked moderately high-to-high in the RFR models, and the average vertical water flux during the irrigation season was highly significant (p < 0.0001) in logistic regression.

Twenty-four halogenated volatile organic compounds (hVOCs) and SF₆ were measured in groundwater samples collected from 312 wells across California at concentrations as low as 10⁻¹² grams per kilogram groundwater. The hVOCs detected are predominately anthropogenic (i.e., "ahVOCs") and as such their distribution delineates where groundwaters are impacted and susceptible to human activity. ahVOC detections were broadly consistent with air-saturated water concentrations in equilibrium with a combination of industrial-era global and regional hVOC atmospheric abundances. However, detection of ahVOCs in nearly all of the samples collected, including ancient groundwaters, suggests the presence of a sampling or analytical artifact that confounds interpretation of the very-low concentration ahVOC data. To increase our confidence in ahVOC detections we establish screening levels based on ahVOC concentrations in deep wells drawing ancient groundwater in Owens Valley. Concentrations of ahVOCs below the Owens Valley screening levels account for a large number of the detections in prenuclear groundwater across California without significant loss of ahVOC detections in shallow, recently recharged groundwaters. Over 80% of the groundwaters in this study contain at least one ahVOC after screening, indicating that the footprint of human industry is nearly ubiquitous and that most California groundwaters are vulnerable to contamination from land-surface activities.