Shallow Coastal Aquifer Research Articles

Two and three-dimensional resistivity imaging in a shallow coastal sand aquifer

Surface resistivity imaging was proposed to map the saline interface produced by the interaction of a tidal creek and a shallow fresh water aquifer. Forward modelling was used to determine the relative merits of using two and three (2D and 3D) dimensional electrode arrays. The results of the forward modelling indicate that under ideal circumstances the surface techniques can resolve the conceptual saline interface near the tidal creek. One 3D and four 2D surveys were completed on the bank of Korogoro Creek. The 2D surveys were orientated parallel (3) and perpendicular (1) to the creek. The apparent resistivity data was inverted using RES2DINV and RES3DINV. The resulting 2D images demonstrate that the conceptual model used in the forward modelling oversimplified the distribution of saline water in the subsurface. The 2D images indicate that freshwater is discharging into the creek via preferred pathways at several locations along the creek bank. The 3D image shows that the distribution of saline water is constrained to an area close to the creek. Neither the 2D or 3D images were able to clearly identify the thin saline wedge observed using downhole techniques (EM39). The results of the resistivity surveys indicate that the techniques are useful in identifying relatively large-scale saline patterns but are unable to identify relatively small anomalies. Most importantly the results highlight the critical need to control interpretation of high resolution surface resistivity data with downhole techniques.

Hydrogeochemistry of groundwater in coastal wetlands: implications for coastal conservation in Scotland

Groundwater in a shallow coastal aquifer in north east Scotland was monitored over the hydrological year October 1996–September 1997. Groundwater flow from inland areas sustained freshwater conditions in a dune–wetland complex of conservation importance. In particular, seasonal flooding of the coastal wetlands due to water table rise provided important roosting and breeding habitats for waterfowl. Hydrogeochemical analysis revealed that groundwater in the shallow sand aquifer was circum-neutral, and non-saline, despite being within 50 m of the sea and only 1 m above the mean high water mark. Calcium and HCO3− were the dominant cation and anion respectively, reflecting weathering processes in the aquifer. Use of the geochemical code NETPATH indicated that calcite weathering in shell fragments within the sand was the primary source of Ca and alkalinity generation. The concentrations of Na and Cl were also important, though these can be explained primarily by atmospheric inputs from precipitation. In detail, the spatial and temporal variation in groundwater chemistry was remarkably complex for what intuitively appeared a simple aquifer system. Temporal variations in groundwater chemistry mainly related to the seasonal event of groundwater recharge. Thus, the main period of rising groundwater levels resulted in a marked dilution of solutes in the aquifer, implying that water storage greatly increased in a relatively short period. A period of several weeks appeared to be required for dissolution processes to proceed to equilibrium. Spatial variation in groundwater chemistry appears to relate to the spatial distribution of geochemical processes in different hydrogeological units. Sulphate reduction, alkalinity generation and Fe precipitation appear to be locally important processes. The chemistry of groundwater maintains the wetland habitat by providing freshwater conditions that allow populations of various plant species to flourish. The potentially large recharge catchments of coastal wetlands, together with increasing pressures in the coastal zone, dictate that pollution can threaten the integrity of hydrochemical processes and requires careful monitoring if freshwater wetlands are to maintain their conservation importance.

Modeling and Analysis of Seawater Intrusion in the Coastal Aquifer of Eastern Cap-Bon, Tunisia

A numerical model that treats density-dependent variably saturated flow and miscible salt transport is used to investigate the occurrence of seawater intrusion in the ‘Korba’ aquifer of the eastern coast of Cap-Bon in northern Tunisia. We examine the interplay between pumping regimes and recharge scenarios and its effect on the saline water distribution. More localized simulations are used to examine, in vertical cross sections, the effects of well location and soil type and the role of the vadose zone in possible remediation actions. The exploratory simulations suggest interesting interactions between the unsaturated zone and the saltwater–freshwater interface with possible implications for groundwater exploitation from shallow unconfined coastal aquifers, involving in one case feedback between seawater intrusion and the high pressure head gradients around the pumping-induced drawdown cone and in another case threshold-like interface displacement for tight soils such as clays. The data processing steps undertaken in this GIS and modeling study are described in some detail, and a critical assessment is given of the data availability and of the requirements for successful monitoring and modeling of seawater intrusion risks in heavily exploited coastal aquifers such as those found in the semi-arid regions of the Mediterranean basin. It is shown how, with the aid of GIS, reasonably reliable information can be assembled from maps, surveys, and other sources of geospatial and hydrogeological data, an approach that is necessary in the many regions of the world with acute water resource problems but with limited means for undertaking systematic data acquisition and environmental monitoring actions. Nonetheless the need for more concerted monitoring of relevant parameters and processes and of closer coordination between monitoring and modeling is stressed. An idea of the extent of over-exploitation of the Korba aquifer is obtained by examining the pumping and rainfall/infiltration data, and the simulation results support groundwater pumping as the mechanism for and seawater intrusion as the origin of the salt contamination observed in the soils and subsurface waters of the Korba plain.

Modelling the potential impact of climate change on a shallow coastal aquifer in northern Scotland

Abstract The potential impact of climate change on a freshwater wetland in a coastal dune system at St. Fergus, north-east Scotland, was assessed using hydrogeological modelling techniques. A range of climate change scenarios were input into a numerical model (MODFLOW) that had been calibrated for the site using contemporary field data. The combination of parameter and hydrogeological uncertainty in models, together with uncertainty over likely site-specific climate change in the future, dictates that such modelling studies are often inconclusive. Nevertheless, a range of scenarios was forecast which indicates that future average recharge and flooding of the wetland is likely to fall within the range of interannual variation observed at the site, regardless of whether recharge increases or decreases. It is therefore likely that any resulting ecological impacts will be limited and reflect subtle hydrological changes that will affect different species in different ways. The utility of such a modelling approach is in providing environmental managers with a range of likely scenarios that can be incorporated into future management plans and policies to maximize flexibility of response.

The Distribution of Microbial Communities in Anaerobic and Aerobic Zones of a Shallow Coastal Plain Aquifer.

Randomly amplified polymorphic DNA (RAPD) fingerprinting was used to determine the genetic similarity of whole-community DNA extracts from unattached microorganisms in several groundwater wells. The study site was a shallow coastal plain aquifer on the Eastern Shore of Virginia that contains distinct regions of anaerobic and aerobic groundwater. Several wells in each region were sampled, and principal component and cluster analyses showed a clear separation of the microbial communities from the two chemical zones of the aquifer. Within these zones, there was no relationship between the genetic relatedness of a pair of communities and their spatial separation. Two additional sets of samples were taken at later times, and the same clear separation between communities in the different zones of the aquifer was observed. The specific relationships between wells within each zone changed over time, however, and the magnitude and direction of these changes corresponded to concurrent changes in the groundwater chemistry at each well. Together, these results suggest that local variation in groundwater chemistry can support genetically distinct microbial communities, and that the composition of the microbial communities can follow seasonal fluctuations in groundwater chemistry.

Impact of Sea Level Rise on Seawater Intrusion into Coastal Aquifer

The anticipated climate change is likely to cause a sea level rise of about 50 cm by the year 2100, which may increase the extent of seawater intrusion into aquifers. This paper identifies the different scenarios of sea level rise and presents, using a numerical simulation model, a scientific and generic evaluation of the impact of sea level rise due to climate change and storm surge on a shallow coastal aquifer in Jamaica. The results of the simulation indicate that, for a confined alluvial aquifer comprising fine–medium sands and having a normal recharge rate, the sea level rise would not cause any significant landward movement of the dispersive seawater interface up to the year 2050. Moreover, for such an aquifer, a storm surge would also have no significant adverse impact. However, in both these cases, the concentration fields within the interface would undergo considerable change. It is hoped that this evaluation would help in decision making for the design of new ground-water based water supply systems in the small tropical islands.

Modelling uncertainty in estimates of recharge to a shallow coastal aquifer

Due to the difficulty of field characterization, one is often forced to employ hydrological models in order to determine estimates of aquifer recharge. A Geographical Information System (GIS) and Digital Elevation Model (DEM) may provide the basic support for such hydrological models. Unfortunately, despite the availability of these tools, the values of many parameters within distributed models remain uncertain for application to real world problems. Using the Bisagno basin (100 km2) in the Liguria region of Italy, the role of model parameter uncertainty on estimates of recharge to the Bisagno Aquifer is demonstrated. Uncertainty estimation is achieved through Monte Carlo simulations with the aid of the Generalised Likelihood Uncertainty Estimation procedure of Beven & Binley (1992).

A Comparison of Methods to Determine K in a Shallow Coastal Aquifer

AbstractFive methods (tidal damping, tracer, slug, permeameter, and grain‐size analysis) for the determination of hydraulic conductivity (K) were compared in both upland and shoreline regions of a sandy outwash aquifer adjacent to a shallow coastal embayment. The 7 m tracer test and tidal damping methods appeared to determine in situ K over sufficient horizontal scales to integrate the spatial variability of K within the aquifer, with the tracer test giving the most reproducible K estimates for the upland, 109 m/d. The tidal K measurements were similar to the tracer results, 106 m/d, integrating K over greater distances (30 to 85 m), but were highly variable due to the large aquifer storativity, small tidal amplitude, and irregular tides.The smaller scale methods yielded lower upland K estimates with the lowest value from permeameters, 29 m/d, only about one‐third of the other four methods. Estimates of upland K by both grain‐size and slug tests were comparable, 86 m/d and 71 m/d, respectively. However, the K derived from grain‐size methods was dependent not only upon the sediment but on the method selected, with consistently higher values of K generated by the Hazen equation than the Krumbein and Monk equation. Verification by in situ tests may be useful in removing this bias.Significant differences in K were found between shoreline and upland regions, though the relative magnitude of the difference was not constant: among the four methods where comparative measurements were possible, shoreline K ranged from 42% to 75% of upland values. Slug tests yielded the largest differences, 30 m/d and 71 m/d, with permeameter, 18 m/d and 24 m/d, the lowest. The lower K of shoreline sediments appears to be the result of processes associated with coastal embayments and should be addressed in the measurement and modeling of coastal ground‐water discharge.

Nitrate Leaching to a Shallow Mid-Atlantic Coastal Plain Aquifer as Influenced by Conventional No-Till and Low-Input Sustainable Grain Production Systems

In December, 1987, the states in the Chesapeake Bay region, along with the federal government, signed an agreement which called for a 40% reduction in nitrogen and phosphorus loadings to the Bay by the year 2000. To accomplish this goal, major reductions in nutrient loadings associated with agricultural management practices were deemed necessary. The objective of this study was to determine if reducing fertilizer inputs to the NT system would result in a reduction in nitrogen contamination of groundwater. In this study, groundwater, soil, and percolate samples were collected from two cropping systems. The first system was a conventional no-till (NT) grain production system with a two-year rotation of corn/winter wheat/double crop soybean. The second system, denoted low-input sustainable agriculture (LISA), produced the same crops using a winter legume and relay-cropped soybeans into standing wheat to reduce nitrogen and herbicide inputs. Nitrate-nitrogen concentrations in groundwater were significantly lower under the LISA system. Over 80% of the NT groundwater samples had NO3-N concentrations greater than 10 mgl-1, compared to only 4% for the LISA cropping system. Significantly lower soil mineral N to a depth of 180 cm was also observed. The NT soil had nearly twice as much mineral N present in the 90-180 cm portion than the LISA cropping system.

Identification and catabolic activity of well-derived gasoline-degrading bacteria from a contaminated aquifer

Approximately 300 gasoline-degrading bacteria were isolated from well water and core material from a shallow coastal aquifer contaminated with unleaded gasoline. Identification of 244 isolates revealed four genera: Pseudomonas, Alcaligenes, Nocardia, and Micrococcus, with pseudomonads making up 86.9% of bacteria identified. A total of 297 isolates was sorted into 111 catabolic groups on the basis of aerobic growth responses on 15 gasoline hydrocarbons. Each test hydrocarbon was degraded by at least one isolate. Toluene, p-xylene, ethylbenzene, and 1,2,4-trimethylbenzene were most frequently utilized as growth substrates, whereas cyclic and branched alkanes were least utilized. Most isolates were able to grow on 2 or 3 different hydrocarbons, and nearly 75% utilized toluene as a sole source of carbon and energy. Isolates were remarkably specific for hydrocarbon usage, often catabolizing only one of several closely related compounds. A subset of 220 isolates was sorted into 51 groups by polyacrylamide gel electrophoresis. Pseudomonas aeruginosa was partitioned into 16 protein-banding groups (i.e., subspecies) whose catabolic activities were largely restricted to substituted aromatics. Different members of subspecies groups defined by protein-banding pattern analysis often exhibited different growth responses on the same hydrocarbon, implying marked strain diversity. The catabolic activities of well-derived, gasoline-degrading bacteria associated with this contaminated aquifer are consonant with in situ adaptation at the site.

The Rainfall/Water-Table Level Relationship of an Unconfined Aquifer.

The soil parameters that determine the infiltration rates of a shallow unconfined coastal aquifer due to rainfall were obtained using the recursive least squares method. The parameters were predicted by minimizing the mean square error between the estimated and measured water-table levels in a bore hole. The parameters λt , a drainage factor, βt , a parameter which accounts for external disturbances, and αt , a set of parameters which determine the infiltration rates due to rainfall, were all treated as time-dependent variables. Their variation is traced by introducing a "forgetting factor" in the algorithm used to determine the parameters. The parameters were determined using rainfall and water-table levels for the year 1979 at a bore hole. The parameters βt and αt varied considerably over the period of time and λt remained constant. From the parameters obtained, it appears that most of the recharge due to rainfall takes place within the first two days of the event. After two days there is very little variation in the water-table level due to rainfall although there is a slight drop on the third, fourth and fifth day, presumably due to the escape of entrapped air beneath the water-table level.

A single‐potential solution for regional interface problems in coastal aquifers

An analytic technique for solving three‐dimensional interface problems in coastal aquifers is presented in this paper. Restriction is made to cases of steady state flow with homogeneous isotropic permeability where the vertical flow rates can be neglected in relation to the horizontal ones (the Dupuit‐Forchheimer assumption). The aquifer is divided into zones defined by the type of flow occurring. These types of flow may be either confined, unconfined, confined interface, or unconfined interface flow, where the interfaces separate freshwater from salt water at rest. The technique is based upon the use of a single potential which is defined throughout all zones of the aquifer. This potential in each zone can be represented in a way similar to that suggested by Girinskii in 1946 and 1947. The potential introduced in this paper is single valued and continuous throughout the multiple‐zone aquifer, and its application does not require that the boundaries between the zones be known in advance. The technique thus avoids the difficulties that result from the discontinuity of both the velocity gradients and the Girinskii potentials at the boundaries between the zones and from the unknown locations of these boundaries. The use of the single‐valued potential is illustrated in this paper for an analytic technique, but it may be used with some advantage in numerical methods such as finite difference or finite element techniques. Applications discussed in this paper involve two interface flow problems in a shallow coastal aquifer with a fully penetrating well. The first problem is one of unconfined interface flow where the upper boundary is a free water table. The second is one of confined interface flow where the upper boundary is horizontal and impervious. Each problem involves two zones. One zone is adjacent to the coast and is bounded below by an interface between freshwater and salt water at rest. The other zone is bounded below by an impervious bottom. It is shown that saltwater intrusion in the well occurs when the discharge of the well surpasses a certain value for which the interface becomes unstable. The conditions that must be met to prevent such saltwater intrusion are established for each problem and are represented graphically.