Search USGS⌕ Search

SEARCH · Search USGS

Results for “Ground Water Modeling Investigation Report”

Search indexed USGS publications on groundwater, aquifers, geologic maps, mineral resources and earthquakes. Explore source records by subject and place.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 775 records · Page 43Linked to original sources
USGS1987

AQMAN; linear and quadratic programming matrix generator using two-dimensional ground-water flow simulation for aquifer management modeling

A FORTRAN-77 computer program code that helps solve a variety of aquifer management problems involving the control of groundwater hydraulics. It is intended for use with any standard mathematical programming package that uses Mathematical Programming System input format. The computer program creates the input files to be used by the optimization program. These files contain all the hydrologic information and management objectives needed to solve the management problem. Used in conjunction with a mathematical programming code, the computer program identifies the pumping or recharge strategy that achieves a user 's management objective while maintaining groundwater hydraulic conditions within desired limits. The objective may be linear or quadratic, and may involve the minimization of pumping and recharge rates or of variable pumping costs. The problem may contain constraints on groundwater heads, gradients, and velocities for a complex, transient hydrologic system. Linear superposition of solutions to the transient, two-dimensional groundwater flow equation is used by the computer program in conjunction with the response matrix optimization method. A unit stress is applied at each decision well and transient responses at all control locations are computed using a modified version of the U.S. Geological Survey two dimensional aquifer simulation model. The program also computes discounted cost coefficients for the objective function and accounts for transient aquifer conditions. (Author 's abstract)

Water-Resources Investigations Report↗
USGS1988

Alternative sources of large seasonal ground-water supplies in the headwaters of the Susquehanna River basin, New York

The northern divide of the Susquehanna River basin crosses 29 broad valleys that contain thick glacial deposits but are drained only by small headwater streams. Much groundwater could be withdrawn from sand and gravel deposits in these valleys with little immediate effect on streamflow. A digital model of the headwater reach of one typical valley suggests that pumping 10.8 million gal/day for 2 months every summer would lower the water table as much as 33 ft, cause the upper 1,900 ft of the stream draining the valley to go dry, and reduce streamflow downvalley by 1.2 million gal/day by the time pumping ceased. Saturated thickness of surficial sand and gravel exceeds 40 ft in about half the headwater valley reaches; the valley floor areas range from 0.2 to 9 sq mi. Seepage losses from small streams that carry runoff from adjacent till-covered uplands are a major source of recharge to aquifers in these valleys under natural conditions and would increase if the water table were lowered by seasonal withdrawals. Some aquifers beneath extensive clay layers in these and other valleys of the Susquehanna River basin may be partially independent of streams but not easily evaluated. (USGS)

Water-Resources Investigations Report↗
USGS1993

Geohydrology and simulation of ground-water flow in the Red Clay Creek Basin, Chester County, Pennsylvania, and New Castle County, Delaware

The 54-square-mile Red Clay Creek Basin, located in the lower Delaware River Basin, is underlain primarily by metamorphic rocks that range from Precambrian to Lower Paleozoic in age. Ground water flows through secondary openings in fractured crystalline rock and through primary openings below the water table in the overlying saprolite. Secondary porosity and permeability vary with hydrogeologic unit, topographic setting, and depth. Thirty-nine percent of the water-bearing zones are encountered within 100 feet of the land surface, and 79 percent are within 200 feet. The fractured crystalline rock and overlying saprolite act as a single aquifer under unconfined conditions. The water table is a subdued replica of the land surface. Local ground-water flow systems predominate in the basin, and natural ground-water discharge is to streams, comprising 62 to 71 percent of streamflow. Water budgets for 1988-90 for the 45-square-mile effective drainage area above the Woodale, Del., streamflow-measurement station show that annual precipitation ranged from 43.59 to 59.14 inches and averaged 49.81 inches, annual streamflow ranged from 15.35 to 26.33 inches and averaged 20.24 inches, and annual evapotranspiration ranged from 27.87 to 30.43 inches and averaged 28.98 inches. The crystalline rocks of the Red Clay Creek Basin were simulated two-dimensionally as a single aquifer under unconfined conditions. The model was calibrated for short-term steady-state conditions on November 2, 1990. Recharge was 8.32 inches per year. Values of aquifer hydraulic conductivity in hillside topographic settings ranged from 0.07 to 2.60 feet per day. Values of streambed hydraulic conductivity ranged from 0.08 to 26.0 feet per day. Prior to simulations where ground-water development was increased, the calibrated steady-state model was modified to approximate long-term average conditions in the basin. Base flow of 11.98 inches per year and a ground-water evapotranspiration rate of 2.17 inches per year were simulated by the model. Different combinations of ground-water supply and wastewater-disposal plans were simulated to assess their effects on the stream-aquifer system. Six of the simulations represent an increase in population of 14,283 and water use of 1.07 million gallons per day. One simulation represents an increase in population of 28,566 and water use of 2.14 million gallons per day. Reduction of average base flow is greatest for development plans with wastewater removed from the basin through sewers and is proportional to the amount of water removed from the basin. The development plan that had the least effect on water levels and base flow included on-lot wells and on-lot septic systems. Five organochlorine insecticides--lindane, DDT, dieldrin, heptachlor, and methoxychlor--were detected in ground water. Four organophosphorus insecticides--malathion, parathion, diazinon, and phorate--were detected in ground water. Four volatile organic compounds--benzene, toluene, tetrachloroethylene, and trichloroethylene--were detected in ground water. Phenol was detected at concentrations up to 8 micrograms per liter in water from 50 percent of 14 wells sampled. The concentration of dissolved nitrate in water from 18 percent of wells sampled exceeded 10 milligrams per liter as nitrogen; concentration of nitrate were as high as 19 milligrams per liter. PCB was detected in the bottom material of West Branch Red Clay Creek at Kennet Square at concentrations up to 5,600 micrograms per kilogram.

Delaware, Pennsylvania↗
USGS2003

Simulated ground-water flow in the Ogallala and Arikaree aquifers, Rosebud Indian Reservation area, South Dakota

The Ogallala and Arikaree aquifers are important water resources in the Rosebud Indian Reservation area and are used extensively for irrigation, municipal, and domestic water supplies. Continued or increased withdrawals from the Ogallala and Arikaree aquifers in the Rosebud Indian Reservation area have the potential to affect water levels in these aquifers. This report describes a conceptual model of ground-water flow in these aquifers and documents the development and calibration of a numerical model to simulate ground-water flow. Data for a twenty-year period (water years 1979 through 1998) were analyzed for the conceptual model and included in steady-state and transient numerical simulations of ground-water flow for the same 20-year period. A three-dimensional ground-water flow model, with two layers, was used to simulate ground-water flow in the Ogallala and Arikaree aquifers. The upper layer represented the Ogallala aquifer, and the lower layer represented the Arikaree aquifer. The study area was divided into grid blocks 1,640 feet (500 meters) on a side, with 153 rows and 180 columns. Areal recharge to the Ogallala and Arikaree aquifers occurs from precipitation on the outcrop areas. The recharge rate for the steady-state simulation was 3.3 inches per year for the Ogallala aquifer and 1.7 inches per year for the Arikaree aquifer for a total recharge rate of 266 cubic feet per second. Discharge from the Ogallala and Arikaree aquifers occurs through evapotranspiration, discharge to streams, and well withdrawals. Discharge rates in cubic feet per second for the steady-state simulation were 184 for evapotranspiration, 46.8 and 19.7 for base flow to the Little White and Keya Paha Rivers, respectively, and 11.6 for well withdrawals from irrigation use. Estimated horizontal hydraulic conductivity used for the numerical model ranged from 0.2 to 120 feet per day in the Ogallala aquifer and 0.1 to 5.4 feet per day in the Arikaree aquifer. A uniform vertical hydraulic conductivity value of 6.6x10 -4 feet per day was applied to the Ogallala aquifer. Vertical hydraulic conductivity was estimated for five zones in the Arikaree aquifer and ranged from 8.6x10 -6 to 7.2x10 -1 feet per day. Average rates of recharge, maximum evapotranspiration, and well withdrawals were included in the steady-state simulation, whereas the time-varying rates were included in the transient simulation. Model calibration was accomplished by varying parameters within plausible ranges to produce the best fit between simulated and observed hydraulic heads and base-flow discharges from the Ogallala and Arikaree aquifers. For the steady-state simulation, the root mean square error for simulated hydraulic heads for all wells was 26.8 feet. Simulated hydraulic heads were within ±50 feet of observed values for 95 percent of the wells. For the transient simulation, the difference between the simulated and observed means for hydrographs was within ±40 feet for all observation wells. The potentiometric surfaces of the two aquifers calculated by the steady-state simulation established initial conditions for the transient simulation. A sensitivity analysis was used to examine the response of the calibrated steady-state model to changes in model parameters including horizontal and vertical hydraulic conductivity, evapotranspiration, recharge, and riverbed conductance. The model was most sensitive to recharge and horizontal hydraulic conductivity.

South Dakota↗
USGS2009

Development and Implementation of a Transport Method for the Transport and Reaction Simulation Engine (TaRSE) based on the Godunov-Mixed Finite Element Method

A model to simulate transport of materials in surface water and ground water has been developed to numerically approximate solutions to the advection-dispersion equation. This model, known as the Transport and Reaction Simulation Engine (TaRSE), uses an algorithm that incorporates a time-splitting technique where the advective part of the equation is solved separately from the dispersive part. An explicit finite-volume Godunov method is used to approximate the advective part, while a mixed-finite element technique is used to approximate the dispersive part. The dispersive part uses an implicit discretization, which allows it to run stably with a larger time step than the explicit advective step. The potential exists to develop algorithms that run several advective steps, and then one dispersive step that encompasses the time interval of the advective steps. Because the dispersive step is computationally most expensive, schemes can be implemented that are more computationally efficient than non-time-split algorithms. This technique enables scientists to solve problems with high grid Peclet numbers, such as transport problems with sharp solute fronts, without spurious oscillations in the numerical approximation to the solution and with virtually no artificial diffusion.

Scientific Investigations Report↗
USGS2004

A method for simulating transient ground-water recharge in deep water-table settings in central Florida by using a simple water-balance/transfer-function model

A relatively simple method is needed that provides estimates of transient ground-water recharge in deep water-table settings that can be incorporated into other hydrologic models. Deep water-table settings are areas where the water table is below the reach of plant roots and virtually all water that is not lost to surface runoff, evaporation at land surface, or evapotranspiration in the root zone eventually becomes ground-water recharge. Areas in central Florida with a deep water table generally are high recharge areas; consequently, simulation of recharge in these areas is of particular interest to water-resource managers. Yet the complexities of meteorological variations and unsaturated flow processes make it difficult to estimate short-term recharge rates, thereby confounding calibration and predictive use of transient hydrologic models. A simple water-balance/transfer-function (WBTF) model was developed for simulating transient ground-water recharge in deep water-table settings. The WBTF model represents a one-dimensional column from the top of the vegetative canopy to the water table and consists of two components: (1) a water-balance module that simulates the water storage capacity of the vegetative canopy and root zone; and (2) a transfer-function module that simulates the traveltime of water as it percolates from the bottom of the root zone to the water table. Data requirements include two time series for the period of interest?precipitation (or precipitation minus surface runoff, if surface runoff is not negligible) and evapotranspiration?and values for five parameters that represent water storage capacity or soil-drainage characteristics. A limiting assumption of the WBTF model is that the percolation of water below the root zone is a linear process. That is, percolating water is assumed to have the same traveltime characteristics, experiencing the same delay and attenuation, as it moves through the unsaturated zone. This assumption is more accurate if the moisture content, and consequently the unsaturated hydraulic conductivity, below the root zone does not vary substantially with time. Results of the WBTF model were compared to those of the U.S. Geological Survey variably saturated flow model, VS2DT, and to field-based estimates of recharge to demonstrate the applicability of the WBTF model for a range of conditions relevant to deep water-table settings in central Florida. The WBTF model reproduced independently obtained estimates of recharge reasonably well for different soil types and water-table depths.

Scientific Investigations Report↗
USGS1995

Analysis of ground-water flow in the Catahoula aquifer system in the vicinity of Laurel and Hattiesburg, Mississippi

The upper, middle, and lower Catahoula aquifers in the vicinity of the cites of Laurel and Hattiesburg in southern Mississippi are made up of irregular, discontinuous sand zones in the Catahoula Formation of Miocene age. In places thee three aquifers may be hydraulically well connected, and are referred to as the Catahoula aquifer system. Withdrawal from the Catahoula aquifers increased from 28 million gallons per day (Mgal/d) to 41 Mgal/d during 1970 to 1985, and decreased to 38 Mgal/d during 1990. Most withdrawal in the Laurel area is from the lower and middle Catahoula, and most withdrawal in the Hattiesburg area is from the middle and upper Catahoula aquifers. In the Laurel area, water levels in selected wells in the lower Catahoula aquifer declined at rates ranging from about 1 to 3.6 feet/ year until the late 1980's in response to the increase in pumping. A three-dimensional model was developed to represent ground-water flow in the Catahoula aquifers. Simulated water levels in the lower Catahoula aquifer, the layer most affected by pumping, were lowered from predevelopment levels as much as 130 feet in the Laurel area and 100 feet in the Hattiesburg area, according to the model analysis of 1992 conditions. Three scenarios of increased pumpage, for the period 1992-2020, were simulated. Under the low-growth scenario, water- level declines would be 20 feet or less below 1992 water levels in the middle and upper Catahoula aquifer in the Hattiesburg area, and about 60 feet in the lower Catahoula aquifer in the Laurel area. Under the moderate-growth scenario, water-level declines would be 40 feet or less below 1992 water levels in the middle Catahoula aquifer in the Hattiesburg area. Water-level declines would be about 110 feet in the lower Catahoula aquifer in the Laurel area, and water levels would approach the top of the aquifer. Under the high-growth scenario, water-level declines would be 40 feet or less in the upper Catahoula aquifer and about 80 feet in the middle Catahoula, with the largest declines occurring in the Hattiesburg area. Water levels would decline about 130 feet and would be drawn down below the top of the lower Catahoula aquifer in the Laurel area under the high-growth scenario.

Water-Resources Investigations Report↗
USGS2006

Deep resistivity structure of Yucca Flat, Nevada Test Site, Nevada

The Department of Energy (DOE) and the National Nuclear Security Administration (NNSA) at their Nevada Site Office are addressing groundwater contamination resulting from historical underground nuclear testing through the Environmental Management program and, in particular, the Underground Test Area project. One issue of concern is the nature of the somewhat poorly constrained pre Tertiary geology and its effects on ground-water flow in the area adjacent to a nuclear test. Ground water modelers would like to know more about the hydrostratigraphy and geologic structure to support a hydrostratigraphic framework model that is under development for the Yucca Flat Corrective Action Unit (CAU). During 2003, the U.S. Geological Survey, supported by the DOE and NNSA-NSO, collected and processed data from 51 magnetotelluric (MT) and audio-magnetotelluric (AMT) stations at the Nevada Test Site in and near Yucca Flat to assist in characterizing the pre-Tertiary geology in that area. The primary purpose was to refine the character, thickness, and lateral extent of pre Tertiary confining units. In particular, a major goal has been to define the upper clastic confining unit (late Devonian - Mississippian-age siliciclastic rocks assigned to the Eleana Formation and Chainman Shale) in the Yucca Flat area. The MT and AMT data have been released in separate USGS Open File Reports. The Nevada Test Site magnetotelluric data interpretation presented in this report includes the results of detailed two-dimensional (2 D) resistivity modeling for each profile (including alternative interpretations) and gross inferences on the three dimensional (3 D) character of the geology beneath each station. The character, thickness, and lateral extent of the Chainman Shale and Eleana Formation that comprise the Upper Clastic Confining Unit are generally well determined in the upper 5 km. Inferences can be made regarding the presence of the Lower Clastic Confining Unit at depths below 5 km. Large fault structures such as the CP Thrust fault, the Carpetbag fault, and the Yucca fault that cross Yucca Flat are also discernable as are other smaller faults. The subsurface electrical resistivity distribution and inferred geologic structures determined by this investigation should help constrain the hydrostratigraphic framework model that is under development.

Open-File Report↗
USGS2005

Simulation of ground-water flow to assess geohydrologic factors and their effect on source-water areas for bedrock wells in Connecticut

Generic ground-water-flow simulation models show that geohydrologic factors?fracture types, fracture geometry, and surficial materials?affect the size, shape, and location of source-water areas for bedrock wells. In this study, conducted by the U.S. Geological Survey in cooperation with the Connecticut Department of Public Health, ground-water flow was simulated to bedrock wells in three settings?on hilltops and hillsides with no surficial aquifer, in a narrow valley with a surficial aquifer, and in a broad valley with a surficial aquifer?to show how different combinations of geohydrologic factors in different topographic settings affect the dimensions and locations of source-water areas in Connecticut. Three principal types of fractures are present in bedrock in Connecticut?(1) Layer-parallel fractures, which developed as partings along bedding in sedimentary rock and compositional layering or foliation in metamorphic rock (dips of these fractures can be gentle or steep); (2) unroofing joints, which developed as strain-release fractures parallel to the land surface as overlying rock was removed by erosion through geologic time; and (3) cross fractures and joints, which developed as a result of tectonically generated stresses that produced typically near-vertical or steeply dipping fractures. Fracture geometry is defined primarily by the presence or absence of layering in the rock unit, and, if layered, by the angle of dip in the layering. Where layered rocks dip steeply, layer-parallel fracturing generally is dominant; unroofing joints also are typically well developed. Where layered rocks dip gently, layer-parallel fracturing also is dominant, and connections among these fractures are provided only by the cross fractures. In gently dipping rocks, unroofing joints generally do not form as a separate fracture set; instead, strain release from unroofing has occurred along gently dipping layer-parallel fractures, enhancing their aperture. In nonlayered and variably layered rocks, layer-parallel fracturing is absent or poorly developed; fracturing is dominated by well-developed subhorizontal unroofing joints and steeply dipping, tectonically generated fractures and (or) cooling joints. Cross fractures (or cooling joints) in nonlayered and variably layered rocks have more random orientations than in layered rocks. Overall, nonlayered or variably layered rocks do not have a strongly developed fracture direction. Generic ground-water-flow simulation models showed that fracture geometry and other geohydrologic factors affect the dimensions and locations of source-water areas for bedrock wells. In general, source-water areas to wells reflect the direction of ground-water flow, which mimics the land-surface topography. Source-water areas to wells in a hilltop setting were not affected greatly by simulated fracture zones, except for an extensive vertical fracture zone. Source-water areas to wells in a hillside setting were not affected greatly by simulated fracture zones, except for the combination of a subhorizontal fracture zone and low bedrock vertical hydraulic conductivity, as might be the case where an extensive subhorizontal fracture zone is not connected or is poorly connected to the surface through vertical fractures. Source-water areas to wells in a narrow valley setting reflect complex ground-water-flow paths. The typical flow path originates in the uplands and passes through either till or bedrock into the surficial aquifer, although only a small area of the surficial aquifer actually contributes water to the well. Source-water areas in uplands can include substantial areas on both sides of a river. Source-water areas for wells in this setting are affected mainly by the rate of ground-water recharge and by the degree of anisotropy. Source-water areas to wells in a broad valley setting (bedrock with a low angle of dip) are affected greatly by fracture properties. The effect of a given fracture is to channel the

Scientific Investigations Report↗
USGS1993

Hydrogeology of, and simulated ground-water flow in, the valley-fill aquifers of the upper Rockaway River basin, Morris County, New Jersey

Public water supply in the Rockaway River valley depends almost entirely on ground water from wells in the valley-fill deposits. Ground-water withdrawals from these deposits in 1986 were about 9.1 million gallons per day. A steady-state ground-water flow model was developed to quantify the effects of ground-water withdrawals on water levels in the valley-fill aquifers and on ground-water discharge to the Rockaway River. The ground-water-flow model, which represents an aquifer system consisting of an unconfined and a confined aquifer separated by a discontinuous confining unit, was implemented to examine aquifer resonse to current and predicted ground-water withdrawals in areas of proposed well sites and the effect of increased ground-water withdrawals on ground-water discharge to the river. Ground-water flow to wells in the valley-fill aquifers is sustained by increased vertical flow between the two aquifers, the diversion of ground water that had discharged to the Rockaway River, and induced seepage resulting from pumping near the Rickaway River. If the rate of ground-water recharge decreases or if the rate of ground-water withdrawals from the valley-fill aquifers increases, ground-water discharge to the Rockaway River above the Boonton Reservoir will decrease by an equivalent amount. The average annual base flow of the Rockaway River above the Boonton Reservoir will meet the minimum passing flow requirement of 7 million gallons per day under conditions of average annual ground-water recharge, and increased ground-water withdrawals anticipated by the years 2000 and 2040. For anticipated increases in withdrawals to 11.5 million gallons per day by the the year 2000, and 14.6 million gallons per day by the year 2040, base flow to the Rockaway River above the Boonton Reservoir may not be sufficient to meet the minimum required reservoir outflow during extended periods of decreased recharge, such as drought, as much as 5 and 11 percent of the time, respectively.

Water-Resources Investigations Report↗
USGS1984

Projected water-level declines in the Ogallala aquifer in Lea County, New Mexico

A two-dimensional digital ground-water flow model was constructed of the Ogallala aquifer in Lea County, New Mexico. Simulations of predevelopment steady-state and historical pumping conditions were used to fit the model. Projections of water-level declines were made based on the condition of no additional development and the condition of a 0.44 percent annual increase in irrigation withdrawals to 1990 and a 0.88 percent annual increase in nonirrigation withdrawals to 2020. Based on no additional development, projected maximum water-level declines from 1980 were 31 feet in 2000 and 59 feet in 2020. The amount of recoverable water remaining in the aquifer as simulated by the model was 24.6 million acre-feet in 2000 and 22.4 million acre-feet in 2020, compared to an estimated 28 million acre-feet in 1980, with increased withdrawals, projected maximum declines were 33 feet in 2000 and 67 feet in 2020. As simulated by the model, 24.4 million acre-feet of recoverable water remained in 2000 and 21.8 million acre-feet remained in 2020. The sensitivity of the model to variation in hydraulic conductivity, specific yield, and recharge was tested over their range of uncertainty. Projected water-level declines were most sensitive to specific yield. In response to changes in specific yield, declines varied by as much as 6.2 feet from the standard simulation to 2020 with no additional development. (USGS)

Water-Resources Investigations Report↗
USGS2000

Estimating the probability of elevated nitrate (NO2+NO3-N) concentrations in ground water in the Columbia Basin Ground Water Management Area, Washington

Logistic regression was used to relate anthropogenic (man-made) and natural factors to the occurrence of elevated concentrations of nitrite plus nitrate as nitrogen in ground water in the Columbia Basin Ground Water Management Area, eastern Washington. Variables that were analyzed included well depth, depth of well casing, ground-water recharge rates, presence of canals, fertilizer application amounts, soils, surficial geology, and land-use types. The variables that best explain the occurrence of nitrate concentrations above 3 milligrams per liter in wells were the amount of fertilizer applied annually within a 2-kilometer radius of a well and the depth of the well casing; the variables that best explain the occurrence of nitrate above 10 milligrams per liter included the amount of fertilizer applied annually within a 3-kilometer radius of a well, the depth of the well casing, and the mean soil hydrologic group, which is a measure of soil infiltration rate. Based on the relations between these variables and elevated nitrate concentrations, models were developed using logistic regression that predict the probability that ground water will exceed a nitrate concentration of either 3 milligrams per liter or 10 milligrams per liter. Maps were produced that illustrate the predicted probability that ground-water nitrate concentrations will exceed 3 milligrams per liter or 10 milligrams per liter for wells cased to 78 feet below land surface (median casing depth) and the predicted depth to which wells would need to be cased in order to have an 80-percent probability of drawing water with a nitrate concentration below either 3 milligrams per liter or 10 milligrams per liter. Maps showing the predicted probability for the occurrence of elevated nitrate concentrations indicate that the irrigated agricultural regions are most at risk. The predicted depths to which wells need to be cased in order to have an 80-percent chance of obtaining low nitrate ground water exceed 600 feet in the irrigated agricultural regions, whereas wells in dryland agricultural areas generally need a casing in excess of 400 feet. The predicted depth to which wells need to be cased to have at least an 80-percent chance to draw water with a nitrate concentration less than 10 milligrams per liter generally did not exceed 800 feet, with a 200-foot casing depth typical of the majority of the area.

Washington↗
USGS2005

Ground-water vulnerability to nitrate contamination at multiple thresholds in the mid-Atlantic region using spatial probability models

The U.S. Geological Survey, in cooperation with the U.S. Environmental Protection Agency?s Regional Vulnerability Assessment Program, has developed a set of statistical tools to support regional-scale, ground-water quality and vulnerability assessments. The Regional Vulnerability Assessment Program?s goals are to develop and demonstrate approaches to comprehensive, regional-scale assessments that effectively inform managers and decision-makers as to the magnitude, extent, distribution, and uncertainty of current and anticipated environmental risks. The U.S. Geological Survey is developing and exploring the use of statistical probability models to characterize the relation between ground-water quality and geographic factors in the Mid-Atlantic Region. Available water-quality data obtained from U.S. Geological Survey National Water-Quality Assessment Program studies conducted in the Mid-Atlantic Region were used in association with geographic data (land cover, geology, soils, and others) to develop logistic-regression equations that use explanatory variables to predict the presence of a selected water-quality parameter exceeding a specified management concentration threshold. The resulting logistic-regression equations were transformed to determine the probability, P(X), of a water-quality parameter exceeding a specified management threshold. Additional statistical procedures modified by the U.S. Geological Survey were used to compare the observed values to model-predicted values at each sample point. In addition, procedures to evaluate the confidence of the model predictions and estimate the uncertainty of the probability value were developed and applied. The resulting logistic-regression models were applied to the Mid-Atlantic Region to predict the spatial probability of nitrate concentrations exceeding specified management thresholds. These thresholds are usually set or established by regulators or managers at National or local levels. At management thresholds of 1 milligram per liter and 3 milligrams per liter as nitrogen, the probability of nitrate concentrations exceeding these levels is greater than 50 percent (0.50) throughout much of the Mid-Atlantic Region. This includes extensive areas throughout central Maryland, southeastern Pennsylvania, northwestern Pennsylvania, and the Delmarva Peninsula. In addition, extensive areas in North Carolina and Virginia also have high probabilities of nitrate concentrations in ground water exceeding management thresholds of 1 milligram per liter and 3 milligrams per liter. The mapped areas showing a high predicted probability of nitrate concentrations in ground water exceeding 1 milligram per liter and 3 milligrams per liter correspond to areas that are mapped as cultivated land cover and/or overlying carbonate rocks. At a management threshold of 10 milligrams per liter (corresponding to the U.S. Environmental Protection Agency standard for nitrate in drinking water of 10 milligrams per liter), the predicted probability of nitrate concentrations in ground water exceeding this level is low for most of the Mid-Atlantic Region, except for the Delmarva Peninsula, southeastern Pennsylvania, and areas mapped as carbonate rocks in Virginia, Maryland, and Pennsylvania.

Scientific Investigations Report↗
USGS2003

Analytical and numerical simulation of the steady-state hydrologic effects of mining aggregate in hypothetical sand-and-gravel and fractured crystalline-rock aquifers

Analytical solutions and numerical models were used to predict the extent of steady-state drawdown caused by mining of aggregate below the water table in hypothetical sand-and-gravel and fractured crystalline-rock aquifers representative of hydrogeologic settings in the Front Range area of Colorado. Analytical solutions were used to predict the extent of drawdown under a wide range of hydrologic and mining conditions that assume aquifer homogeneity, isotropy, and infinite extent. Numerical ground-water flow models were used to estimate the extent of drawdown under conditions that consider heterogeneity, anisotropy, and hydrologic boundaries and to simulate complex or unusual conditions not readily simulated using analytical solutions. Analytical simulations indicated that the drawdown radius (or distance) of influence increased as horizontal hydraulic conductivity of the aquifer, mine penetration of the water table, and mine radius increased; radius of influence decreased as aquifer recharge increased. Sensitivity analysis of analytical simulations under intermediate conditions in sand-and-gravel and fractured crystalline-rock aquifers indicated that the drawdown radius of influence was most sensitive to mine penetration of the water table and least sensitive to mine radius. Radius of influence was equally sensitive to changes in horizontal hydraulic conductivity and recharge. Numerical simulations of pits in sand-and- gravel aquifers indicated that the area of influence in a vertically anisotropic sand-and-gravel aquifer of medium size was nearly identical to that in an isotropic aquifer of the same size. Simulated area of influence increased as aquifer size increased and aquifer boundaries were farther away from the pit, and simulated drawdown was greater near the pit when aquifer boundaries were close to the pit. Pits simulated as lined with slurry walls caused mounding to occur upgradient from the pits and drawdown to occur downgradient from the pits. Pits simulated as refilled with water and undergoing evaporative losses had little hydro- logic effect on the aquifer. Numerical sensitivity analyses for simulations of pits in sand-and-gravel aquifers indicated that simulated head was most sensitive to horizontal hydraulic conductivity and the hydraulic conductance of general-head boundaries in the models. Simulated head was less sensitive to riverbed conductance and recharge and relatively insensitive to vertical hydraulic conductivity. Numerical simulations of quarries in fractured crystalline-rock aquifers indicated that the area of influence in a horizontally anisotropic aquifer was elongated in the direction of higher horizontal hydraulic conductivity and shortened in the direction of lower horizontal hydraulic conductivity compared to area of influence in a homogeneous, isotropic aquifer. Area of influence was larger in an aquifer with ground-water flow in deep, low-permeability fractures than in a homogeneous, isotropic aquifer. Area of influence was larger for a quarry intersected by a hydraulically conductive fault zone and smaller for a quarry intersected by a low-conductivity fault zone. Numerical sensitivity analyses for simulations of quarries in fractured crystalline-rock aquifers indicated simulated head was most sensitive to variations in recharge and horizontal hydraulic conductivity, had little sensitivity to vertical hydraulic conductivity and drain cells used to simulate valleys, and was relatively insensitive to drain cells used to simulate the quarry.

Water-Resources Investigations Report↗
USGS1984

Shallow ground-water flow and drainage characteristics of the Brown ditch basin near the East Unit, Indiana Dunes National Lakeshore, Indiana, 1982

Brown ditch drains wetlands between three parallel ridges of sand dunes near the East Unit of Indiana Dunes National Lakeshore in Poter County, Indiana. Dune and lacustrine sands form a surficial aquifer that is the source of base flow to the ditch. Profiles established in July and August 1982 show that the average streambed slope of the ditch in the Lakeshore (0.19 percent) is six times that of the upstream east reach in the adjacent town, The Pines (0.03 percent). Although the ditch contains debris and vegetation, the reach in the Lakeshore seems to convey all the base flow it receives. In contrast, the upstream east arm contains several ponded sections where flow is sluggish. Digital model simulations show that dredging to form a uniformly graded streambed and eliminate ponding in the upstream east arm of the ditch south of The Pines could lower the water table 0.2 to 2.0 feet in The Pines. Lowering the ditch stage in the Lakeshore by 0.5 to 1.0 foot would cause additional water-table decline of less than 0.5 foot in The Pines. Lowering the ditch stage both in the Lakeshore and in The Pines could lower the water-table in the Lakeshore by nearly 1.0 foot. (USGS)

Water-Resources Investigations Report↗
USGS1992

Water resources and aquifer yields in the Charles River basin, Massachusetts

In 1984, about 66 Mgal/d (million gallons per day) of municipally supplied water was used by towns in the middle and upper Charles River basin and by the city of Cambridge in the lower basin. The Division of Water Resources of the Massachusetts Department of Environmental Management estimates that by the year 2020 an additional 11 Mgal/d of water will be needed to meet demands. The largest sources of un used water in the basin are 15 stratified-drift aquifers along the Charles River that are hydraulically con nected to the river and to its major tributaries. These aquifers, which consist mainly of medium to coarse sand and gravel, are generally unconfined, narrow, and discontinuous. They are as much as 100 feet thick, and transmissivity ranges from about 1,000 to 25,000 cubic feet per day per square foot times foot of aquifer thickness. The depth to the water table from the land surface generally is less than 15 feet and the water table fluctuates about 3 to 5 feet annually. Recharge to the aquifers is mostly derived from precipita tion, and the aquifers continuously discharge ground water to streams, ponds, lakes, and wetlands. Aquifer yields from the 15 aquifers under normal climatic conditions were related to the estimated duration of flow of the streams that drain the aquifers. Long-term (180 days or more) aquifer yields that are derived from interception of ground-water discharge range from less than 1 to more than 5 Mgal/d at the streamflow that is exceeded 95 percent of the time. However, because withdrawal may reduce streamflow in the Charles River to unacceptable levels, these yield estimates were adjusted to meet commonly cited minimum-streamflow criteria. For example, if minimum streamflows in the Charles River are maintained at no less than the annual 7-day, 10-year, mean low flow for 95 percent of the time, only three of the 15 aquifers could sustain yields of more than 1 Mgal/d from intercepted ground-water discharge. However, none of the aquifers could yield an appreciable amount of water from this source if streamflows are maintained at no less than this minimum for 99 percent of the time. Yields also are available to wells front water in duced by pumping to infiltrate through a streambed. Yields front induced infiltration from the Charles River were estimated for four aquifers and range from about 4 to 11 Mgal/d at the 95-percent flow duration. If minimum-streamflow criteria are to be maintained, these yield estimates would be substantially lower. A streamflow-accounting model was used to demonstrate the possible streamflow reductions that could result from further development of the water resources of the middle and upper Charles River basin. Results indicate that pumpage of an additional 11 Mgal/d could reduce flow in the river at Waltham by about 6 ft 3 /s (cubic feet per second) (4 Mgal/d) throughout the year--alt amount equal to about 24 percent of the flow that is exceeded 95 percent of the time. Also, model results indicate that increases in punzpage and the export of was-tewater from the basin of 23 Mgal/d would reduce streamflow at Waltham by about 14 ft 3 / s (9 Mgal/d) throughout the year.

Massachusetts↗
USGS2002

Hydrologic investigation of Powell Marsh and its relation to Dead Pike Lake, Vilas County, Wisconsin

An analytic element ground-water-flow model was constructed to help understand the ground- and surface-water hydrology in the vicinity of Dead Pike Lake and Powell Marsh, Vilas County, Wisconsin. The model was used to simulate the effect of removing Powell Marsh control structures (ditches and Vista Pond) on the hydrology of Dead Pike Lake. Measurements and model simulation results show that ground water in the vicinity of Powell Marsh moves to the northwest and west. If Powell Marsh structures are removed from the simulation, it does not affect the general direction of ground-water flow nor the total flow to Dead Pike Lake. Without the simulated structures, slightly more ground-water flow enters Dead Pike Lake and slightly less surface-water flows at the Dead Pike Lake inlet than with the simulated structures. Ground-water levels measured in piezometers installed along a flow path indicated that ground-water flow primarily is horizontal in the marsh and moves upward in the vicinity of a ditch where it discharges. Flow from Vista Pond is downward to the ground-water system but eventually also discharges upward to the ditches. Based on analyses of water samples from piezometers, the ditch, and Vista Pond, it was shown that dissolved iron is transported in the ground water. When ground water is discharged, iron and manganese react with dissolved oxygen, then precipitates, and forms the oxyhydroxide floc present in the Powell Marsh ditches. The processes involved in the transport and floc formation are not unique to the ditches, but are an expected outcome where discharging ground water and oxygenated surface water meet. Therefore, although floc formed in the ditches would no longer be available for transport if ditches were removed, it is likely that the floc formation would be redirected to the near-shore areas of Dead Pike Lake where increased groundwater discharge is expected.

Wisconsin↗
USGS1996

Simulated effects of alternative withdrawal strategies on ground-water-flow patterns, New Jersey Pinelands

A steady-state, three-dimensional ground-water- flow model of the unconfined part of the Kirkwood- Cohasey aquifer system beneath the upper parts of the Rancocas Creek and Wading River Basins in the New Jersey Pinelands was developed to (1) define ground-water-flow patterns and residence times in an aquifer system typical of the New Jersey Coastal Plain and (2) demonstrate the effects of alternative withdrawal strategies of ground- water-flow patterns and streams. Ground-water flow near the McDonald's-Middle Branch area was analyzed by using a particle tracker to demonstrate the effects of three hypothetical withdrawal scenarios on the configurations of source areas of ground-water flow to withdrawal wells, streams, and other discharge outlets in the Kirkwood-Cohansey aquifer system. Under natural conditions, most ground-water discharge to streams and wetlands. Ground-water residence times ranged from slightly greater than zero to about 200 years. Much of the ground water remained in the system for less than 20 years because it discharged to streams. Residence times of ground water were reduced significantly by persistent withdrawals. The configurations of source areas of flow to local stream systems and to the Piney Point aquifer are affected by the location of a withdrawal well. Results of withdrawal simulations indicate that well-location strategies applied in the Kirkwood-Cohansey aquifer system can alleviate the adverse effects of withdrawals on streams and that large-scale regional withdrawals in confined aquifers can adversely effect streams although the effects are dispersed over numerous streams.

Water-Resources Investigations Report↗