John A. Izbicki

Artificial Recharge Through a Thick, Heterogeneous Unsaturated Zone

Thick, heterogeneous unsaturated zones away from large streams in desert areas have not previously been considered suitable for artificial recharge from ponds. To test the potential for recharge in these settings, 1.3 x 10(6) m(3) of water was infiltrated through a 0.36-ha pond along Oro Grande Wash near Victorville, California, between October 2002 and January 2006. The pond overlies a regional pumping depression 117 m below land surface and is located where thickness and permeability of unsaturated deposits allowed infiltration and saturated alluvial deposits were sufficiently permeable to allow recovery of water. Because large changes in water levels caused by nearby pumping would obscure arrival of water at the water table, downward movement of water was measured using sensors in the unsaturated zone. The downward rate of water movement was initially as high as 6 m/d and decreased with depth to 0.07 m/d; the initial time to reach the water table was 3 years. After the unsaturated zone was wetted, water reached the water table in 1 year. Soluble salts and nitrate moved readily with the infiltrated water, whereas arsenic and chromium were less mobile. Numerical simulations done using the computer program TOUGH2 duplicated the downward rate of water movement, accumulation of water on perched zones, and its arrival at the water table. Assuming 10 x 10(6) m(3) of recharge annually for 20 years, a regional ground water flow model predicted water level rises of 30 m beneath the ponds, and rises exceeding 3 m in most wells serving the nearby urban area.

Application of Flowmeter and Depth‐Dependent Water Quality Data for Improved Production Well Construction

Ground water production wells commonly are designed to maximize well yield and, therefore, may be screened over several water-bearing zones. These water-bearing zones usually are identified, and their hydrogeologic characteristics and water quality are inferred, on the basis of indirect data such as geologic and geophysical logs. Production well designs based on these data may result in wells that are drilled deeper than necessary and are screened through zones having low permeability or poor-quality ground water. In this study, we examined the application of flowmeter logging and depth-dependent water quality samples for the improved design of production wells in a complex hydrogeologic setting. As a demonstration of these techniques, a flowmeter log and depth-dependent water quality data were collected from a long-screened production well within a multilayered coastal aquifer system in the Santa Clara-Calleguas Basin, Ventura County, California. Results showed that the well yields most of its water from four zones that constitute 58% of the screened interval. The importance of these zones to well yield was not readily discernible from indirect geologic or geophysical data. The flowmeter logs and downhole water quality data also show that small quantities of poor-quality water could degrade the overall quality of water from the well. The data obtained from one well can be applied to other proposed wells in the same hydrologic basin. The application of flowmeter and depth-dependent water quality data to well design can reduce installation costs and improve the quantity and quality of water produced from wells in complex multiple-aquifer systems.

Simulation of fluid, heat transport to estimate desert stream infiltration

In semiarid regions, the contribution of infiltration from intermittent streamflow to ground water recharge may be quantified by comparing simulations of fluid and heat transport beneath stream channels to observed ground temperatures. In addition to quantifying natural recharge, streamflow infiltration estimates provide a means to characterize the physical properties of stream channel sediments and to identify suitable locations for artificial recharge sites. Rates of winter streamflow infiltration along stream channels are estimated based on the cooling effect of infiltrated water on streambed sediments, combined with the simulation of two-dimensional fluid and heat transport using the computer program VS2DH. The cooling effect of ground water is determined by measuring ground temperatures at regular intervals beneath stream channels and nearby channel banks in order to calculate temperature-depth profiles. Additional data inputs included the physical, hydraulic, and thermal properties of unsaturated alluvium, and monthly ground temperatures measurements over an annual cycle. Observed temperatures and simulation results can provide estimates of the minimum threshold for deep infiltration, the variability of infiltration along stream channels, and also the frequency of infiltration events.

Wastewater compounds in urban shallow groundwater wells correspond to exfiltration probabilities of nearby sewers

Wastewater compounds are frequently detected in urban shallow groundwater. Sources include sewage or reclaimed wastewater, but origins are often unknown. In a prior study, wastewater compounds were quantified in waters sampled from shallow groundwater wells in a small coastal California city. Here, we resampled those wells and expanded sample analyses to include sewage- or reclaimed water-specific indicators, i.e. pharmaceutical and personal care product chemicals or disinfection byproducts. Also, we developed a geographic information system (GIS)-based model of sanitary sewer exfiltration probability--combining a published pipe failure model accounting for sewer pipe size, age, materials of construction, with interpolated depths to groundwater--to determine if sewer system attributes relate to wastewater compounds in urban shallow groundwater. Across the wells, groundwater samples contained varying wastewater compounds, including acesulfame, sucralose, bisphenol A, 4-tert-octylphenol, estrone and perfluorobutanesulfonic acid (PFBS). Fecal indicator bacterial concentrations and toxicological bioactivities were less than known benchmarks. However, the reclaimed water in this study was positive for all bioactivity tested. Excluding one well intruded by seawater, the similarity of groundwater to sewage, based on multiple indicators, increased with increasing sanitary sewer exfiltration probability (modeled from infrastructure within ca. 300 m of each well). In the absence of direct exfiltration or defect measurements, sewer exfiltration probabilities modeled from the collection systems physical data can indicate potential locations where urban shallow groundwater is contaminated by sewage.

Fate of Nutrients in Shallow Groundwater Receiving Treated Septage, Malibu, CA

Treated wastewater discharged from more than 400 onsite wastewater treatment systems (OWTS) near the Civic Center area of Malibu, California, 40 km west of downtown Los Angeles, composes 28% of the recharge to a 3.4 km2 alluvial aquifer. On the basis of δ18O and δD data, the fraction of wastewater in some samples was >70%. Ammonium and nitrate concentrations in water from 15 water-table wells sampled in July 2009 and April 2010 ranged from <0.01 to 12 milligrams per liter as nitrogen (mg/L as N), and from <0.01 to 11 mg/L as N, respectively. Chemical and isotopic data (δ15N of ammonium and nitrate, and δ18O of nitrate) show two processes remove nitrogen discharged from OWTS. Where groundwater was reducing, sorption of ammonium resulted in 30 to 50% nitrogen removal. Where groundwater was initially oxic, nitrification with subsequent denitrification as reducing conditions developed, resulted in up to 60% nitrogen removal. Nitrogen removal through sorption dominated during the cooler April sample period, and denitrification dominated during the warmer July sample period. The combination of mixing and nitrogen removal due to denitrification, sorption, and volatilization produces a δ15N apparent fractionation factor (εapp = −5), that can be explained using laboratory-derived fractionation factors (ε) for the individual processes. Phosphate concentrations ranged from < 0.04 to 2 mg/L as phosphorous. Sorption to iron oxides on the surfaces of mineral grains at near-neutral pHs removed some phosphate; however, little removal occurred at more alkaline pHs (>7.3).

Electromagnetic-Induction Logging to Monitor Changing Chloride Concentrations

Water from the San Joaquin Delta, having chloride concentrations up to 3590 mg/L, has intruded fresh water aquifers underlying Stockton, California. Changes in chloride concentrations at depth within these aquifers were evaluated using sequential electromagnetic (EM) induction logs collected during 2004 through 2007 at seven multiple-well sites as deep as 268 m. Sequential EM logging is useful for identifying changes in groundwater quality through polyvinyl chloride-cased wells in intervals not screened by wells. These unscreened intervals represent more than 90% of the aquifer at the sites studied. Sequential EM logging suggested degrading groundwater quality in numerous thin intervals, typically between 1 and 7 m in thickness, especially in the northern part of the study area. Some of these intervals were unscreened by wells, and would not have been identified by traditional groundwater sample collection. Sequential logging also identified intervals with improving water quality-possibly due to groundwater management practices that have limited pumping and promoted artificial recharge. EM resistivity was correlated with chloride concentrations in sampled wells and in water from core material. Natural gamma log data were used to account for the effect of aquifer lithology on EM resistivity. Results of this study show that a sequential EM logging is useful for identifying and monitoring the movement of high-chloride water, having lower salinities and chloride concentrations than sea water, in aquifer intervals not screened by wells, and that increases in chloride in water from wells in the area are consistent with high-chloride water originating from the San Joaquin Delta rather than from the underlying saline aquifer.

Movement of Water Infiltrated from a Recharge Basin to Wells

Local surface water and stormflow were infiltrated intermittently from a 40-ha basin between September 2003 and September 2007 to determine the feasibility of recharging alluvial aquifers pumped for public supply, near Stockton, California. Infiltration of water produced a pressure response that propagated through unconsolidated alluvial-fan deposits to 125 m below land surface (bls) in 5 d and through deeper, more consolidated alluvial deposits to 194 m bls in 25 d, resulting in increased water levels in nearby monitoring wells. The top of the saturated zone near the basin fluctuates seasonally from depths of about 15 to 20 m. Since the start of recharge, water infiltrated from the basin has reached depths as great as 165 m bls. On the basis of sulfur hexafluoride tracer test data, basin water moved downward through the saturated alluvial deposits until reaching more permeable zones about 110 m bls. Once reaching these permeable zones, water moved rapidly to nearby pumping wells at rates as high as 13 m/d. Flow to wells through highly permeable material was confirmed on the basis of flowmeter logging, and simulated numerically using a two-dimensional radial groundwater flow model. Arsenic concentrations increased slightly as a result of recharge from 2 to 6 µg/L immediately below the basin. Although few water-quality issues were identified during sample collection, high groundwater velocities and short travel times to nearby wells may have implications for groundwater management at this and at other sites in heterogeneous alluvial aquifers.

Concentration, UV-spectroscopic characteristics and fractionation of DOC in stormflow from an urban stream, Southern California, USA

Environmental context. There is concern that there may be public health issues associated with dissolved organic carbon (DOC) because of its unknown composition and reactivity to disinfection in water used for public supply. Here, changes in the concentration and composition of DOC in stormflow (1995–2004) in the Santa Ana River and its tributaries is evaluated based on its optical properties, molecular weight and solubility differences of the contributing DOC components. Such a study may allow improved water management in the future. Ongoing work in the study area has shown that DOC in surface water diverted for groundwater recharge degrades to near natural levels in underlying aquifers pumped for public supply. Abstract. The composition of dissolved organic carbon (DOC) in stormflow from urban areas has been greatly altered, both directly and indirectly, by human activities and there is concern that there may be public health issues associated with DOC, which has unknown composition from different sources within urban watersheds. This study evaluated changes in the concentration and composition of DOC in stormflow in the Santa Ana River and its tributaries between 1995 and 2004 using a simplified approach based on the differences in the optical properties of DOC and using operationally defined differences in molecular weight and solubility. The data show changes in the composition of DOC in stormflow during the rainy season and differences associated with runoff from different parts of the basin, including extensive upland areas burned prior to the 2004 rainy season. Samples were collected from the Santa Ana River, which drains ~6950 km2 of the densely populated coastal area of southern California, during 23 stormflows between 1995 and 2004. Dissolved organic carbon (DOC) concentrations during the first stormflows of the ‘winter’ (November to March) rainy season increased rapidly with streamflow and were positively correlated with increased faecal indicator bacteria concentrations. DOC concentrations were not correlated with streamflow or with other constituents during stormflows later in the rainy season and DOC had increasing UV absorbance per unit carbon as the rainy season progressed. DOC concentrations in stormflow from an urban drain tributary to the river also increased during stormflow and were greater than concentrations in the river. DOC concentrations in stormflow from a tributary stream, draining urban and agricultural land that contained more than 320000 animals, mostly dairy cows, were higher than concentrations in stormflow from the river and from the urban drain. Fires that burned large areas of the basin before the 2004 rainy season did not increase DOC concentrations in the river during stormflow after the fires – possibly because the large watershed of the river damped the effect of the fires. However, the fires increased the hydrophobic neutral organic carbon fraction of DOC in stormflow from the urban drain and the tributary stream.

Selected techniques for monitoring water movement through unsaturated alluvium during managed aquifer recharge

Printed on recycled paper Figure 1. Water-level and pumping data from a managed aquifer-recharge site near Victorville, California, 2001–07. The data show water-level changes caused by nearby pumping that can make it difficult to identify when infiltrated water arrives at the water table (Izbicki and others, 2008). Figure 2. Experimental field-scale recharge sites at the Water Supply Stabilization Project-2 site, Antelope Valley, California. The AVUZ-2 and AVUZ-4 test sites include recharge ponds and instrumentation for monitoring the downward movement of water infiltrated from the pond. Managed aquifer recharge is used to augment natural recharge to aquifers. It can be used to replenish aquifers depleted by pumping or to store water during wetter years for withdrawal during drier years. Infiltration from ponds is a commonly used, inexpensive approach for managed aquifer recharge. At some managed aquiferrecharge sites, the time when infiltrated water arrives at the water table is not always clearly shown by water-level data (fig. 1). As part of site characterization and operation, it can be desirable to track downward movement of infiltrated water through the unsaturated zone to identify when it arrives at the water table.

Data from a thick unsaturated zone in Joshua Tree, San Bernardino County, California, 2007--09

..........................................................................................................................................................