Document cortlandcountyny_gov_doc_d6cd7d24c6

HYDROGEOLOGY, WATER QUALITY, AND SIMULATION OF GROUND-WATER FLOW IN A GLACIAL-AQUIFER SYSTEM, CORTLAND COUNTY, NEW YORK Cosmos Hill West Branch East Branch \ \ Carr Hill Fine sand, silt and clay Confined sand and gravel aquifer Till Bedrock confining unit Prepared in cooperation with unconfined sand Cortland County Department of Health and gravel aquifer ancI Cortland County Department of Planning USGS U'S - GEOLOGICAL SURVEY science for a changing world Water-Resources Investigations Report 96-4255 ---PAGE BREAK--- ---PAGE BREAK--- Hydrology, Water-Quality, and Simulation of Ground-Water Flow in a Glacial Aquifer System, Cortland County, New York SyTodd S. Miller, Donald A. Sherwood, Peter M. Jeffers, and Nancy Mueller U.S. GEOLOGICAL SURVEY Water-Resources Investigations Report 96-4255 Prepared in cooperation with the CORTLAND COUNTY DEPARTMENT OF HEALTH CORTLAND COUNTY DEPARTMENT OF PLANNING science for a changing world USGS Ithaca, New York 1998 ---PAGE BREAK--- U.S. DEPARTMENT OF THE INTERIOR BRUCE BABBITT, Secretary U.S. GEOLOGICAL SURVEY Thomas J. Casadevall, Acting Director For additional information write to: Subdistrict Chief U.S. Geological Survey 903 Hanshaw Road Ithaca, NY14850 Copies of this report can be purchased from: U.S. Geological Survey Branch of Information Services Box 25286 Denver, CO 80225-0286 ---PAGE BREAK--- CONTENTS Abstract 1 Introduction 2 Purpose and Scope 3 Previous 3 Methods of Investigation 3 Well Inventory, Test Drilling, Levels, Water-Level Measurements, and Aquifer Test 4 Streamflow Measurements 4 Seismic-Refraction Surveys 4 Water Sampling and 4 Simulation Models 5 Description of the Study Area 5 Physiography and Bedrock Geology 5 Climate 6 Land 8 Acknowledgments 9 9 Glacial Geology 10 Bedrock Scouring 10 Kames and Kame Terraces 12 Valley Heads Moraine System 15 Proglacial Lakes and Lacustrine Deposits in Valleys 15 Glaciofluvial Sediments (Outwash) of the Valley Heads Episode 16 Ice Readvance and Drainage of Proglacial Lake During Late Stages of Deglaciation................. 16 Hydrology of the Glacial-Aquifer System 17 Geometry of Aquifers and Confining 18 Hydraulic 18 Unconfined Aquifer 19 Confined Aquifer 20 Confining Unit 20 Ground-Water Levels and Flowpaths 21 Unconfined Aquifer 21 Confined Aquifer 21 Sources of Recharge 22 Direct Infiltration of Precipitation on the Valley 22 Upland Sources 22 Induced Infiltration by Ground-Water Withdrawals. 25 Infiltration from Recharge Basins 25 Ground-Water Discharge 25 Seepage to Streams 25 Municipal and Industrial 26 Underflow From the Study Area 26 Water 26 Volatile Organic Compounds 28 Extent and Migration of Trichloroethylene in the Unconfined Aquifer 28 Horizontal Migration 28 Vertical Migration 29 Effects of Remediation Efforts 29 Long-Term Trends of Trichloroethylene Concentrations 35 Fate and Migration of Trichloroethylene 35 Inorganic Chemical Constituents and Physical Characteristics 38 Specific Conductance, pH, and Alkalinity 38 Chloride and Sodium 38 Trace Elements (Iron and 39 Nitrogen 41 Temporal Changes 42 Ground-Water Chemistry and Land 42 Contents Hi ---PAGE BREAK--- Simulation of Ground-Water 45 Description and Design of Numerical Model. 45 Model Grid 47 Geometry of Model 47 Hydraulic Conductivity. 48 Unconfmed Aquifer 48 Confined Aquifer 48 Stream-Aquifer 48 Boundary Conditions 49 Ground-Water Withdrawals and Recharge Basins 50 Model Calibration. 50 Model Sensitivity 56 Model Applications 58 Areas Contributing Recharge to Municipal Wells 58 Flowpaths to Wells and Streams from Sources of Contamination 59 Typewriter 59 Superfund Site 59 Summary 61 References Cited 64 APPENDIXES 1. Records of wells and test borings in the Cortland, N.Y., study 67 2. Inorganic chemical analyses of ground water from selected wells in the glacial-aquifer system in the Cortland, N.Y. study 75 3. Organic chemical analyses of ground water from selected wells in the glacial-aquifer system in the Cortland, N.Y. study area. 81 PLATES 1-5 Maps of Cortland, N.Y. study area showing: 1. Well locations and surficial geology in a glacial-aquifer system. 2. Altitude of potentiometric surface, direction of ground-water flow, and contributing areas to public-water supply wells during high-recharge conditions in the unconfined glacial aquifer, Cortland, New York 3. Altitude of potentiometric surface, direction of ground-water flow, and contributing areas to public-water supply wells during average-recharge conditions in the unconfined glacial aquifer, Cortland, New York 4. Altitude of potentiometric surface, direction of ground-water flow, and contributing areas to public-water supply wells during low-recharge conditions in the unconfined glacial aquifer, Cortland, New York 5. Finite-diiference grid and boundaries used to model the glacial-aquifer system FIGURES 1. Map showing location and principal geographic features of study area, Cortland County, N.Y 2 2. Map showing physiographic features of New 6 3. Diagram showing geologic time scale and when geologic units in the study area were formed 7 4. Graph showing annual precipitation at Cortland, 1973-92 8 5. Graph showing precipitation and departure from normal at Cortland, from January 1990 through May 1993 9 6. Map showing preglacial drainage in study 10 7. Hydrogeologic sections: A. Section A-A' 11 B. Section B-B' 12 C. Section C-C' 13 iv Hydrogeology, Water-Quality, and Simulation of Ground-Water Flow in a Glacial Aquifer System, Cortland County, New York ---PAGE BREAK--- D. Section 14 8. Map showing bedrock surface altitude in Cortland study area 14 9. Map showing location of Valley Heads ice and moraines that dammed drainage in valleys of the Tioughnioga River basin 15 10. Map showing upper surface altitude of lacustrine unit in Cortland study 17 11. Map showing positions of the Valley Heads ice, and of outwash that was deposited in valleys of the Tioughnioga River basin 18 12. Block diagram showing generalized hydrogeologic framework of the glacial-aquifer system in the Cortland study area 19 13. Block diagram showing sources of recharge to stratified drift in valleys in the glaciated Northeast 24 14. Map showing locations of streamflow-measurement sites in Cortland study area 27 15. Maps showing extent of trichloroethylene plume and contours of equal trichloroethylene concentration in Cortland study area: A. April 3-5, 1990 30 B. September 17-20, 31 C. April 27, 1993 32 16. Diagrams showing vertical distribution of TCE in ground water at two well sites: A. Well 341, 4,000 feet northeast of TCE source, July 19, 1990. B. Well 381, 750 feet from TCE source, September 33 17. Graphs showing trichloroethylene concentration at three wells: A. Well 350, November 4, 1986- May 20, 1993. B. Well 347, December 12, 1986-May 20, 1993. C. Well 364, December 12, 1986- April 29, 1993 34 18. Diagram showing physical, chemical, and biological processes that affect the fate and distribution of trichloroethylene in the unconfined aquifer at Cortland 36 19. Diagram showing transformation pathways of trichloroethylene under anaerobic 37 20. Graph showing dissolved chloride concentrations in water from wells at the City of Cortland well field for 1943-76, 1980, and 1990 41 21. Map showing distribution of dissolved chloride concentration in the unconfined aquifer in the Cortland area, April 1990 41 22. Boxplots showing concentration of alkalinity, dissolved calcium, nitrite plus nitrate, and dissolved chloride from selected wells in the Cortland, study area in 1976, 1980, and 1990: A. By year. B. By land use. 44 23. Geohydrologic section showing conceptual model of ground-water flow in the glacial-aquifer system in the Cortland study 46 24. Maps showing location of active and inactive grid boundaries in layers 1-3 of the three-dimensional ground-water flow model in the Cortland study area 47 25. Maps showing simulated head for in model layers 1-3 during steady-state simulations: A. high-recharge conditions 53 B. average-recharge conditions 54 C. low-recharge conditions 55 26. Graph showing results of sensitivity analyses for hydraulic head in the glacial-aquifer system in Cortland study area 56 27. Graphs showing results of sensitivity analyses for ground-water discharge to major streams during average recharge conditions in the glacial aquifer system in the Cortland study area: A. West Branch Tioughnioga River B. Reaches A and B of Tioughnioga 57 28. Maps showing flowpaths and traveltimes of ground water from two sources of contamination in Cortland study area during high-, average-, and low-recharge conditions 60 TABLES 1. Land use in the Otter Creek-Dry Creek basin, Cortland County, N.Y., 1992 8 2. Results of aquifer tests in the glacial-aquifer system in the Cortland, N.Y. study 20 3. Streamflow losses and gains in tributary streams in the Cortland study area, 23 4. Ground-water discharge to West Branch Tioughnioga River, Tioughnioga River, and to Otter Creek during average- and low-flow conditions, 25 Contents v ---PAGE BREAK--- 5. Ground-water withdrawals from the Cortland study area by major pumping wells during low-, average-, and high-recharge conditions, 1990-91 27 6. Range of half-lives for trichloroethylene and its degradation products. 38 7. Minimum, maximum, mean, median, and interquartile range of concentration or value for selected constituents or properties of ground-water samples collected from the glacial aquifer in the Cortland study area during April and September 1990 39 8. Minimum, maximum, mean, median, and interquartile range of concentration or value for selected constituents or properties of ground-water samples collected from the glacial aquifer in the Cortland, N.Y. study area during: A. April 1990. B. September 1990. 40 9. Minimum, maximum, mean, median, and interquartile range of concentration or value for selected constituents or properties of ground water sampled during three studies in Cortland, N.Y. study 43 10. Minimum, maximum, mean, median, and interquartile range of concentration or value for selected constituents or properties of ground water in the Cortland, N.Y. study area, by land 45 11. Difference between measured and simulated heads at 49 selected wells in the Cortland, N.Y. study area for high-, average-, and low-recharge 51 12. Measured and simulated streamflow gains and losses, for selected stream reaches during high-, average-, and low-recharge conditions in the Cortland, N.Y. study 52 13. Steady-state water budgets for the glacial-aquifer system at Cortland N.Y, for high-, average-, and low-recharge 52 CONVERSION FACTORS AND VERTICAL DATUM Multiply By To Obtain Length inch (in.) 2.54 centimeter foot (ft) 0.3048 meter mile (mi) 1.609 kilometer Slope foot per mile (ft/mi) 0.1894 meter per kilometer foot per foot (ft/ft) 0.3084 meter per meter Area square mile (mi2) 2.590 square kilometer acre 0.40483 hectare Flow cubic foot per second (ft3/s) 0.02832 cubic meter per second inch per year (in/yr) 25.4 millimeter per year million gallons per day (Mgal/d) 3.785 cubic meters per day gallons per minute (gal/min) 0.06309 liter per second feet per day 0.3084 meter per day Temperature degrees Fahrenheit °C = 5/9 (°F-32) degrees Celsius Specific Conductance microsiemens per centimeter at 25° Celsius (jlS/cm) Equivalent Concentration Terms milligrams per liter (mg/L) = micrograms per gram (mg/g) = parts per million micrograms per liter (IXg/L) = micrograms per kilogram (Ilg/kg) = parts per billion Vertical Datum: In this report, "sea level" refers to the National Geodetic Vertical Datum of 1929 a geodetic datum derived from a general adjustment of the first-order level nets of the United States and Canada, formerly called Sea Level Datum of 1929. vi Hydrogeology, Water-Quality, and Simulation of Ground-Water Flow in a Glacial Aquifer System, Cortland County, New York ---PAGE BREAK--- Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York By Todd S. Miller, Donald A. Sherwood, Peter M. Jeffers, and Nancy Mueller Abstract The glacial-aquifer system in Cortland County consists of an unconfined sand and gravel aquifer 40 to 80 feet thick, underlain by a lacustrine and till unit 1 to 155 feet thick that, in turn, is under- lain by a confined sand and gravel aquifer 1 to 170 feet thick. The two aquifers are hydraulically connected in some places along the valley walls where the confining layer is absent. Water in the unconfined aquifer generally moves from areas of recharge along the valley walls toward the center of the valley, then flows northeastward down the Otter Creek-Dry Creek valley and discharges to pumping wells, the West Branch Tioughnioga River, and the Tioughnioga River. Water is pumped from the unconfined aquifer at municipal and industrial wells at a rate of 6.76 to 7.20 million gallons per day. Trichloroethylene (TCE) that was detected in water samples from several wells in 1986 was considered a threat to municipal-water supplies. Results of ground-water sampling in April and September 1990 and in April 1993 indicated that a TCE plume, as defined by concentrations equal to or greater than 5 micrograms per liter, had migrated 1.25 miles northeastward in the uncon- fined aquifer from a spill area at a typewriter production plant in the west-central part of the aquifer. The extent of the plume was the same in all three sampling periods, indicating that steady- state conditions had been reached. TCE concen- trations were below the U.S. Environmental Protection Agency's "Maximum Contaminant Level" of 5 micrograms per liter 1 mile upgradient from the City of Cortland municipal wells, which are 2.25 miles downgradient (northeast) of the spill area. Inorganic- and organic-chemical analy- ses of ground-water samples collected during April and September 1990 indicate that water generally meets New York State drinking-water standards except in the part of the unconfined aquifer that is contaminated by TCE. A ground-water flow model (MODFLOW) was used to simulate high-, average-, and low-recharge conditions in the aquifer system, and a particle- tracking routine (MODPATH) was used with output from the flow model to estimate the area contributing water to municipal wells and to delineate flowpaths of ground water from two sources of contamination. The simulated water budgets indicate that the largest source of recharge to the aquifer system (55 to 58 percent of total recharge) is from the uplands; this recharge includes seepage losses from upland streams that flow onto the aquifer, and unchanneled runoff and ground-water inflow from the uplands. The second-largest source of recharge is precipitation (33 to 39 percent of total recharge) that directly falls over the aquifer. Most ground-water discharge (57 to 71 percent of total) occurs as seepage from the aquifer system into streams, and some (26 to 40 percent of the total) occurs as discharge to pumping wells. Results of particle- tracking analyses indicate that the contributing areas are U-shaped and extend over most of the aquifer upgradient from the pumping wells. The largest contributing areas were obtained in the simulation of low-recharge conditions, and the smallest areas were obtained in the simulation of high-recharge conditions. Ground-water flowpaths from sources of contamination shifted southward in the low-recharge simulation in response to changes in the distribution of recharge. Abstract 1 ---PAGE BREAK--- INTRODUCTION Glacial aquifers are the principal sources of water for many communities in upstate New York, but the high permeability and shallow depth to the water table in these aquifers make them highly susceptible to contamination. Potential contamination sources can include leaking petroleum-product storage tanks, leachate from landfills and septic systems, road- deicing salts, agricultural pesticides and fertilizers, and chemical spills (such as solvents and degreasers) at commercial and industrial facilities. Protection of these aquifers from contamination is critical in areas where ground-water use is great and alternative sources of drinking water are not readily available. Water managers need information on the size and location of the areas that contribute recharge to their public-supply wells, and the direction and rate of flow of ground water from known or potential sources of contamination, to protect the quality of the water. The City of Cortland and surrounding communi- ties obtain water supplies from a highly productive glacial aquifer system (fig. This system has been designated as a "Primary Aquifer" by the New York 76°15' 12'30 EXPLANATION STUDY AREA BOUNDARY OF GLACIAL AQUIFER MUNICIPALITY BOUNDARY DRAINAGE DIVIDE between Oswego River basin and Susquehanna River basin Area shown in this figure DIRECTION OF STREAMFLOW MUNICIPAL WELL FIELD Cortland County 0 5 10 KILOMETERS Area shown in figures 8 and 10 Area shown in figure 1 City of Cortland well field Sewage-treatment plant CITY 0F CORTLAND Lime Hollow Road municipal well errace Road well field 0 1 2 TOWN OF CORTLANDVILLE 9.5 miles to Marathon 3 KILOMETERS TOWN OF VIRGIL Base from U.S.Geological Survey 1:62,500 series: Cortland (1903) and Groton (1903) Figure 1. Location and principal geographic features of study area, Cortland County, N.Y. 2 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- State Department of Environmental Conservation and as a "Sole Source Aquifer" by the U.S. Environmental Protection Agency (EPA) under the Safe Drinking Water Act. Several parts of the glacial-aquifer system have been contaminated by: solvents and degreasers, including trichloroethylene (TCE), trichloroethane (TCA), and dichloroethylene (DCE); gasoline that leaked from tanks from at least two service stations; bacteria from failing septic systems; and leachate from inactive hazardous-waste sites. Concen- trations of chloride and nitrate in the aquifer generally meet State standards for drinking water but have been steadily increasing since the 1950's as a result of road salt, lawn and garden fertilizers, and leachate from septic systems (Buller and others, 1978). A study of ground-water contamination (O'Brien and Gere Engineers, Inc., 1987) at a typewriter - production plant in the western part of the aquifer, at which solvents and degreasers were found, resulted in a legal settlement that included funds to study the hydrogeologic framework of the aquifer, the migration of contaminants from the site, and the investigation of other possible sources of off site contamination. In 1989, the U.S. Geological Survey (USGS), in cooper- ation with the Cortland County Departments of Health and Planning, began a 31/2-year hydrogeologic, water-quality, and numerical modeling study of the glacial-aquifer system in Cortland County. Major objectives of the study were to increase knowledge on how the aquifer system works; provide a numeri- cal ground-water flow model to help manage the ground-water resource; and to determine the extent of a TCE plume, the current ground-water quality in the aquifer, changes (if any) in water quality during the 15-year period since wells were sampled by Buller and others (1978), and whether the ground water meets New York State drinking-water standards. Purpose and Scope This report describes the hydrogeology of the glacial-aquifer system including the geologic framework; the ground-water flow system, includ- ing ground-water and surface-water interaction, water levels, and water budgets; the water quality, with an emphasis on the extent, trends, and fate of TCE; and results of model simulations of ground-water flow for high-, average-, and low-recharge conditions, including the estimation of recharge areas to municipal wells, and the advective flowpaths of contaminants migrating from sources of contamina- tion. Also included are geologic sections; maps and diagrams depicting well locations, geology, ground-water levels, and direction of ground-water flow for measured and simulated high-, average-, and low-recharge conditions; recharge areas to public water supplies, concentrations of TCE and several other selected chemical constituents, and flowpaths and traveltimes of contaminants; and tables of climate and land-use data, well records, a water budget, and data on hydraulic properties of the aquifers and on water quality. Previous Investigations USGS investigations of glacial aquifers in Cortland County began when Asselstine (1946) identi- fied sources of ground-water supplies and collected well and water-quality data. Randall (1972) collected well and test-boring records in the study area. Buller and others (1978) and Miller and others (1981) studied surficial geology, movement of ground water, and water quality in the Otter Creek-Dry Creek valley. Cosner and Harsh (1978) constructed a two-dimen- sional ground-water model of the glacial aquifer in the Otter Creek-Dry Creek valley; this model was later modified by Reynolds (1985). Several hydrogeological and engineering consultants conducted site-specific studies at chemical-spill sites in the study area. O'Brien and Gere Engineers, Inc. (1987 and 1990) studied the hydrogeology and organic-chemical contamination in ground water at the typewriter-production plant in the western part of the study area. Blasland, Bouck and Lee, Engineers (1992) investigated the hydrogeology and extent of chemical contamination of an EPA-designated "Superfund site," locally known as the Rosen Superfund Site, in the southeastern part of the study area (pi. Apfel (1967) studied the availability of ground water at a machine- tool plant in the southwestern part of the study area, and Galson Technical Services, Inc., (1988) investi- gated the hydrogeology and extent of a petroleum- product spill at that site. Resource Engineering (1987) investigated the hydrogeology and extent of contami- nation from paint-stripping activities at a center for the handicapped in the northern part of the study area. Methods of Investigation Hydrogeologic data were collected from several sources, including published reports by the USGS Introduction 3 ---PAGE BREAK--- consulting firms, and data in files of local drillers and government agencies. USGS fieldwork during this study included a well inventory, test drilling, leveling, an aquifer test, water-level measurements, streamflow measurements, two seismic-refraction surveys, and three rounds of ground-water sampling. Well Inventory, Test Drilling, Levels, Water-level Measurements, and Aquifer Test Records of 215 wells were collected and compiled (appendix The well inventory was augmented by test drilling in which 20 observation wells and 4 test holes were installed by auger, cable-tool, and air-rotary rigs to obtain data on stratigraphy, hydraulic proper- ties, water quality, and water levels in the aquifers. Leveling was done to about 100 wells to determine the elevations of water-level-measuring points, which were mostly the tops of the well casings. These eleva- tions were determined by standard surveying methods (Kennedy, 1990) and are generally accurate to within 0.01 ft. Levels were also run to determine channel profiles of major streams that flow over the aquifer. The stream-channel elevations were used in the stream-routing package for the ground-water models. Water levels were measured in 50 wells from July 1989 through October 1991 and during three synoptic rounds in more than 100 wells during March 28-29,1990 (high-recharge conditions), May 28-June 4,1991 (average-recharge conditions), and October 7-9,1991 (low-recharge conditions). These water-level data were contoured to show the potentio- metric surfaces (pis. 2 through 4) for high-, average-, and low-recharge conditions and were used to calibrate the models. An aquifer test was conducted July 16,1991 at the Town of Cortlandville municipal well at Lime Hollow Road (fig. 1) to calculate the hydraulic conductivity of the aquifer materials. Drawdown data were analyzed through a curve-fitting procedure that uses type curves developed by Nueman (1974) for partially penetrating wells in an unconfined aquifer. Streamflow Measurements Streamflow was measured by current meters in several reaches of most streams in the study area during each of the three recharge conditions that were modeled; techniques are described by Buchanan and Somers (1969). Streamflow measurements were used to identify the location and the amount of gain or loss in streams. Seismic-Refraction Surveys Seismic-refraction surveys were conducted at two sites (pi. 1) to supplement data from test drilling. These surveys obtained continuous records on depth to water table and to bedrock. Seismic-refraction techniques used in this study are described by Haeni (1988). A series of 12 geophones spaced 100 ft apart were laid on the ground, and arrival times of compres- sional waves generated by explosives were recorded and plotted as a function of source-to-geophone distances. A three-layer (unsaturated unconsolidated sediments, saturated unconsolidated sediments, and top of bedrock) boundary-formula computer analysis (Scott and others, 1972) was used to calculate depths to water table and to bedrock. Water Sampling and Analysis Water samples were collected from wells and streams in April 1990, September 1990, and April 1993. The main purpose of the two 1990 sampling rounds was to map the extent of TCE migrating from the former typewriter-manufacturing plant in the western part of the aquifer under high- and low- recharge conditions and to define the general chemical quality of ground water throughout the study area. The purpose of the April 1993 sampling was to evaluate what effects remedial practices used at the typewriter plant were having on concentra- tions of TCE in ground water. Selection of purging techniques to remove stand- ing water from well casings before sample collection depended on well construction and water-yielding capacity of the well. At least three volumes of water were pumped or bailed from monitoring wells with good yield prior to sample collection. Monitoring wells with low yield were pumped dry, then allowed to partly recover before they were sampled. Large wells (6-in. diameter and larger) were purged with a stain- less-steel submersible pump or a 4-in. diameter bailer; wells of 2-in. diameter and smaller were purged either by a Teflon 1 bladder pump, a peristaltic pump (with Tygon tubing), or a stainless-steel bailer. The pumps of domestic wells were turned on for 10 to 15 minutes (the estimated time for evacuation of three volumes of standing water in the casing) before sample was 1 Use of trade, product, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government. 4 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- collected. All sampling and purging equipment were thoroughly cleaned between sampling. The Chemistry Department at State University of New York (SUNY) College at Cortland analyzed the water samples collected in April 1990 for pH, specific conductance, and total alkalinity; common ions (calcium, chloride, magnesium, sodium, and potassium; nitrate as nitrogen); and volatile organic compounds (VOCs). Temperature and specific conductance of water was measured in the field at the time samples were collected. The SUNY Chemistry Department also analyzed water samples collected in September 1990 and June 1993 for VOCs. The USGS National Water-Quality Laboratory in Denver, Co., analyzed water samples collected in September 1990 for nutrients, metals and common anions. VOC analyses were done by purge- and-trap/gas chromatograph methods, modified from USEPA 501 methods (U.S. Environmental Protection Agency, 1979). Quality-control procedures for samples collected during April 1990 consisted of analyses of duplicate samples, replicate samples, and field and laboratory spikes of inorganic constituents. Two samples that were to be analyzed for VOCs were split and spiked; one set consisting of two samples was sent to the USGS National Water Quality Laboratory, and one set was sent to the Chemistry Department at SUNY College at Cortland. Analytical results and recoveries reported by the two laboratories compared favorably. Simulation Models A three-dimensional, finite-difference, ground- water flow model, MODFLOW (McDonald and Harbaugh, 1988), was developed to represent the glacial aquifer system and was used for steady-state simulation of three recharge conditions (high, average, and low). The models were used to compute water levels and water budgets. Particle-tracking postprocessing programs, MODPATH and MODPATH-PLOT (Pollock, 1989), were used to estimate the recharge areas of municipal wells, the flowpaths of the advective phase of chemical migra- tion, and traveltime of the advective phase of contam- inants migrating from chemical-spill sites to discharge areas. Description of Study Area The study area, in the Otter Creek-Dry Creek valley and in parts of West Branch, East Branch, and Tioughnioga River valleys (fig. forms a rectangle 6 mi long by 2 mi wide in the Town of Cortlandville, in the southwestern part of Cortland County (fig. The City of Cortland lies within the study area and within the Town of Cortlandville. The population of the Cortland area increased about 7 percent during 1970-90 and, in 1990, was about 24,100 (U.S. Bureau of Census, 1990). Physiography and Bedrock Geology The study area is in the glaciated part of the Allegheny Plateau (fig. The rocks of the plateau consist of gently folded layers of shale with some siltstone, sandstone, and limestone that dip to the south at 40 ft/mi. These rocks are part of the "Catskill Delta" complex (Woodrow and Sevon, 1985) and were deposited in marine seas during late Devonian time (fig. The rocks were then uplifted above sea level millions of years later and then eroded to a nearly flat plain by the middle of the Cenozoic (Isachsen and others, 1991). The rocks were again uplifted during Late Cenozoic time to form the Allegheny Plateau, which was dissected by streams to form a hilly terrain. Bedrock in the study area is far less permeable than the overlying glacial sand and gravel aquifers. Bedrock forms the bottom of the aquifer system in the study area. Relief in the study area is about 720 ft, and summits are as high as 1,800 ft above sea level. The lowest elevation, about 1,080 ft, is the channel of Tioughnioga River, in the eastern part of the study area. The bedrock surface has a relief of about 1,000 ft in areas where it is more than 300 ft below the flood plain. Uplands have moderately sloping hillsides ranging from 9 to 18 percent. Muller (1970) described the drainage pattern at Cortland as "the hub of five valleys that converge from the north, northeast, east, south, and southwest" (fig. All five valleys drain into the Tioughnioga River in the eastern part of the study area. Most of the study area is drained by Otter Creek and Dry Creek, which flow northeastward and drain into the West Branch Tioughnioga River in the north-central part of the study area. The western part of the study area contains several kettle ponds, some with bottom sediments of marl that have been mined in the past. Most of the ponds are dry during late summer and fall. Stupke Introduction 5 ---PAGE BREAK--- 80° 79° 45' 44' 43' 42' 41' 78° CANADA 77° 76° I St. Lawrence- Champlain Lowlands 75° 74° 73° 72° Area shown in figures 9 and 11 NEW YORK . 0 25 50 75 100 MILES 0 25 50 75 100 KILOMETERS \JERSEY , t I ATLANTIC OCEAN Figure 2. Physiographic features of New York. (Modified from Cressey, 1966, fig. 9, p. 26.) pond (fig. 1) is a large kettle pond that has no inlet; its outlet is a headwater tributary of Otter Creek. Climate Two types of air masses provide the dominant weather characteristics of Cortland cold, dry air- masses from the northern interior of the continent, which prevail during the cold half of the year, and warm, humid airmasses from the Gulf of Mexico and southwestern part of the continent, which are typical during the warm half of the year (Dethier, 1966). A third, but less common type is the marine airmasses, which move along the North Atlantic Coast (northeast- ers) and produce occasional cool, wet, cloudy condi- tions. Large bodies of water near Cortland, such as Lake Ontario, 70 mi to the north, tend to moderate air temperatures and supply moisture to the cold, conti- nental airmasses during the cold season. Mean annual temperature of Cortland is 45.7 °F (National Oceanic and Atmospheric Administration, 1990). Winters are long, cloudy, and cold, with minimum temperatures ranging from 0 to -10 °F for an average of 13 days per winter (National Oceanic and Atmospheric Administration, 1992). Summer has warm daytime temperatures, cool evenings, and occasional periods of high, uncomfortable humidity. Summer daytime temperatures typically range from the mid-70s to mid-80s reach 90 °F or higher on average of 8 days a year (Roffner, 1985). The annual precipitation at Cortland for 1973-92 is shown in figure 4; the average for the 20-year period was 41.2 in., and the range was from 32.3 in. in 1982 to 50.6 in. in 1977. precipitation is normally distributed fairly uniformly throughout the year, but from January 1990 to May 1993 it ranged from 1.35 in. in March 1990 to 8.42 in. in July 1992 (fig. Irriga- tion for agricultural crops is rarely needed. Severe droughts are rare, but minor droughts that occur periodically cause concern about degradation of water quality of streams during extreme low-flow conditions and offish survival in reaches that are downgradient of sewage discharges. Snowfall in Cortland averaged 90.9 in. during 1973-92 (National Oceanic and Atmospheric Admin- istration, 1992). Continuous snow cover typically starts in late December and lasts until mid-March. Some snow is a result of the "lake effect," a process whereby an artic cold front from Canada crosses Lake 6 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- ERA PERIOD EPOCH Millions of years ago Important geologic events in Central New York 500 1,000 2,000 COcc < HI LL O CO § 3,000 4,000 5,000 Mesozoic Cenozoic Paleozoic Precambrian Time Quaternary Tertiary Cretaceous Jurassic Triassic Origin of the earth about 4.5 billion years ago Permian Mississipian Devonian Silurian Ordovician Cambrian Holocene Pliocene Miocene Oligocene Eocene Paleocene Late Early Late Middle Early Late diddle Late Early Middle Early Late Early Late Middle Early Late Middle Early 0.1 1.6 5 24 38 55 66 96 138 205 240 290 330 360 374 387 410 435 500 570 Cortland glacial aquifer formed 14,000 years ago Erosion of Paleozoic rocks to form dissected Allegheny Plateau Sediments deposited in seas will form the sedimentary rocks underlying Cortland Figure 3. Geologic time scale. Description of Study Area 7 ---PAGE BREAK--- Mean precipitation for period of record (1973-92) YEAR Figure 4. Annual precipitation at Cortland, N.Y., 1973-92. (Data from National Oceanic and Atmospheric Administration, 1992) Ontario and the Finger Lakes, is warmed in the low elevations and picks up moisture^ and then releases some of the moisture as snow as the air rises over the uplands (orographic effects). When the lakes freeze by midwinter,"lake effect" snow becomes infrequent because the ice prevents evaporation off the lake. The heaviest snowfalls, typically 1 to 2 ft, occur when northeasters move inland and cross over or close to Cortland. A severe "northeaster," locally known as the "Blizzard of 93," dumped 3 ft of snow on Cortland during March 13-14,1993 (National Oceanic and Atmospheric Administration, 1993). Land Use Land use in the Otter Creek-Dry Creek basin was inventoried from the 1992 Cortland County Tax Assessment Rolls and from field inspections made in 1992 (table Agricultural land occupies 35.5 percent of the Otter Creek-Dry Creek drainage basin (5 percent of land over the aquifer). The study area contains three working farms that grow mostly hay and corn. Agricultural land has been decreasing since the 1950's. The prime farmland is in flat outwash valleys, which are also most suitable for urban development; consequently, pressure to convert farmland to residen- tial, commercial, and industrial has increased. The trend of decreasing farm land will probably continue. Residential, the largest land use, accounts for 29 percent of the Otter Creek-Dry Creek drainage basin. Most of the residential area is within the City of Cortland, and all residential areas within the City of Cortland are sewered. Surrounding the city, in the Town of Cortlandville, are many large suburban housing developments. Most of the residential areas in the Town of Cortlandville are sewered, although residential growth has recently begun to outpace the extension of sewer lines. Most of the residential areas in the bedrock uplands are unsewered. Commercial land use (8 percent of the Otter Creek- Dry Creek drainage basin) has been continually expand- ing, especially since 1985, primarily in the form of strip development along major transportation corridors. Most commerce consists of retail sales and services. Industrial land use also constitutes 8 percent of the Otter Creek-Dry Creek basin. Manufacturing, process- ing, and warehousing are included in this category. Major products in 1992 were foundation garments, plastic houseware, packaging, industrial filters, fishing line, machine tools, and wood products. (The typewriter plant moved to Mexico in 1994.) Sand and gravel mining constitutes about 2.8 percent of land in the Otter Creek-Dry Creek drainage basin. Most of the gravel is extracted for asphalt production. Table 1 . Land use in the Otter Creek-Dry Creek basin, Cortland County, N.Y., 1992 [Data from Cortland Department of Planning, 1992] Percentage of use__________total basin area Agricultural cropland and pasture 35.5 Residential (sewered) 19 Residential (unsewered) 10 Commercial 8 Industrial 8 Forest and brushland 5 Churches, institutions, and public services 4.4 Parks, athletic fields, and golf courses 3 Sand and gravel mining 2.8 Major transportation corridors 2.0 Water and wetlands 0.9 Other recreational areas 0.8 Cemeteries 0.6 8 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- Public land constitutes 7 percent of the Otter Creek-Dry Creek drainage basin. Fortuitously, the City of Cortland owns a 180-acre forested watershed that borders the city's well field; thus, the city's water managers have been able to protect their ground-water supply from contamination, largely by controlling the land use on this property. Other substantial public lands in the study area are the SUNY college campus and a 100-acre fish laboratory owned by the U.S. Fish and Wildlife Service. The campus extends over the aquifer and a bedrock hill (pi. The fish laboratory is on the west side of the ground-water divide in the western part of the aquifer. Other small parcels of public lands include municipal recreational parks and government institution and service sites. Regulation is the primary means of controlling land use in the study area. Both the City of Cortland and the Town of Cortlandville have zoning ordinances, subdivision controls, and land-use plans. At the County level, the Cortland County Health Department enforces the New York State Sanitary Code, and the Cortland County Planning Board reviews land-use proposals of county wide significance. The Town of Cortlandville adopted an Aquifer Protection District amendment to its zoning ordinance in 1988, to regulate certain land uses over the aquifer. Acknowledgments Special thanks are extended to James Feuss, Director of Cortland County Health Department, for providing well and chemical data. Thanks are also extended to Douglas Withey, Superintendent of Cortland County Water Board, for providing pumping data and access for collection of hydrogeo- logic data in the Waterworks area. Thanks are also extended to Hayne Smith, Town of Cortlandville Engineer, for arranging the aquifer test at the munic- ipal well at Lime Hollow Road. HYDROGEOLOGY The glacial aquifer system in the study area, and the ground-water and surface-water flow conditions within it, are the result of glacial processes. Hydro- geologists use concepts and principles of geologic processes and the chronological order in which they II 1 1 1 III 1 oO 1 ' ' ' 1 ' ' CO VALUE D NORMAL 5 DEPARTURE FROM NORMAL, IN INCHES O 00 co T CM CO - "co CM CO CO 00 T- d - 1 o c\i. I 1 1 1 CM O ^ CM T. pS S9 9 Ti" i n ~ ill L L 1 J FMAMJ J ASC 1990 CM CM O) CM 0 o * ^ o in CO 0 ' ' 0 " CO odd oq o - 9 o II II _ I CM I, 0 TT 0> CO 0 1 1 1 CD CD ]Bj 9 7 JASON 1991 n ^ fen 9 d 9 gj CM I-i " I-i r i ^ co V o J.L L L L L L DJ FMAMJ J COo CM CD in- d 8 co TJ- d d T ' CO 9 O) W CD CD . L L L L L L * _ S" 1 1 ASONDJ FMAM 1992 1993 i 5 a. O Figure 5. precipitation and departure from normal at Cortland, N.Y., January 1990 through May 1993. (Data from National Oceanic and Atmospheric Administration.) Hydrogeology 9 ---PAGE BREAK--- have occurred to conceptionalize and interpret the hydrogeologic framework of an aquifer. These processes, when sufficiently understood, form a basis for interpolation and correlation between data points. Glacial processes are highly diverse and have resulted in complex stratigraphy and ground-water flow patterns in the study area; therefore, extensive data are needed for accurate representation of the hydrogeo- logic system. Glacial Geology Erosion by ice and meltwaters during the last glaciation modified the bedrock topography in the study area and removed most of the previously depos- ited unconsolidated materials (Muller and others, 1988); thus, sediments from the last glaciation (Late Wisconsinan) are prevalent. Some of the sediments that were eroded by ice became entrained in the glacier, or were dragged along its bottom, where they were ground up and later deposited as lodgment till atop bedrock or older glacial deposits. Lodgment till consists of poorly sorted clay, silt, sand, and stones that were compacted by the ice and is a poor aquifer; it is often referred to by drillers and farmers as "hardpan" or "boulder clay" their testimony to its toughness to drill into or plow. Valleys in the study area form a dendritic drainage pattern and increase in width to the southwest, indicat- ing that water in the preglacial drainage system in the East and West Branch Tioughnioga River Valleys flowed southward, that Trout Brook flowed westward, and that a stream in the reach of the Tioughnioga River Valley between Polkville and Blodgett Mills flowed northwestward (fig. All four streams converged at Cortland to form a major ancestral river that flowed south westward in the Otter Creek valley. The south- westward route of the Tioughnioga River was blocked by ice in the Fall Creek valley during Late Wisconsian time and was diverted to the southeast, where it now flows in a narrow valley and exits the study area near Bloggett Mills (fig. The diversion of the Tiough- nioga River in the study area resulted in a "separated valley" (Randall and others, 1988), which is a broad valley that is partly filled with stratified drift, is drained by minor streams (Otter Creek and Dry Creek), and abuts a large stream at one end (Tiough- nioga River). The small streams in separated valleys are vulnerable to depletion by large ground-water withdrawals, which lower the water table below the streambed and thereby induces water to seep through the streambed into the aquifer. 76° 15' 12'30" 10' 7'30" Factory Brook Stupke Pond Tioughnioga River P reglacial drainage flowed southwi Blodgett Mills 9.5 miles to Marathon 1 Present drainage flows south Base from U.S. Geological Survey 1:62,500 series: Cortland (1903) and Groton(1903) 1 2 MILES 0 1 2 KILOMETERS EXPLANATION STUDY AREA AQUIFER BOUNDARY PREGLACIAL DRAINAGE- arrow shows direction of preglacial streamflow PRESENT DRAINAGE DIVIDE between Oswego River basin and Susquehanna River basin Figure 6. Preglacial drainage in Cortland, N.Y., study area. Bedrock Scouring Rowing ice and subglacial meltwaters first removed older unconsolidated deposits, then scoured the underlying Devonian-age shales and siltstones, resulting in widened and deepened bedrock valleys characteristic of glacial terrain. The bedrock valleys are asymmetrical in some places, such as in the Otter Creek Valley near South Cortland (geologic sections A-A' and B-B', figs. 7A and 7B), where a buried bedrock bench runs along the northern valley wall, and are nearly symmetrical in other places, such as in the West Branch Tioughnioga River valley (sections C-C' and D-D', figs. 7C and 7D). The longitudinal bedrock profile is nearly flat in the West Branch Tioughnioga River and Otter Creek- Dry Creek valleys, where elevations range from 900 to 950 ft above sea level (figs. 7C and The bedrock floor slopes southward in the East Branch Tioughnioga 10 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- 1,000 - 950 - 900 - 850 VERTICAL EXAGGERATION X 20 DATUM IS SEA LEVEL EXPLANATION I 500 METERS cc cc LU o KAME MORAINE- poorly sorted silt, sand, and gravel deposited during Valley Heads episode OUTWASH- stratified sand and gravel LACUSTRINE DEPOSITS- stratified very fine sand, silt, and clay KAME DEPOSITS- hummocks composed of stratified, poorly sorted silt, sand, and gravel TILL- poorly sorted clay, silt, sand, and stones 326 SHALE AND SILTSTONE WATER TABLE- water levels measured during May 28-29, 1991 WELL AND WELL NUMBER Figure 7A. Hydrogeologic section A-A'. (Line of section and well locations are shown in pi. River Valley (Reynolds, 1987) and slopes northwest- ward in the Tioughnioga River Valley from the col (preglacial divide) near Messengerville to the eastern part of the City of Cortland. A conspicuous depression in the bedrock floor in the eastern part of the city (fig. 8) at well 430 (pi. 1) is the location of the lowest Hydrogeology 11 ---PAGE BREAK--- West B 1,050 1,000 - 950 900 VERTICAL EXAGGERATION X 20 DATUM IS SEA LEVEL EXPLANATION o m, KAME MORAINE- poorly sorted silt, sand, and gravel deposited during Valley Heads episode OUTWASH- stratified sand and gravel LACUSTRINE DEPOSITS- stratified very fine sand, silt, and clay KAME DEPOSITS- hummocks composed of stratified, poorly sorted silt, sand, and gravel TILL- poorly sorted clay, silt, sand, and stones SHALE AND SILTSTONE WELL AND WELL NUMBER Figure 7B. Hydrogeologic section B-B'. (Line of section and well locations shown in pi. elevation of the bedrock surface (801 ft) and the thick- est valley fill found in the study area (313 ft thick). In the center of the city is an isolated, 120-ft-high, egg-shaped (from top view) bedrock hill (fig. 1) that is surrounded by valley-fill deposits more than 200 ft thick. This landform is known as an umlaufberg (Muller, 1970), which formed when meltwaters depos- ited outwash around the hill and buried a preglacial bedrock spur between the northeast side of the hill and the bedrock massif hills to the north. The SUNY campus is on this hill. Kames and Kame Terraces As the glacier began to retreat from its maximum extent in northern New Jersey about 19,000 years before present (Late Wisconsinan time) it took about 5,000 years for the ice front to recede northward and give ground to the Cortland area (Muller and Calkin, 1993). As the glacier receded from the study area (between 14,000 and 15,000 years before present), the uplands emerged first because the ice was thinner there than in the valleys. Meltwater and upland streams that 12 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- 900 - 950 900 850 850 VERTICAL EXAGGERATION X 20 DATUM IS SEA LEVEL EXPLANATION 1,000 2,000 FEET 500 METERS in o DC < O Qk OUTWASH- stratified sand and gravel LACUSTRINE DEPOSITS- stratified very fine sand, silt, and clay KAME DEPOSITS- hummocks composed of statified, poorly sorted silt, sand, and gravel TILL- poorly sorted clay, silt, sand, and stones SHALE AND SILTSTONE 177 WATER TABLE- water levels measured during May 28-29, 1991 WELL AND WELL NUMBER Figure 7C. Hydrogeologic section C-C'. (Line of section and well locations are shown on pi. ---PAGE BREAK--- West D FEET 1,200 -i 1,150 - 1,100 - 1,050 - 1,000 - 950- East D' 900 West Branch Tioughnioga River 447 projected VERTICAL EXAGGERATION X 20 DATUM IS SEA LEVEL 0 1,000 2,000 FEET QUATERNARY AGE -Qlsc: Qk EXPLANATION OUTWASH- stratified sand and gravel LACUSTRINE DEPOSITS- stratified very fine sand, silt, and clay KAME DEPOSITS- hummocks of poorly sorted silt, sand, and gravel TILL- poorly sorted clay, silt, sand, and stones Shale and siltstone WATER TABLE.- Water levels measured during May 28-29, 1991 188 WELL AND WELL NUMBER Figure 7D. Hydrogeologic section D-D'. (Line of section and well locations shown on pi. 0 500 METERS BEDROCK CONTOUR- Shows altitude of bedrock surface. Contour interval is 50 and 100 feet. Datum is sea level Base from U.S. Geological Survey 1:62,500 series: Cortland (1903) Figure 8. Bedrock-surface altitude in Cortland, N.Y., study area. 14 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- drained into the valleys deposited sand and gravel, silt, and clay on the remaining ice and between the ice and the bedrock hillside. When the ice melted, the sediments atop and next to it collapsed, forming hummocky mounds of ice-contact deposits known as kames and kame terraces. Kames typically consist of poorly sorted silty sand and gravel of moderate perme- ability, but discontinuous zones of well-sorted, highly permeable, coarse sand and gravel are found in some places. Large areas of kames are found on hillsides at Cortland High School in the southeast part of the study area, South Cortland, and in the East Branch Tiough- nioga River Valley (pi. 76° 15' 10' 45' - )NJDAGA (DRTLAN EXPLANATION PROGLACIAL LAKE 5 KILOMETERS ERVAL DATUM IS SEA LEVEL Base from U.S. Geological Survey, 1954, 1:250,000 Figure 9. Location of Valley Heads ice and moraines that dammed drainage in valleys of the Tioghnioga River basin. (Location is shown in fig. Kame deposits on hillsides are typically 10 to 80 ft thick and form minor aquifers where they are saturated; in some places, however, kame deposits may be only seasonally saturated. Kame deposits in the central parts of the valley form confined aquifers where they are overlain by lacustrine very fine sand, silt, and clay. Buried kame deposits are typically 60 to 170 ft thick in the western and eastern parts of the study area (figs. 7B, 7C), and 2 to 30 ft thick in the West Branch Tioughnioga River Valley (fig. 7D). Kame deposits in areas where the confining layer is absent, or pinches out before extending to the edges of the valley, are overlain by outwash sand and gravel (fig. 7C) and form a continuous unconfined aquifer. Valley Heads Moraine System A major standstill of the ice front in the northern part of the Allegheny Plateau of central New York between 14,000 and 14,900 years ago resulted in deposition of large amounts of sediment that formed the Valley Heads Moraine system (Muller and Calkin, 1993). A Valley Heads Moraine formed in the western part of the study area and in valleys north of it (fig. This moraine is called "Valley Heads" because it formed a drainage divide between, and headwater areas of, the southward draining Susquehanna River basin and the northward draining St. Lawrence River basin. The crest of the Valley Heads Moraine in the western part of the study area is characterized by kame and kettle topography (pi. Kettles formed where blocks of ice melted, leaving depressions. The ice side (southwest side) of the moraine is relatively steep and hummocky, whereas the northeast side grades from hummocky terrain to a moderately sloping, pitted outwash plain that, in turn, grades to a sloping, smooth outwash plain. The moraine deposits consist of heterogeneous sediments including coarse kamic sand and gravel in the upper part, and fine-grained sediments, such as till and lacustrine material, in the lower part. Proglacial Lakes and Lacustrine Deposits in Valleys Glaciers were imposing obstructions to streams in northward draining valleys the ice formed dams that impounded water in valleys. These impoundments are known as proglacial lakes. During the retreat of Late Wisconsinan ice in central New York, proglacial lakes occupied large parts of valley systems, including the Tioughnioga River drainage system (fig. Glacial Geology 15 ---PAGE BREAK--- Upland streams and meltwaters from the glacier transported and deposited gravel, sand, silt, and clay into proglacial lakes. The coarse-grained, heavy sediments carried by these streams (sand and gravel) were deposited nearshore to form deltas, whereas the fine, light sediments (very fine sand, silt, and clay) were carried further out into the lake, were they settled to form a lake-bottom deposit (lacustrine unit). This lacustrine unit is found throughout the valley in the study area except where it thins and pinches out along the valley edges (fig. 10). It is as much as 90 ft thick in the eastern part of the study area, 15 ft thick in the northern and central parts, and 170 ft thick in the southwestern part. This unit underlies the upper outwash aquifer and overlies the confined aquifer. A delta was found at the well 341 site (pi. on McLean Road in the central part of the aquifer. The well penetrated a sequence of coarse sediments that graded with increasing depth to fine sediments from 0 to 83 ft below land surface. The sequence is inter- preted as coarse outwash from land surface to 37 ft below land surface that overlies deltaic sediments from 37 to 83 ft below land surface. This sequence, in turn, overlies fine-grained lacustrine sediment. The proglacial-lake outlet was south of Messenger- ville (fig. Elevations of proglacial-lake outlets are useful because they indicate the maximum lake level and, thus, the maximum elevation at which lacustrine deposits can be found. The exact elevation of this outlet is uncertain, however, because the Tioughnioga River eroded much of the sediment plug that formed the outlet channel. The lake's water level can be estimated from other evidence, however. For example, the absence of beaches and hanging deltas that would be expected to flank hillsides along the valleys suggest that the highest water level was lower than the present valley-floor elevation (about 1,100 ft). The highest elevation at which lacustrine deposits were found at drilling sites in the study was 1,090 ft (penetrated by well 429 in the southern part of the city); therefore, the maximum lake level was probably between 1,090 and 1,100 ft. Glaciofluvial Sediments (Outwash) of the Valley Heads Episode As the sediment plug that dammed the proglacial lake at Messingerville (fig. 9) was eroded, allowing the proglacial lake to drain to the south in the Tiough- nioga River valley, meltwater from retreating Valley Heads ice throughout the area developed a braided- stream system that deposited large amounts of glacio- fluvial sediments (outwash) on top of the lacustrine sediments. The outwash forms a wedge-shaped deposit that is more than 100 ft thick near the head of outwash in the western part of the study area (fig. 7B, geologic section B-B') and thins eastward to 35 to 45 ft thick in the southern and eastern parts of Cortland (fig. 7C, geologic sectionC-C'). Outwash in the study area grades from coarse boulder gravel at the head of outwash in the southwestern part to coarse cobble and pebbly, sand and gravel in distal reaches in the central and eastern parts of the study area. The outwash, along with local deltaic deposits, forms most of the highly permeable unconfined aquifer in the study area (fig. 11). Drilling with percussion-tool and air-rotary drilling rigs is difficult near the head of outwash, where boulder gravel is present, and auger- ing is nearly impossible. The large amount of very coarse sediments that form the outwash in the study area indicate that large discharges of fast-flowing meltwaters characterized the period of the Valley Heads standstill. Ice Readvance and Drainage of Proglacial Lake During Late Stages of Deglaciation A discontinuous 10- to 20-ft-thick layer of till that overlies the morainal and outwash deposits in the western part of the study area indicates that ice readvanced or surged northeastward to about Stupke Pond during the late stages of the Valley Heads episode. The till is typically found 10 to 30 ft below land surface and is overlain by 10 to 30 ft of sand and gravel deposited by subsequent meltwater. The final glacial event to affect the study area was the draining of a proglacial lake that formed in the Fall Creek valley between the retreating ice front and the western side of the Valley Heads Moraine. The lake drained eastward through an outlet on the moraine at an elevation of about 1,195 ft and eroded a 200-ft- wide, 10- to 20-ft deep channel in the moraine and created a widening swath through the outwash in the western and central parts of the study area (pi. This erosion of outwash left 10- to 20-ft-high cutbanks (pi. 1) along the channel, and outwash terraces above it. Test boring 345, in the middle of the channel, encoun- tered boulders the size of bowling balls to depth of 20 ft below land surface, indicating that large volumes of fast-moving water had flowed through the channel. 16 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- 76° 15' Branch Tioughnioga River East Branch Tioughnioga River EXPLANATION EXTENT OF LACUSTRINE UNIT STRUCTURE CONTOUR- Shows altitude of top of lacustrine unit. Contour interval 10 feet. Datum is sea level BOUNDARY OF GLACIAL AQUIFER DATA POINT 0 0.5 1.0 KILOMETERS Base from U.S.Geological Survey 1:62,500 series: Cortland (1903) Figure 10. Upper surface altitude of lacustrine unit in Cortland, N.Y., study area. Hydrology of the Glacial-Aquifer System The aquifer system in the study area is part of a large regional glacial-aquifer system that occupies major valleys in the Tioughnioga River basin (fig. 11). The regional aquifer system is bounded laterally by till-covered bedrock hillsides (pi. 1) and beneath by the bedrock valley floor. The aquifer system in the study area consists of a 40-to 80-ft-thick unconfined sand and gravel aquifer that overlies a 1- to 155-ft- thick lacustrine confining layer that, in turn, overlies a 1- to 170-ft-thick confined sand and gravel aquifer (fig. 12). The base of the confined aquifer is the till and bedrock. Although the confining unit impedes ground- water movement between the upper and lower aquifers in the middle of the valley, the two aquifers are in hydraulic contact where the confining layer is absent in many places along the valley walls. The unconfined sand and gravel aquifer in the study area is one of the most productive aquifers in New York State and has been designated as a "Primary Aquifer" by the New York State Department of Environmental Conserva- tion, and as a "Sole Source Aquifer" by the U.S. Environmental Protection Agency. Hydrology of the Glacial Aquifer System 17 ---PAGE BREAK--- 76° 15' ONDAG ~R~fLAN EXPLANATION OUTWASH-Dashesjndicate direction of meltwater flow 30'- Base from U.S. Geological Survey, 1954,1:250,000 Figure 11. Positions of Valley Heads ice, and of outwash that was deposited in valleys of the Tioughnioga River basin. (Location is shown in fig. Geometry of Aquifers and Confining Unit The unconfined aquifer forms a wedge-shaped deposit that thins and slopes downward to the north- east. The saturated thickness of the unconfined aquifer (from the water table to the top of the underlying lacustrine confining unit) is as much as 80 ft in the western part of the study area and thins northeastward to about 40 ft at the Tioughnioga River in the eastern part. The unconfined aquifer is underlain in most places by the lacustrine unit, which has a flat or gently northeastward sloping surface. The lacustrine unit, which forms a confining unit that separates the unconfined aquifer from the confined aquifer (fig. 7, geologic sections A-A', B-B', C-C, and is an extensive lens-shaped deposit that consists of interbedded very fine sand, silt, and clay. It is typically 60 to 140 ft thick but is much as 155 ft thick in the middle of the valley. It thins toward the valley walls, where it pinches out. It typically lies 100 to 130 ft below land surface in the western part of the study area, 55 to 85 ft in the central part, and 35 to 65 ft in the eastern part. This unit extends beyond the study area in all valleys (fig. 10). Beneath the confining unit is the confined sand and gravel aquifer, which is found everywhere except in the middle of the valley in the western part of the study area, where several well records indicate the lacustrine deposit to lie directly over bedrock. The bottom of the confined aquifer is the bedrock valley floor or, in some places, the top of till that overlies bedrock. The top of the confined aquifer has an undulating surface that is typical of hummocky kames and(or) bead-shaped glaciofluvial deltas that form the aquifer. The confined aquifer ranges from 1 to 170 ft thick. Hydraulic Conductivity Hydraulic conductivity is a measure of the ability of deposits to transmit water. Well-sorted, coarse- grained sediments have high hydraulic conductivity because they have many large interconnected pore spaces through which water can flow, whereas fine- grained sediments have low hydraulic conductivity because they have fewer and smaller interconnected pore spaces. Although clay may be well sorted and can have high porosity, it has the lowest hydraulic conduc- tivity because the pores are so small that molecular attraction between clay particles and water prevents significant movement of water. Hydraulic conductivity of well-sorted sand and gravel typically exceeds 500 ft/d; that of medium to coarse sand and moder- ately sorted, fine sandy gravel typically ranges from 25 to 500 ft/d; that of fine sand and poorly sorted silty gravel typically ranges from 1 to 25 ft/d; and that of silt, clay, and till typically is less than 1 ft/d (Heath, 1983). Hydraulic conductivity of outwash can differ widely from place to place because the complex depositional processes in meltwater environments resulted in a heterogeneous mixture of particle sizes. As stream channels were abandoned, buried, moved 18 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- laterally, aggraded, or downcut, the outwash materials became mixed; the particle size in a channel reflects the flow velocity of meltwater discharging from the ice and also the place of deposition within the channel, such as, inside a meander bend, where the velocity is much lower than along the outside of the bend. Wells that pump large amounts of water are preferred for aquifer tests because tests at these wells affect a relatively large volume of the aquifer and tend to average the effects of local anomalies within the aquifer, whereas slug tests and aquifer tests that remove only small amounts of water affect a relatively small volume of aquifer and are, therefore, more likely to tap a zone that might not be representative of the aquifer. Hydraulic-conductivity values of aquifers in the study area, as determined from aquifer tests that used large pumping rates, are given in table 2. Unconfined Aquifer Hydraulic-conductivity values of the unconfined aquifer in the study area, as determined from aquifer tests that used large pumping rates, ranged from 85 to 1,150 ft/d (table The largest values (880 to 1,150 ft/d) were in the unconfined outwash deposits in the western and central parts of the study area, which were close to the ice front during the Valley Heads standstill, where fast-flowing meltwater deposited well-sorted coarse gravel and washed away most of the fine-grained sediment, and redeposited it in more distal reaches. Moderately high values (220 to 380 ft/d) were measured in distal reaches of uncon- fined outwash aquifer in the eastern part of the study area, where meltwater velocity was lower than that near the ice front and resulted in deposition of finer, less sorted sand and gravel. A moderately low value Cosmos Hill West Branch East Branch \ Till Bedrock Not to scale Figure 12. Generalized hydrogeologic framework of the glacial-aquifer system in the Cortland, N.Y., study area. Hydrology of the Glacial Aquifer System 19 ---PAGE BREAK--- Table 2. Results of aquifer tests in the glacial-aquifer system in the Cortland, N.Y. study area. [Well locations are shown in pi. 1, gal/min = gallons per minute, ft/d = feet per day.] USGS well number Location 317 11 355 432 434 386 447 435 151 409 City of Cortland Barry School Town of Cortlandville (Terrace Road) Town of Cortlandville (Lime Hollow Road) Smith Corona Corporation ETL Testing Laboratories Rosen Superfund site Monarch Tool Corporation Tunison Fish Hatchery Tunison Fish Hatchery Date 9-12-75 4-26-76 3-15-76 7-16-91 12-26-90 9-11-80 1-19-95 8-4-67 3-12-68 8-14-62 Aquifer Unconfined (outwash) Unconfined (outwash) Unconfined (outwash) Unconfined (outwash) Unconfined (outwash) Unconfined (outwash) Unconfined (outwash) Unconfined (kame) Confined (kame) Confined (kame) Average Average pumping rate hydraulic (gal/min) conductivity (ft/d) 3,000 1,200 630 1,430 975 350 80 150 125 82 l,150a l,050a 950a 880b 900C 350d 220e 85f 60a 65a a Cosner and Harsh, 1978 b Determined during this study c O'Brien and Gere Engineers, Inc., 1990 d Reynolds, 1985 e Blasland, Bouck and Lee, Engineers, 1992 f Apfel, 1967 (85 ft/d) was determined from an aquifer test at an industrial well installed in unconfined kame deposits in the western part of the unconfined aquifer; this reflects the poorly sorted silty sand and gravel that is typical of kame material. The gradation from high hydraulic conductivity in the western part of the study area to moderately high values in the eastern part is consistent with the observed northeastward decrease in grain size of sediment samples collected during test drilling. Coarse cobble gravel predominates in the western and central parts, whereas coarse sand and gravel is common in the eastern part. The northeastward fining of sediments reflects the increasing distance from the source of sediments at the ice front, and the decrease in the gradients of meltwater channels (from 40 ft/mi in the western part to 20 ft/mi in the eastern part). It also reflects a decrease in bedload particle-size through mechanical erosion during transport. Moderately high hydraulic conductivity values (260 to 380 ft/d) were determined by three diffusivity analyses (transmissivity divided by storage coeffi- cient), and one specific capacity analysis, in the unconfined outwash deposits along the Tioughnioga River in the eastern part of the study area by Reynolds (1985,1987). Low to moderately low values (3.4 to 140 ft/d) were obtained by slug tests in small-diameter monitoring wells installed in the unconfined aquifer near the valley wall at the Rosen Superfund site (pi. 1) (Blasland, Bouck and Lee, Engineers, 1992). These low values probably result from the mixing along the valley walls of poorly sorted, silty sand and gravel inwash from uplands with well-sorted outwash. Confined Aquifer Low- to moderately low hydraulic conductivity values (16 to 65 ft/d) were obtained for the confined aquifer from two aquifer tests at the fish hatchery in the southwestern part of the study area and one slug test at the Rosen Superfund site (fig. Hydraulic conductivity values determined from aquifer tests at the fish hatchery ranged from 60 to 65 ft/d, and the value determined by a slug test at the Rosen Superfund site was 16 ft/d. These aquifer tests probably repre- sented poorly sorted kame deposits. Confining Unit A hydraulic conductivity value of 2.0 ft/d was obtained from analysis of a slug test of a 2-in.-diame- ter monitoring well installed in the confining unit at the Rosen Superfund site (Blasland, Bouck and Lee, Engineers, 1992). Hydraulic conductivity values of the confining layer were estimated in other locations from grain-size characteristics observed in split-spoon samples and from air-rotary and cable-tool drill cuttings. A value of 2.0 ft/d was estimated for the confining unit where it consists of very fine to fine sand and silt; this value also was estimated for a similar fine-grained deposit in New Hampshire 20 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- (Tepper and others, 1990). A value of 0.1 ft/d is estimated for areas where the confining layer consists of silt and clay. Ground-Water Levels and Flowpaths Water-level measurements can be used to define the slope of the water table in the unconfined aquifer and of the potentiometric surface in the confined aquifer. The slope determines the direction of ground- water flow. Water levels in about 100 wells were measured during March 28-29,1990; from May 28- June 4,1991; and October 7-9,1991, to obtain values representing periods of high, average, and low recharge, respectively (pi. 2-4). These measurements were used to calibrate the ground-water flow model, as discussed further on. Ground-water levels also were measured in about 50 wells from July 1989 through October 1991 to determine annual water-level fluctuations. The water-level measurements made during this study indicated a hydraulic head loss of about 100 ft over the 6-mi length of the study area, from the ground-water divide in the western part to the Tioughnioga River in the east. Ground water moves from areas of high head to areas of low head, and the directions of flow depend on local factors such as aquifer geometry, distribution of recharge, and location of discharge areas (streams, ponds, wetlands, and pumping wells). These factors also determine the altitude and slope of the water table (or of the potentiometric surface in a confined aquifer). Ground water flows roughly perpendicular to the potentiometric contours (lines of equal head), as shown in plates 2-4. Unconfined Aquifer Water in the unconfined aquifer generally moves laterally from the edges of the valley toward the center, then northeastward along the axis of the Otter Creek-Dry Creek valley and discharges to pumping wells and to the West Branch Tioughnioga and the Tioughnioga Rivers. The study area contains four major pumping centers: the City of Cortland well field, the well field for the Town of Cortlandville on Terrace Road, a well for the Town of Cort- landville on Lime Hollow Road, and the purge well at the typewriter plant in the western part of the study area (pi. 2-4). The hydraulic gradient in the unconfined aquifer in the middle of the valley typically ranges from 0.0027 ft/ft during periods of low water levels to 0.0037 ft/ft during periods of high water levels. The steeper hydraulic gradients along the edges of the valley in some areas than in the middle probably result from kame deposits of low hydraulic conductivity and(or) proximity to major recharge areas, such as where upland streams flow onto the aquifer and lose water to the coarse sediments. Ground-water flowpaths during periods of high, average, and low recharge shift with the magnitude and distribution of recharge. For example, water infil- trating from upland tributary streams that flow onto the Otter Creek-Dry Creek valley forms a fan-shaped ground-water mound within the unconfined aquifer. Water spreads from the apex of the mound (where the stream enters the edge of the valley) toward middle of the valley where, in effect, it "pushes" the flowpath of ground water in the middle of the valley (the water is moving down the axis of the valley) toward the opposite side of the valley. The extent of the "push effect" depends upon the local geology, the size of the upland tributary basin, and recharge conditions at the opposite valley wall. The median annual-minimum streamflow is highly dependent on the size of a drainage basin and the amount of sand and gravel in it (Thomas, 1966). Drainage basins that are underlain by significant amounts of sand and gravel (more than 20 percent of the basin) have significantly higher streamflow during low-flow periods than basins underlain entirely by till or bedrock. During low-recharge conditions, stream- flow in small upland till and bedrock basins and on unchanneled hillsides is minimal or absent and would provide little or no recharge to the main valley from these streams. Streams in large upland tributary basins that contain significant amounts of sand and gravel tend to be perennial, however, and provide a continu- ous amount of recharge to the aquifer (Thomas, 1966). Confined Aquifer The study area contains only 15 wells that tap the confined aquifer because the unconfined aquifer has adequate yields in most places, making more costly deep wells unnecessary. Therefore, the inferred direc- tions of ground-water movement in the confined aquifer are less certain than those in the unconfined aquifer, which are based on water-level measurements in more than 100 wells. The direction of ground-water flow in the confined aquifer, as inferred from water-level measurements Hydrology of the Glacial Aquifer System 21 ---PAGE BREAK--- made in 15 wells finished in the confined aquifer, generally follows that in the unconfined aquifer; that is, from the edges of the Otter Creek-Dry Creek valley toward the center, then northeastward down the valley axis. Ground water in the West Branch and East Branch Tioughnioga River valleys flows southward and south westward, respectively, then, southeastward as underflow down the Tioughnioga River valley at the confluence with the Otter Creek-Dry Creek valley. Vertical (downward) gradients were measured in some nested piezometers along the edges of the valley where the confining unit is absent. These areas are where water in the unconfined aquifer flows vertically downward to the confined aquifer. Conceptually, water in the confined aquifer may move up into the unconfined aquifer along the edges of the valley in the eastern part of the study area, as indicated by ground-water-flow simulations (discussed in the modeling section). Monitoring wells would need to be installed along the edges of the valley in this area to determine whether such upward flow actually occurs. Sources of Recharge The aquifer system receives recharge from three sources under natural (nonpumping) conditions infiltration of precipitation on the aquifer, upland sources, such as runoff from unchanneled hillsides and seepage from bedrock that border the aquifer, and seepage from tributary streams that flow onto the aquifer (fig. 13). The unconfined aquifer receives additional recharge from infiltration beneath recharge basins at an industrial site in the western part of the aquifer, and induced infiltration from streams and ponds near the major pumping wells. Direct Infiltration of Precipitation on the Valley Part of the precipitation that falls on surficial sand and gravel is returned to the atmosphere by evapotranspiration; the remainder infiltrates and recharges the aquifer. Thus, the amount of recharge by infiltration (Rj) equals precipitation minus evapotranspiration (ET), assuming that ground-water evapotranspiration is negligible: t = P - ET Rates of recharge from precipitation vary season- ally. Most of the precipitation that falls during the dormant period of vegetation (typically from mid- October through the end of April) infiltrates into the aquifer, but little reaches the aquifer during the growing season (May through September) because it is lost through evapotranspiration. In the Northeast, evapotranspiration in any year can be estimated as the long-term average annual precipitation minus long-term average annual stream runoff (Lyford and Cohen, 1987). Average annual runoff for 53 years (1939-91) in the Tioughnioga River basin, measured at a streamflow gage at Cortland (USGS station 01509000, pis. 2-4) was 23.0 in. (Campbell and others, 1992); thus, the average annual evapotranspiration would equal 18.2 in. (long-term average annual precipitation of 41.2 in. minus long- term average annual runoff of 23.0 in.), and the long- term average annual recharge would be 23.0 in. (long- term average annual precipitation of 41.2 in. minus average annual evapotranspiration of 18.2 in.). Upland Sources Although direct infiltration of precipitation on the valley has been considered the chief source of recharge to sand and gravel aquifers, recent studies in the glaciated Northeast indicate that, in regions of moderate to high topographic relief (where hills rise more than 500 ft above the adjacent valley floor), more than half of the natural recharge to surficial strat- ified-drift aquifers in the valleys can be derived from upland runoff (Morrissey and others, 1987). For example, 58 percent or more of the recharge in two glacial aquifers in valleys in central New York was derived from unchanneled runoff from hillsides that border the valley-fill aquifer, and seepage loss from upland tributary streams that flow over the aquifer (Randall and others, 1988). Recharge to valley-fill aquifers from adjacent unchanneled hillsides include surface runoff and the lateral movement of ground water (within upland till, sand and gravel, and bedrock) that flows toward the valley and seeps into the aquifer along the its edges (fig. 13). All precipitation that is not lost through evapotranspiration in the unchanneled drainage areas in the uplands is assumed to become either runoff or ground water that will eventually reach the valley and infiltrate into the aquifer. If a stream is at the base of the hill, however, runoff and ground water from the hillside will discharge into the stream rather than to the aquifer. The amount of recharge from runoff from adjacent unchanneled hillsides annually can be calculated by the following equation: 22 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- = P-ETxDA where R = recharge from runoff from unchanneled hillsides P = precipitation ET = evapotranspiration DA = drainage area of hillside Tributary streams (channeled flow) that drain till- and-bedrock basins in the uplands are major sources of aquifer recharge where they enter and flow over the permeable surficial sand and gravel in the main valley. Small tributaries that flow onto the aquifer typically go dry, especially during the summer, when flow in streams is small and the water table in the valley-fill aquifer falls below the streambed, allowing water in the stream to seep below the channel. The rate of recharge from losing tributaries is controlled largely by the vertical hydraulic conductivity and thickness of the streambed and by the hydraulic heads in the stream and aquifer. Rates of recharge from tributaries were calculated from several sets of streamflow measure- ments made during periods of high, average, and low recharge in 1990-91 in most tributary streams that flow onto the aquifer (table Discharges at the measuring sites are shown in plates 2 through 4. Most tributary streams in the study area lose water to the aquifer where they flow onto the valley; an exception is Otter Creek, which can be divided into three reaches that have two flow regimes; these are the reach from the point where Otter Creek enters the valley to where it joins the outlet to Stupke Pond, the reach from Stupke Pond to the umlaufberg, and the reach where Otter Creek flows north of the umlaufberg to its mouth. Reaches 1 and 3 lose water during the entire year; and reach 2 typically gains water during high- recharge conditions but loses water during periods of average and low recharge. Reach 3 is hydraulically connected to the aquifer during most of the year, except when the water table is lower than the streambed in the summer, causing Otter Creek to dry up. Otter Creek can be considered the main trunk stream in the Otter Creek-Dry Creek valley because it has the only reach in the valley that gains water, even though for only part of the year. Only part of the water in upland tributaries recharges the aquifer during periods of high stream- flow (during spring and during large storms through- out the year); the remainder flows over the aquifer and discharges into the West Branch Tioughnioga and Tioughnioga Rivers. During average- and low-flow conditions (summer, fall, and winter), however, all of the water in most tributary streams recharges the aquifer, except in Dry Creek, which goes dry only during extended droughts. As the water table declines through the summer, losing streams typically dry up in the upstream direction, starting at the mouth, whereas gaining streams typically dry up in the direction, starting at the headwaters. Table 3. Streamflow losses and gains in tributary streams above unconfined aquifer in the Cortland study area, 1990-91. [Values are in cubic feet per second; location of tributaries and streamflow-measurement sites are shown in pis. 2-4.] Streamflow loss or gain between measurement sites near valley wall and near mouth, or where streams dry up March 29, 1990 Tributary stream (high recharge) UPPER REACH OF OTTER CREEK (from valley wall to outlet to Stupke Pond) Dry Creek Perplexity Creek Tributary A Tributary B Tributary C Tributary D Tributary E Estimated loss from other tributaries not measured Sum of tributary losses LOWER REACH OF OTTER CREEK (Stupke Pond to mouth) -1.20 -0.81 -0.65 -0.95 -0.40 -0.41 -2.10 -1.22 -0.50 -8.24 +8.60 May 30-June 4, 1991 (average recharge) -0.4] -0.54 -0.20 -0.12 -0.12 Dry at valley wall -0.23 -0.11 -0.05 -2.73 -0.42 October 9, 1 991 (low recharge) Dry at valley wall -0.37 Dry at valley wall Dry at valley wall Dry at valley wall Dry at valley wall Dry at valley wall Dry at valley wall Dry at valley wall -0.37 Entire reach was dry Hydrology of the Glacial Aquifer System 23 ---PAGE BREAK--- Large upland tributary EXPLANATION SAND AND GRAVEL LACUSTRINE FINE SAND, SILT, AND CLAY TILL BEDROCK Reach of tributary stream that gains water from ground-water discharge ' Reach of tributary stream that loses water to aquifer SOURCES OF RECHARGE TO SAND AND GRAVEL AQUIFERS IN LARGE VALLEYS Infiltration of precipitation that falls directly over the valley Unchanneled runoff and (or) ground-water flow from hillsides Regional ground-water flow through bedrock and till Infiltration from tributary stream (see letters A and B) Recharge that moves laterally to river Recharge that moves downward to confined aquifer Figure 13. Sources of recharge to stratified drift in valleys in the glaciated Northeast. (Modified from Morrissey and others, 1987, fig. 24 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- Induced Infiltration by Ground-Water Withdrawals Pumping affects Otter Creek where it flows through the City of Cortland well field because that reach of the creek is hydraulically connected to the aquifer (pi. 2-4). Pumping lowers the water table beneath the stream surface and thereby induces infil- tration from the stream to the aquifer. Pumping of a City of Cortland municipal well during low-flow conditions once caused a nearby reach of Otter Creek to dry up, and when the pump was turned off, water reappeared in the creek (James Roberts, former manager of Cortland Water Department, oral commun., 1990). The rate of induced infiltration for a pumping rate of 4.29 Mgal/d, as estimated from results of numerical model simulation of average-conditions, is 0.6 ft3/s. No induced infiltration can occur when Otter Creek dries up. The municipal well for the Town of Cortlandville on Lime Hollow Road may induce some infiltration from a pond 400 ft south of the well. The similarity of water levels in the pond to those in the municipal well suggests that the pond is hydraulically connected to the aquifer. Infiltration From Recharge Basins A recovery well (well 434, pi. 1) at the typewriter plant in the western part of the study area is pumped at a rate of about 980 gal/min. The water is routed through an air stripper, where most volatile-organic chemicals are removed, then discharged to recharge basins about 750 ft north of the plant (pi. 2-4), where it infiltrates back to the aquifer. Ground-Water Discharge Water in the unconfined aquifer discharges as seepage into major streams and, seasonally, to some reaches of Otter Creek; to pumping wells; and as seepage to springs that are headwaters to the Fall Creek valley in the western part of the study area. Most of the water pumped from municipal wells is eventually piped to the sewage-treatment plant as wastewater, where it is treated and discharged into the Tioughnioga River in the eastern part of the study area (pi. Most water in the confined aquifer leaves the study area as underflow in the Tiough- nioga River valley in the southeastern part of the study area. Seepage to Streams Most of the ground water that discharges to streams enters the West Branch Tioughnioga and Tioughnioga Rivers; lesser amounts enter the reach of Otter Creek from the Stupke Pond outlet to the umlauf- berg in the central part of the study area. The confining unit prevents most water in the confined aquifer from flowing upward into the unconfined aquifer and into the streams. Measurements of streamflow in the West Branch Tioughnioga and Tioughnioga Rivers and in Otter Creek (fig. 14) indicate that total ground-water discharge to these streams was 17.8 ft3/s for average- recharge conditions and 8.4 ft3/s for low-recharge conditions (table measurements in these large streams during periods of high recharge were impracti- cal (for safety reasons). Table 4. Ground-water discharge to West Branch Tioughnioga River, Tioughnioga River, and Otter Creek during average- and low-flow conditions, 1991. [All values are in cubic feet per second. Location of measurement sites and reaches shown in fig. 14]. Average-flow conditions Low-flow conditions Reach Site no. Net gain within reach Measured (upstream value minus discharge value) Net gain within reach Measured (upstream value minus discharge value) A. West Branch Tioughnioga River A and B . Tioughnioga River minus discharge from sewage treatment pland B-E. Otter Creek 1 2 1 3 subtotal 3 7 31.4 40.6 83.2 99.2 -8.4 90.8 1.4 2.4 TOTALS 9.2 7.6 LQ 17.8 12.8 18.7 53.3 64.4 -8.6 55.7 0.0 0.0 5.9 2.5 OQ 8.4 Hydrology of the Glacial Aquifer System 25 ---PAGE BREAK--- Municipal and Industrial Withdrawals Total ground-water withdrawal from the glacial- aquifer system by municipal and industrial wells ranges from 6.76 to 7.20 Mgal/d (table most of which is from the unconfined aquifer. The largest user is the City of Cortland, which, during 1984-92, pumped 3.9 to 4.3 Mgal/d (data from records at the Cortland Water Department). Substantial leaks in the distribution system (Douglas Withey, manager of Cortland Water Department, oral commun., 1994) result in conveyance losses that return some of the water to the aquifer; therefore, actual discharge from pumping is somewhat less than the reported values. The Town of Cortlandville pumped from 0.65 to 0.99 Mgal/d of water from its two well fields during 1990-91 (Hayne Smith, Town of Cortlandville Engineer, oral commun., 1992). A purge well at the typewriter plant pumped at a rate of 0.69 Mgal/d from December 1989 through October 1990, and at a rate of 1.43 Mgal/d from October 1990 to at least December 1994. All pumped water from the purge well is returned to the aquifer through recharge basins 750 ft north of the plant. When the purge well was pumped continuously (since October 1990),the other produc- tion wells for the plant were not used. Underflow From the Study Area Within the confined aquifer, water that is northeast of the ground-water divide in the western part of the study area flows northeastward and leaves the study area as underflow through the Tioughnioga River valley at an estimated rate of 40,500 ft3/d (0.5 ft3/s), as calculated from Darcy's equation for one-dimensional flow in a prism of porous material: Q = where Q = flow (LV1) K = hydraulic conductivity of the aquifer (Lr A = cross-sectional area perpendicular to flow (L2); h2-hj = head difference across the prism of flow L = length of flowpam The following hydraulic values were used to calculate underflow out of the Tioughnioga River valley: hydraulic conductivity 150 ft/d cross-sectional area of valley 135,000 ft2 head difference 0.002 ft length 1 ft. Ground water that is west of the divide in the western part of the study area flows southwestward and leaves the study area as underflow to the Fall Creek valley, also at an estimated rate of about 40,500 ft3/d, or 0.5 ft3/s. WATER QUALITY Water from 15 wells was sampled about by the Cortland State College and Cortland County Department of Planning during 1987-93. In addition, the USGS and these two groups conducted three synoptic rounds of sampling to define the extent of a trichloroemylene (TCE) plume that resulted from a spill at the typewriter plant in the western part of the study area; assess the contamination-remediation efforts at the source (the plant site); and compare current concentrations of common ions, metals, and nutrients in ground water throughout the study area with historical data to detect changes or trends in water quality over time. TCE was used at the typewriter plant for degreasing mechanical parts from 1958 through October 1986 (O'Brien and Gere Engineers, Inc., 1987). Results of chemical analyses are given in appendix 3. The first round of sampling (April 4-5, 1990, a high-recharge period), entailed collection of samples from 60 wells (31 observation wells, 26 private wells, and 3 municipal wells). All samples were analyzed for common ions, metals, nutrients, and volatile-organic compounds (VOC's). The second round (September 17-20, 1990, a low-recharge period), entailed collec- tion of samples from 31 of the 60 wells sampled previ- ously to verify results from the first sampling, and from an additional 9 observation wells that were installed after the first sampling to obtain data needed to define the extent of the TCE plume. Of the 40 samples collected during the second round, all were analyzed for nutrients, and 30 were analyzed for common ions, metals, and VOCs. The third round of sampling, 3 years later (April 27, 1993), entailed collection and analysis of samples from 18 wells for TCE and its degradation products to determine whether remedial efforts at the typewriter plant site had improved water quality downgradient of the site. 26 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- 76"15' 42'37' Branch Tioughnioga River East Branch Tioughnioga River Tributary to Perplexity Creek EXPLANATION STUDY AREA A 2 STREAMFLOW-MEASUREMENTSITES- 4 Numbers are measurement-site numbers; letters are reaches of streams listed in table 4 BOUNDARY OF GLACIAL AQUIFER - - - - GROUND-WATER DIVIDE 0.5 1.0 MILES I I 1 Unnamed , tributary 0 0.5 1.0 KILOMETERS Base from U.S.Geological Survey 1:62,500 series: Cortland (1903) Figure 14. Location of streamflow-measurement sites in Cortland, N.Y., study area. Table 5. Ground-water withdrawals from the Cortland, N.Y. study area by major pumping wells during low-, average-, and high-recharge conditions, 1990-91. [All values are in millions of gallons per day. Locations of wells are shown on pi. Recharge conditions and dates of measurement Well number Owner 317 City of Cortland 354, 355 Town of Cortlandville, Terrace Road 432 Town of Cortlandville, Lime Hollow Road 434 Smith Corona Corp., recovery well 392 Smith Corona Corp., production wells 386 ETL Testing Labs 373 Cortland Hospital 140, 139 Tunison Fish Hatchery 138 Tunison Fish Hatchery TOTALS Low recharge Average recharge High recharge (Oct7-9, 1991) (May 28-June 4, 1991) (Mar. 28-29, 1990) 4.09 0.56 .16 1.43 Not pumping .22 Not pumping .20 JO 6.76 4.28 0.40 .59 .69 .48 .22 .18 .20 JQ 7.14 4.41 0.65 Not pumping .69 .36 .22 .18 .20 JQ 7.20 Water Quality 27 ---PAGE BREAK--- Volatile Organic Compounds The U.S. chemical industry produces large amounts of volatile organic compounds (VOCs) that are used extensively by industries and by homeowners. Large-scale production and use of VOCs has resulted in many instances of soil- and ground-water contamination, most of which has resulted from leaks or spills at manufacturing plants, petroleum-product storage tanks, and chemical- waste disposal sites (Plumb, 1991). VOCs are low- molecular-weight hydrocarbons having very low (less than 2 percent) solubility in water, and rapid volatization. VOCs on the USEPA list of "Priority Pollutants" (U.S. Environmental Protection Agency, 1991) include trichloroethylene (TCE), benzene, 1,1,1 -trichloroethane, and cis- 1,2-dichloroethylene. Each is toxic to some degree, and most, including TCE, are believed to be carcinogenic (National Academy of Sciences, 1977). The USEPA "Maximum Contaminant Levels" (MCLs) for drinking water supplied by municipal water systems is 5 jig/L for TCE and benzene; 200 jig/L for 1,1,1-trichloroethane, and 70 ^ig/L for cis-1,2- dichloroethylene (U.S. Environmental Protection Agency, 1991). The MCL is an enforceable, health- based regulation set by the USEPA. Many VOCs, including many of the chlorinated hydrocarbons (CHCs), are solvents used in degreasers and cleaners for metal and electronic parts, and also in paint removers, dry-cleaning fluids, drain cleaners, spot removers, and septic-tank cleaners. Other VOCs, such as benzene and toluene, are used in fuels. TCE is a CHC and a common industrial solvent that has been in commercial use for about 40 years. CHCs are persistent in the environment because they resist chemical and biological degradation; the average half-life for abiotic CHC transformations range from 2 months to 1010 years, and CHC half-lives for bio- degradation range from 2 weeks to 8 months (Barbee, 1994). Anaerobic reductive dehalogenation (typically dechlorination) of CHCs is the primary biodegra- dation process in ground water; for example, cis-l ,2-dichloroethylene is a reductive dechlorination product of TCE (Barrio-Lage and others, 1986; Wilson and others, 1991). Extent and Migration of Trichloroethylene Contamination in the Unconfined Aquifer The Chemistry Department at SUNY College at Cortland detected TCE in samples from a well in the central part of the unconfined aquifer in 1986. The well is 0.25 mi north of the Cortlandville municipal well field and 1.25 mi southwest and upgradient of the Cortland well field, and City and Town water manag- ers and Cortland County health officials were concerned that the TCE could migrate to drinking- water supplies. In December 1986, samples were taken from about 23 observation wells and domestic wells by the college Chemistry Department and the Cortland County Department of Health to determine the extent and source of VOC contamination. Most of these wells were resampled about during 1986-93. Most samples were analyzed by the college Chemistry Department; the rest were analyzed by the USGS laboratory in Denver, Co., and Buck Laboratory in Cortland. From December 1986 through 1993, TCE was found in concentrations above the MCL of 5 jig/L at many wells in the central part of the unconfined aquifer. The highest TCE concentration found beyond the typewriter plant was 222 jig/L, in well 350 (owner Pace) on Pheasant Run (pi. 1) on April 9,1987. This well 350 is 1,000 ft downgradient from the TCE source area at the plant, cis- 1,2-Dichloroethylene, a solvent and reductive dehalogenation product of TCE, was also found in off site wells where TCE was detected, but concentrations were below the MCL of 70 jig/L. Horizontal Migration Analyses of samples collected offsite by the Cortland County Department of Health and the college Chemistry Department, and of samples collected on the typewriter plant site by O'Brien and Gere Engin- eers, Inc. (1987), indicated that TCE in ground water was migrating northeastward. Hydrogeological consultants working for the plant found much higher concentrations of TCE and cw-l,2-dichloroethylene in wells at the plant site than in offsite wells and identified TCE-contaminated soils near the northwest side of the main plant building (O'Brien and Gere Engineers, Inc., 1987). The highest TCE concentrations found were at well site 448 (local well no. MW-7), installed by the consultants along the northwest perimeter of the main plant building (pi. where TCE and cis-1,2-dichloroethylene concentrations of 10,000 and 7,600 [ig/L, respectively, were found on August 3,1989. TCE concentrations that day in 19 other wells on the site ranged from <1 to 270 ^ig/L,andci5-l,2-dichloroethylene concentrations ranged from <1.0 to 1,700 [ig/L (O'Brien and Gere Engineers, Inc., 1990). 28 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- The extent of the TCE plume during the three rounds of sampling is shown in figures 15A, B, and C, respectively; the plume is defined where concentra- tions exceeded the MCL of 5 jig/L. The extent of the plume was roughly the same during all three sampling rounds, indicating that steady-state conditions proba- bly had been reached. The plume extends 1.25 mi northeast from the plant and follows head gradients and the direction of ground-water flow. Vertical Migration The vertical distribution of TCE in the unconfined aquifer at wells 341 and 381 is plotted in figure 16. The TCE concentrations at well 341 (fig. 16A), which is 4,000 ft northeast of the TCE source at the plant, are higher in the middle zone than in the lower and upper zones of the unconfined aquifer, but the TCE concen- tration at well 381, which is 750 ft northeast of the TCE source, was uniformly distributed with depth (fig. 16B). The vertical distribution of TCE at well 341 (fig. 16A) is likely the result of vertical dispersion of TCE with increasing distance from the source; the vertical distribution of hydraulic conductivity in the aquifer, and dilution by recharge (from precipita- tion). Coarse-grained sediments that form the top and middle zones of the aquifer and have high hydraulic conductivity provide a preferred pathway for ground- water flow and dissolved contaminants, but the fine- grained sediments of lower hydraulic conductivity in the lower zone retard downward ground-water movement and are a less preferred pathway. Little or no TCE was detected in the confined aquifer, which is protected by the lacustrine confining unit, which prevents the movement of water in the unconfined aquifer into the confined aquifer. Effects of Remediation Efforts A recovery well was installed at the typewriter plant as part of a "pump and treat" method for onsite remediation. Contaminated water is pumped from the aquifer and routed through an air stripper, where VOCs are volatilized and released into the atmosphere. The treated water is then piped 750 ft north of the plant and discharged into recharge basins, where it infil- trates to the water table. The return of pumped water to the aquifer causes little or no net loss of ground water. A cone of depression that has formed around the recovery well, and a ground-water mound that has formed beneath the recharge basins, alters the ground- water flow paths in the vicinity of the plant. The effects of pumping by the TCE-recovery well and the rate of ground-water flow are essential consid- erations in evaluations of TCE concentrations in the plume. If the pump-and-treat system keeps the contaminated water from leaving the site, concentra- tions of VOCs will decrease over time through degra- dation and dispersion, and through downvalley movement of the tail of the plume that developed before pumping . The recovery well was pumped intermittently from December 1989 through October 1990 at a rate of about 980 gal/min for 12 to 18 hours daily and contin- uously at this rate since October 1990. The TCE concentrations in most wells that were sampled in April 1990 (110 days after the beginning of intermit- tent pumping) were probably not significantly affected by the pumping because all wells that were sampled, except for well 22 (local well number CT-22), were more than 1,000 ft from the source of TCE- the distance ground water would move in 110 days at a velocity of 9 ft/d (velocity = hydraulic conductivity of 1,000 ft/d, divided by porosity of 0.3, times the hydraulic gradient of 0.0027 ft/ft). The highest TCE concentrations found during the April 1990 sampling were just above 100 [ig/L at well 350 (owner Pace) and well 121 (local well number CT- 21), which are 1,000 and 2,400 ft northeast of the source of TCE, respectively (fig. 15A and pi. These two wells have had the highest TCE concentrations of any off site wells. Pumping of the recovery well might have affected TCE concentrations at wells as far as 0.5 mi (2,640 ft) northeast of the source by the time of the second sampling round (September 17-20,1990), as calcu- lated for a pumping duration of 288 days (9.5 months) and a ground-water flow rate of 9 ft/d. TCE concentra- tions beyond this distance from the source were proba- bly not affected yet. The extent of the TCE plume in September 1990 was about the same as in April 1990, but TCE concen- trations were typically 30 to 50 percent lower; this could partly reflect the tendency of TCE concentra- tions at wells not yet affected by the recovery well to be lower during low-recharge periods (summer and fall) than during high-recharge periods (spring), as explained further on. By the time of the third sampling round, 3 years later (in April 1993), the recovery well had been operating for 1,186 days (3.25 years), the last 2.5 years of which it had pumped continuously. During those 2.5 Water Quality 29 ---PAGE BREAK--- 76° 14' 13' 42°36' 35' 34' CORTLAN MUNICIPAL, A1RP CHAS EXPLANATION LINE OF EQUAL TRICHLOROETHYLENE CONCENTRATION- Shows trichloro- 5 ethylene concentration in micrograms per liter. Samples collected during April 3-5, 1990 AQUIFER BOUNDARY 65 WELL SITE AND TRICHLOROETHYLENE CONCENTRATION, in micrograms per liter SURFACE-WATER SITE AND TRICHLOROETHYLENE CONCENTRATION, in micrograms per liter PUBLIC-WATER SUPPLY WELL AND TRICHLOROETHYLENE CONCEN- TRATIONS, in micrograms per liter LESS THAN Base from New York State Department of Transportation 1:24,000 series: Cortland, New York (1980) Figure 15A. Extent of trichloroethylene plume and lines of equal trichloroethylene concentration in study area, April 3-5, 1990. 30 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- 76° 14' 42°36' 35' CORTLANB'-CQ EXPLANATION LINE OF EQUAL TRICHLOROETHYLENE CONCENTRATION- Shows trichloro- 5 ethylene concentration in micrograms per liter. Samples collected during September 17-20, 1990 AQUIFER BOUNDARY 65 WELL SITE AND TRICHLOROETHYLENE CONCENTRATION, in micrograms per liter PUBLIC-WATER SUPPLY WELL AND TRICHLOROETHYLENE CONCEN- TRATIONS, in micrograms per liter 34' Base from NewYork State Department of Transportation 1:24,000 series: Cortland, NewYork (1980) Figure 15B. Extent of trichloroethylene plume and lines of equal trichloroethylene concentration in study area. September 17-20, 1990. Volatile Organic Compounds 31 ---PAGE BREAK--- 76° 14' 42°36' EXPLANATION LINE OF EQUAL TRICHLOROETHYLENE CONCENTRATION- Shows trichloro- _5 ethylene concentration in micrograms r per liter. Samples collected during April 27, 1993 AQUIFER BOUNDARY 65 WELL SITE AND TRICHLOROETHYLENE * CONCENTRATION, in micrograms per liter alllm \tf S ^ rfv^^ [ / ,000 FEET IT [ . PUBLIC-WATER SUPPLY WELL LESS THAN Base from New York State Department of Transportation 1:24,000 series: Cortland, New York (1980) Figure 15C. Extent of trichloroethylene plume and lines of equal trichloroethylene concentration in study area, April 27, 1993. 32 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- years, ground water would had moved more than 1.5 mi from the source-a distance greater than the length of the plume (about 1.25 mi). Therefore, water at all wells sampled during April 1993 would have been affected by the recovery-well operation. The "pump and treat" remedial program was expected to capture all contaminated ground water on the site; TCE concentrations at the site were expected to decline rapidly, and those offsite were expected to decline significantly. These expectations were bolstered by analyses of several random samples collected in the fall of 1992, which showed that TCE concentrations had decreased significantly, some to the lowest values since December 1986 (fig. 17). These decreases prior to the April 1993 sampling suggested that the pump-and-treat system at the plant site was effective. A large amount of recharge in the winter and spring of 1993 (fig. which included the blizzard of 1993, reversed this trend of improvement, however. Results of the April 1993 sampling indicated that TCE concentrations at most wells in the middle and distal parts of the plume (figs. 15C, 17B, 17C) had increased significantly since the preceding fall and were nearly as high as those detected during April 1990 (fig. ISA) and September 1990 (fig. 14B). These elevated concentrations could not have been due to additional TCE emanating from the plant area because the ground-water traveltime from there to the middle and distal parts of the plume is from 1.5 to 2.5 years. The elevated TCE concentrations can be explained, however, by desorption of TCE that had previously sorbed onto sediments in the unsaturated zone. The above-normal recharge of March and April 1993 raised the water table to its highest level in at least 10 years and resaturated upper parts of the aquifer that had been A. WELL 341 Geologic description Stratigraphic column 10 20- 30- 40 - 50 - 60 - 70- 80 QO 00- 10- 20- 30- 37 Fill Outwash silty gravel Outwash, interbedded layers of sand and gravel and coarse sand Deltaic sand, grades with increasing depth from very coarse sand to fine sand Lacustrine silty sand depth to sandy silt Pebbly sand Sand and gravel Silty gravel (till?) Sand and gravel 1 1 1 1 1 SsS oooc m& JD '5 Unconfined ac 9 _03 Cc o 0 .CD oT T3 CD _c o O I I I Water table AN / *^ ^^0.7 X K^ ,<0.1 K^ I I u - 20 - 40 L 60 - 80 -100 -120 -140 -160 -180 -200 -220 -240 -PRO 0 10 20 TCE CONCENTRATION, in micrograms per liter Figure 16. Vertical distribution of trichloroethylene (TCE) in ground water at two wells in Cortland, N.Y., study area: A. Well 341, 4,000 feet northeast of typewriter plant site, July 19, 1990. B. Well 381, 750 feet northeast of plant site, September 23-26, 1992. (Well locations shown on pi. Volatile Organic Compounds 33 ---PAGE BREAK--- 240 O 22° < 200 DC DC I LU -100 Z H 18° O ^ 160 g Q- 140 21 ^ 120 d £l 100 80 60 40 20 0 I O fcg DC A. Well 350 (1,000 feet f romTCE source) Intermittent pumping of recovery well 19861 1987 Continuous pumping of recovery well 1988 1989 1990 I 1991 1992 I 1993 Zg < DCI- LLJ OZ O O LLJ Z LLJ 140 £ 120 t _i DC 100 LU CL W 80 DCO 60 O DCI 40 Z ~ 20 0 Zg t DC DC feP S = 8- Ujg LLJ < I O I- DC LU o ° ^ DC ^ g z I g DC B. Well 347 (4,000 feet f romTCE source) 1986 1 1987 50 UMMMIII 1988 1989 1990 I 1991 1992 I 1993 tC.Well 364 (4,700 feet f romTCE source) 1986' 1987 1988 1989 ' 1990 ' 1991 1992 1993 Figure 17. Trichloroethylene concentration at wells 350, 347, and 364 from November 1986 through May 1993. (Locations shown on pi. 34 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- unsaturated and that contained TCE that had not been thoroughly flushed by precipitation. Results of the April 1993 sampling indicated that TCE concentrations had increased in the middle and distal wells but had declined significantly at some wells near the source, such as well 350 (owner Pace, fig. 17A), and at other wells on Pheasant Run (pi. 1) that had the highest TCE concentrations of the offsite wells. The lowered TCE concentrations at these wells could be explained by changes of ground-water flow paths near the plant site as a result of pumping by the recovery well and infiltration of the pumped water in recharge basins. Long-Term Trends In TCE Concentrations The year in which the TCE spill occurred was estimated from the arrival of the front of the plume at well 364 early in 1987 (fig. 17C) and from results of a particle-tracking program (MODPATH) in the back- tracking mode. The program calculated the ground- water traveltime to range from 21 months for high- recharge conditions to 23 months for low-recharge conditions. Subtracting 21 to 23 months from the time when the front of the TCE plume reached well 364 (early in 1987) indicates the time of the spill to be as early 1985. After the front of the plume reached well 364 in early 1987, the TCE concentrations at this well fluctuated only through late 1992 (fig. 17C). The highest concentrations of TCE in most offsite wells since the first sampling began in 1986 were found during 1987-88. Concentrations decreased thereafter from 1989 through mid-1992, then decreased sharply at the end of the 1992, probably in response to removal of contaminated soil, and capture of contaminated ground water by the recovery well. These decreases were followed in the spring of 1993 by increases to about the same levels as those during 1989-91, however. As explained previously, the simultaneous increases of TCE concentrations in middle and distal parts of the plume suggest local releases of TCE to the aquifer, such as through desorp- tion from a diffuse TCE source in the previously unsaturated zone of the aquifer, rather than to new TCE migrating from the source area at the typewriter plant. Desorption of TCE in the unsaturated zone can occur when precipitation infiltrates the soil, and a rising water table saturates the previously unsatur- ated zone. TCE concentrations at wells within 1,500 ft of the typewriter plant, such as well 350 (fig. 17A), showed seasonal fluctuations. TCE concentrations were highest during the spring and lowest during the summer and fall. Hydrogeologic consultants (O'Brien and Gere, Engineers., 1990) also reported seasonal TCE fluctuations in wells at the plant. TCE concentra- tions at wells more than 1,500 ft from the plant showed little fluctuation, however (for example, well 364, about 4,500 ft downgradient of the plant, fig. 17C and pi. indicating that seasonal fluctuations of TCE concentrations decrease with increasing distance from the TCE source. Fate and Migration ofTrichloroethylene The fate and migration of TCE through an aquifer system are affected by physical, chemical, and biolog- ical processes. Physical processes include advection, dispersion, and nonaqueous-phase flow; chemical processes include volatilization, sorption, and dissolu- tion; and biological processes include aerobic and anaerobic biotransformations (MacKay and others, 1985). Some of these processes are illustrated in figure 18. The physical processes of advection, dispersion, and nonaqueous-phase flow affect the transport of TCE in the study area. The advective phase of TCE plume generally follows the direction of ground-water flow as it moves northeastward from the typewriter plant and either discharges into Stupke Pond and Otter Creek or remains in the central part of the aquifer. Physical dispersion of TCE results from differing rates of ground-water flow that result from varying of flowpaths (some flowpaths follow a more circuitous route than others) and local variations in the hydraulic conductivity of the aquifer. The relatively constant TCE concentrations that were found in most offsite wells before operation of the recovery well (1987 through the end of 1990) suggest a slow desorption of TCE from contaminated aquifer sediments and (or) dissolution of a dense, nonaqueous- phase liquid (DNAPL) in the aquifer. DNAPLs migrate downward under gravitational force in the unsaturated zone, where they may disperse, dissolve, degrade, or be removed by pumping or excavation. DNAPL's in the unsaturated zone tend to partition into a gaseous phase and migrate as volatilized constit- uents. Most soil-gas TCE will diffuse upward in response to natural concentration gradients and will eventually escape into the atmosphere, but some will be adsorbed onto sediments in the unsaturated zone (Cho, and others, 1991) by one or more of the follow- Volatile Organic Compounds 35 ---PAGE BREAK--- ing sorption mechanisms: mineral-surface absorp- tion, partitioning into natural organic matter, partitioning into water that coats the surface of sediments, and adsorption into micropores. The amount of natural organic matter already on the unsat- urated sediments is reported as the dominant factor for absorption of CHCs onto sediments, but where the organic content of the sediment is low, absorption in micropores may contribute significantly to sorbate uptake (Farrell and Reinhard, 1994). Where a spilled DNAPL exceeds the retaining capacity of the unsaturated zone, it seeps to the water table and sinks downward within the aquifer (fig. 18), where some of it becomes trapped in pore spaces, some dissolves in the ground water, and the rest continues downward (DNAPL is denser than water) until it encounters a stratigraphic layer of low perme- ability (such as at the bottom of the aquifer), where it will accumulate as a pool. Continuous dissolution and dispersion of DNAPL's within the pore spaces of the saturated sediments and in the pool at the bottom of the aquifer then provides a constant release of contami- nants (Barbee, 1994). DNAPL pools have been diffi- cult to locate because they are relatively small and commonly migrate downslope along the bottom of aquifer and accumulate in local depressions. No such pools have been found in the study area to date; they may have dispersed or migrated elsewhere, or they could have eluded detection by the few test wells that extend to the bottom of the aquifer. The rate and extent of chemical and microbial transformation of TCE in the subsurface are controlled by the physiochemical properties of TCE, by chemical properties of soil and water, and the microbial popula- tion in the soil (Barbee, 1994). The three major isomers formed by reductive dehalogenation of TCE TCE LAND spill ^ SURFACE siter+ Lengthof plume in study area about 1.5 miles UNSATURATED ZONE flushed by rain and snowmelt several weeks of the year ZONE OF ANNUAL FLUCTUATION OF THE WATER TABLE flushed for several months each year during high-recharge period in the spring AQUIFER, (SATURATED ZONE) flushed continuously BOTTOM OF AQUIFER tillorglaciolacustrine fine sand,silt.and clay____ Direction of ground-water flow Volatilization and sorption of some of contaminant onto soil Aerobic degradation / / / / / / / Aerobic and anaerobic degradation depending on the time of year DNAPL TCE piume Anaerobic degradation and sorption Path of dissolved contaminant Dense nonaqueous phase liquid (DNAPL) pod Figure 18. Physical, chemical, and biological processes that affect the fate and distribution of trichloroethylene (TCE) in and above the unconfined aquifer at Cortland, N.Y. 36 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- under methanogenic conditions by anaerobic soil microorganisms are a's-l,2-DCE, trans-\ ,2-DCE and 1,1-DCE (Ehlke and others, 1991; Imbrigiotta and others, 1991; Davis and Olsen, 1990). C7s-l,2-DCE is by far the most common transformation product in ground water in the study area, as well as in most other parts in the United States. The ratio of the cis form to the trans form of 1,2-DCE in laboratory experiments is about 30:1 (Davis and Olsen, 1990). Primary and secondary chemical and microbiological transforma- tion pathways of TCE under anaerobic conditions are depicted in figure 19. Anaerobic mechanisms such as hydrolysis and dehalogenation are the primary abiotic transformation pathways of TCE breakdown in typically reduced subsurface environments, but reactivity ranges over 10 orders of magnitude (Barbee, 1994). Under typical ground-water conditions, chlorinated ethenes such as TCE are relatively nonreactive; and 1,2-DCE is the least reactive product. Only under strongly basic conditions will chlorinated ethenes hydrolyze (Jeffers and others, 1989). Biological transformation of CHCs can occur under a wide range of environmental conditions. The most significant process in the subsurface is microbi- ally mediated dehalogenation (McCarty, 1991). The bacterial process, called cometabolism, involves an inducer/substrate (such as methane) that activates as an enzyme pathway that, in turn, oxidizes the substrate to generate carbon and energy for bacterial growth, while also oxidizing CHCs, such as TCE (McCarty, 1991). The cometabolic reductive dehalogenation process results in degradation products of halogenated and nonhalogenated aliphatic hydrocarbons (fig. 19) that can serve as primary substrates for further biodegra dation of CHCs (Barbee, 1994). The biotransformation of TCE may produce degradation products that are persistent in the environment ( such as a's-l,2-DCE and trans-1 ,2-DCE), and that are greater health threats than TCE, (such as vinyl chloride, which is carcino- genic to humans). The half-lives for TCE and its chlorinated degradation products are given in table 6. Further transformation of cis-1,2-DCE is uncertain. The half-life of cis- 1,2-DCE through abiotic hydrolysis is at least 85 million years (table thus, the likelihood of detecting the products of further degradation would be small-only on rare occasions and in trace amounts was 1,2-DCA detected. Other possible transformation compounds, such as vinyl chloride, were not detected. The ultimate transformation products of TCE are carbon dioxide, water, and chloride (fig. 19). Trichloroethylene (TCE) c/ xci t 1,1 Dichloroethylene c/s-1,2-Dichloroethylene frans-1 ,2-Dichloroethylene H Cl Cl Cl H Cl\ Cl Cl Cl \ / c=c c=c ' \ / \ H H H Cl\ / C = C H Cl 1,2-Dichloroethane Cl Cl Carbon dioxide Water Chloride O = C = O H-O-H Cl ~ EXPLANATION PRIMARY TRANSFORMATION PATHWAY SECONDARY TRANSFORMATION PATHWAY Figure 19. Transformation pathways of trichloroethylene under anaerobic conditions. (Modified from Davis and Olsen, 1990, fig. Volatile Organic Compounds 37 ---PAGE BREAK--- Table 6. Range of half-lives for trichloroethylene and its degradation products. = greater than, data from Barbee, 1994] Half-life process Chlorinated hydrocarbon Trichloroethylene cw-l,2-dichloro- ethlene Irons- 1 ,2-dichloro- ethylene 1 ,1-Dichloroethylene 1 ,2-Dichloroethane Vinyl chloride Abiotic hydrolysis Anaerobic or dehalogenation biodegradation half-life (years) half-life (days) 0.42 8.5 xlO7 - 8.5 xlO9 - 4.7 x 107 - 24 > - 1.1 2.1 x 10 10 2.1 x 10 10 1.2xl08 -61 10 33 - 320 88 - 339 53 - 147 81 - 173 >60 >60 Inorganic Chemical Constituents and Physical Characteristics Chemical quality of ground water is generally affected by several factors: the chemical composi- tion of precipitation that recharges the aquifer; chemical reactions with the soil as the recharge passes through the unsaturated zone; chemical reactions between the aquifer material and ground water in the aquifer; residence time of water within the matrix, and land use above the aquifer. Chemical analyses of ground-water samples collected during April and September 1990 (table 7) in parts of the unconfined part of the aquifer that are not contaminated by TCE indicate that the quality of ground water in the study area generally meets New York State drinking-water standards. Concentrations of some constituents at some wells exceeded the drinking-water standards, however. Specific Conductance, pH, and Alkalinity Specific conductance is a measure of the capacity of water to conduct an electrical current and is related to the type and concentration of ions in solution. Specific conductance is affected by precipitation and by chemical and physical reactions such as adsorption, ion exchange, oxidation, and reduction. Specific conductance of water samples collected during April 1990 ranged from 257 to 1,440 piS/cm, with a median value of 544 piS/cm (table 8A), and values in Septem- ber 1990 ranged from 288 to 3,850 piS/cm, with a median of 612 piS/cm (table 8B). Median specific conductance values for the April samples did not differ significantly from the September values, nor from the median value of 440 piS/cm for samples collected in 1976 (Duller and others, 1978). Median values for the 1990 sampling from Cortland aquifer seemed to be higher than those for other aquifers in upstate New York, however (Miller and others, 1988). The pH of a solution is a measure of the effective hydrogen-ion concentration. The primary determinant of pH in ground water is the interaction of soil and rock molecules with gaseous and dissolved carbon dioxide, bicarbonate, and carbonate ions. The pH of ground water in the study area ranged from 7.2 to 8.4 with a median value of 7.6 for the April 1990 sampling, and from 6.9 to 8.9 with a median value of 7.7 for the September 1990 sampling. These median values indicate that the water in the study area is basic. Alkalinity is a measure of the capacity of water to neutralize an acid by chemical buffering; most alkalin- ity results from bicarbonate (HCO3) and carbonate (CO3) ions. Median values for alkalinity in the study area were 190 mg/L during the April 1990 sampling and 202 mg/L for those collected in the September 1990 sampling. These values are within the range of most ground water in the State. Chloride and Sodium Potential sources of chloride and sodium in ground water include road-deicing salts, septic-tank effluents, and sodium-bearing minerals within the aquifer. In some places, elevated concentrations of these ions can result from the discharge of mineralized water in bedrock into sand and gravel aquifers. Elevated concentrations of chloride were generally found in ground water near major roads that are heavily salted. Median concentrations of chloride for the April and September 1990 samplings were 37 and 39 mg/L, respectively. The chloride concentration of 274 mg/L at the Cortland County office building during April 1990 exceeded the New York State drink- ing-water standard of 250 mg/L (well 177, pi. Chloride concentrations in ground water at the City of Cortland well field have been increasing since the early 1940's (Buller and others 1978) (fig. 20). Although still well below the New York State drink- ing-water standards, chloride concentrations there are significantly higher (at the 95-percent confidence level) than in 1976. Chloride has migrated downward to some degree throughout the study area. The median concentration of samples from depths greater than 22.5 ft below the 38 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- water table was 30 mg/L, and the median concentra- tion of samples from depths less than that was 39 mg/L. The chloride concentration of a sample from well 306, which is 255 ft deep, was 28 mg/L. Chloride concentrations during the April 1990 sampling were generally less than 40 mg/L in the western part of the study area, except at the north end of the municipal airport, where the concentration at well 330 was 80 mg/L. Concentrations in the central part of the aquifer ranged from 40 to 80 mg/L, and those in the eastern part ranged from 40 mg/L to 120 mg/L (fig. 21). Dissolved sodium concentrations in the study area ranged from 4.8 mg/L to 530 mg/L, with a median value of 28 mg/L. Although no standard has been established for sodium, the USEPA (1976) recom- mends less than 20 mg/L in drinking water for people on sodium-restricted diets. Of the 30 wells that were sampled during September 1990,20 had sodium concentrations above the 20-mg/L limit. Trace Elements (Iron and Manganese) Trace elements generally occur naturally in ground water in extremely low concentrations, but industrial processes and urbanization tend to increase their abundance within the hydrologic system and thereby can seriously degrade water quality. Except for iron and manganese, trace elements detected in ground water in the study area did not exceed drinking-water standards. Iron and manganese concentrations are generally considered together; State drinking-water standards specify a combined maximum limit of 300 ^ig/L. Concentrations of iron and manganese exceeded the 300-pig/L limit at 7 of the wells sampled during September 1990; the median concentration of iron and manganese at each was 16 pig/L. Extremely high concentrations of iron and manganese (7,900 pig/L and 3,600 pig/L, respectively) were found in water from well 365 (pi. in the eastern part of the study area. This well is at an abandoned gasoline station on the edge of the TCE plume, and dissolution of old under- Table 7. Minimum, maximum, mean, median, and interquartile range of concentrations for selected constituents or properties of ground-water samples collected from the glacial aquifer in the Cortland, N.Y. study area during April and September 1990. [Values are in milligrams per liter, mg/L, unless otherwise noted, ^iS/cm, microsiemens per centimeter; ^ig/L, micrograms per liter] Constituent Number of or property samples Well depth, in feet Alkalinity (as CaCO3) pH Specific conductance, piS/cm Phosphorus, total as PO4 Nitrate as N Chloride Sodium Potassium Calcium Nitrite + nitrate as N Magnesium Sulfate Silica Barium (pig/L) Berylium (pig/L) Cadmium (pig/L) Chromium (pig/L) Cobalt (\ng/L) Copper (jig/L) Iron (pig/L) Lead (pig/L) Manganese (IAg/L) 86 101 60 60 59 59 69 89 40 30 40 30 30 30 30 30 30 30 30 30 30 30 30 Minimum 14 62 6.9 257 .01 <.01 5.0 4.8 .2 .1 .10 6.0 17 .55 29 .50 <1 <5 <3 <10 <3 10 <1 Maximum 255 312 8.9 3,850 1.29 12 270 530 8.3 780 9.9 39 25 14 290 1.5 3.0 15 9 50 7,900 30 3,600 Mean 55 191 7.7 673 .44 4.1 52 37 1.7 94 4.4 15 22 7.2 77 .55 1.1 5.3 3.3 13 334 11 206 Median 47 196 7.7 592 .43 4.3 38 23 1.4 79 4.6 14 23 7.4 62 .50 < 1 <5 <3 <10 16 10 16 Interquartile range 25th 75th 34 167 7.6 513 .12 2.4 28 15 1.0 72 3.4 14 21 6.7 50 .50 <1 <5 <3 <10 10 10 4 67 220 7.8 705 .68 5.6 61 39 2.0 89 5.6 16 24 7.9 74 .50 <1 <5 <3 <10 38 10 72 Inorganic Chemical Constituents and Physical Characteristics 39 ---PAGE BREAK--- Table 8. Minimum, maximum, mean, median, and interquartile range of concentration or value for selected constituents or properties of ground-water samples collected from the glacial aquifer in the Cortland, N.Y. study area during April 1990 and September 1990. [Values are in milligrams per liter (mg/L) unless otherwise noted, uS/cm = microseimens per centimeter, ft = feet] Constituent or property A. April 1990 Well depth (ft) Depth to water surface (ft) Depth below water table (ft) Alkalinity (as CaCO3) pH Specific conductance, u.S/cm Phosphorus, total as P Nitrate as N Chloride Sodium Potassium B. September 1990 Well depth (ft) Depth to water surface (ft) Depth below water table (ft) Specific conductance pH Alkalinity (as CaCO3) Nitrite + nitrate as N Calcium, dissolved Magnesium, dissolved Sodium, dissolved Chloride, dissolved Sulfate Silica, dissolved Barium, dissolved (pig/L) Berylium, dissolved (pig/L) Cadmium, dissolved (IAg/L) Chromium, dissolved (pig/L) Cobalt, dissolved (IAg/L) Copper, dissolved (pig/L) Iron, dissolved ([Ag/L) Lead, dissolved (Iig/L) Manganese, dissolved (IAg/L) Number of samples 43 40 32 59 28 28 59 59 59 59 59 43 29 28 32 32 42 40 30 30 30 10 10 30 30 30 30 30 30 30 30 30 30 Interquartile range Minimum 14 3.3 4.3 62 7.2 257 .01 .1 5.0 5.0 .15 14 5.5 2.5 288 6.9 104 .10 13 6.0 4.8 21 17 .55 29 .50 < 1 <5 <3 <10 <3 10 <1 Maximum 255 61.6 247 265 8.4 1,440 1.3 12 270 200 8.3 217 72.8 207.3 3,850 8.9 312 9.9 230 39 530 79 25 14 290 1.5 3 15 9 50 7,900 30 3,600 Mean 55 15.5 39.7 182 7.7 581 .44 4.1 52 29 1.7 55 22.9 33.6 753 7.7 204 4.4 82 15 54 47 22 7 77 .55 1.1 5.3 3.3 13 330 11 210 Median 45 11.5 27.2 190 7.6 544 .43 4.3 37 22 1.4 48 17.5 22.4 612 7.7 202 4.6 79 14 28 39 23 7.4 62 .50 <1 <5 <3 <10 16 10 16 25th 28 7.2 17.6 154 7.5 477 .12 2.4 28 14 1.02 34 13.1 14.6 555 7.6 187 3.4 74 14 18 30 21 6.7 50 .50 < 1 <5 <3 <10 10 10 4 75th 67 18.2 45.5 212 7.7 660 .68 5.6 61 35 2.0 67 28.0 35.7 796 7.8 224 5.6 90 16 47 70 24 7.9 74 .50 <1 <5 <3 <10 38 10 72 40 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- 1960 1970 YEAR 1980 1990 Figure 20. from wells and 1990. Dissolved chloride concentrations in water at City of Cortland well field, 1943-76, 1980, (Modified from Buller and others, 1978, fig. 20.) ground gasoline-storage tanks may have been the source of the iron and manganese. Nitrogen Nitrogen, which constitutes about 80 percent of the atmosphere, generally occurs in nature in combination with other elements and is a common degradation product of organic wastes. Nitrate sources include the decomposition of organic nitrogen that is introduced to the soil by nitrogen-fixing plants and bacteria, human and animal wastes, and organic and inorganic fertilizers. Total nitrate concentrations (NO3~ as N) in ground-water samples collected in April 1990 ranged from 0.1 to 12 mg/L, with a median value of 4.3 mg/L. The largest concentrations (12 and 8.6 mg/L) were at wells 304 and 204, respectively, both of which are close to Otter Creek and just downgradient of a large dairy farm. 76'15' 42°37' West Branch Tioughnioga River East Branch Tioughnioga River CHLORIDE CONCENTRATION- in milligrams per liter LESS THAN 40 40-80 80-120 NO DATA BOUNDARY OF GLACIAL AQUIFER DATA POINT 0.5 1.0 KILOMETER 33' - Base from U.S. Geolocial Survey, Cortland 1:62,500, 1903 Figure 21 . Distribution of chloride concentrations in the unconfined aquifer in Cortland study area, April 1990. Inorganic Chemical Constituents and Physical Characteristics 41 ---PAGE BREAK--- Ground-water samples collected in September 1990 were analyzed for dissolved nitrite plus nitrate (NO2~+NO3~) as N. Nitrate (NO3~) is typically the most common nitrogen species in surface water and ground water, whereas nitrite (NO2~) is unstable and usually undergoes nitrification or denitrification and typically occurs in concentrations of less than 0.1 mg/L (Behnke, 1974). Concentrations of nitrite plus nitrate ranged from O.I to 9.9 mg/L in September 1990 with a median value of 4.6 mg/L. The highest concentrations were 9.9 mg/L and 8.3 mg/L in wells 5 and 207, respectively, both of which are subject to agricultural influences. Wells 304 and 204, which had the highest nitrate concentrations in April 1990, had concentra- tions of 4.0 and 4.6 mg/L, respectively, in September 1990. Data from several well pairs (a shallow well and a deep well) indicate that the vertical distribution of nitrogen is fairly uniform throughout the upper part of the unconfined aquifer. Nitrate-concentration data from six wells finished in the confined aquifer indicate that nitrogen is present in that aquifer, but in lower concentrations (median value 2.0 mg/L). Temporal Changes Data collected in this study (April and September 1990) were compared with data collected in 1978 by Duller and others (1978), and with data collected from November 1979 through January 1981 by Cortland County, to discern trends in water quality in the glacial-aquifer system. Only wells that also had been sampled in the 1978-81 studies were represented in the comparison. Constituents that were compared were alkalinity, nitrite plus nitrate as N, calcium, chloride, and sulfate. Chloride was the only constituent for which concentrations for all four periods were avail- able. Maximum, minimum, mean, median, and inter- quartile ranges of the constituents that were compared are shown in table 9. Nonparametric statistical tests were used to discern year-to-year differences in median values. A 0.05 level of significance (95-percent confidence level) was used for all tests to describe the error probability of falsely detecting differences. Simple descriptive statistics such as boxplots and median and interquar- tile range were used first to examine the distribution of the sample population of each constituent from the three time periods. Because water-quality data typically do not have a normal distribution, the median and interquartile range provide a better measure of the sample-population distribution than do the mean and the standard deviation. For this reason, a Kruskal- Wallis one-way ANOVA (analysis of variance) was done on ranked data to compare the distribution of concentrations by year (Conover and Iman, 1981). Differences among sample populations (by year) indicated by the Kruskal-Wallis test were further defined by Tukey's multiple-comparison test. Concentrations of chloride, the only constituent with data from all four time periods, were significantly greater during April and September 1990 than in 1976 or 1980. Median values for April and September 1990 were similar (36 and 38 mg/L, respectively), and the median concentration for 1980 (21 mg/L) was greater than for 1976 (18 mg/L). The increase in chloride concentration over the years (fig. 20) is due to increased use of road salt and, to a lesser extent, leakage from aging septic systems. Median concentrations of alkalinity were greater in September 1990 (210 mg/L) than in April 1990 (191 mg/L), and the September median was signifi- cantly greater than that for 1976 (167 mg/L). Median concentrations of calcium for April and September 1990 where nearly equal (80 and 77 mg/L, respec- tively) but were significantly higher than in 1976 (63 mg/L). Median concentrations of nitrite plus nitrate differed significantly among sampling periods- September 1990 had the highest, and 1976 had the lowest. The trend of increasing nitrogen concentra- tions may result from increased application of agricul- tural and lawn fertilizer, as well as leaking septic systems. The significant difference between nitrate concentrations in April 1990 and those in September 1990 may result from three factors: recharge during this period was fairly large; thus, concentrations may have been diluted; ground-water levels in Septem- ber had declined by an average of 7.4 ft, decreasing the amount of dilution; and nitrogen from fertilizers applied during late spring and summer probably had reached the water table by September. The distribution of concentration data for alkalinity, nitrite plus nitrate, calcium, and chloride in 1976 and the two 1990 sampling periods are shown by year in figure 22A. Ground-Water Chemistry and Land Use The available data allow only generalized compar- isons between ground-water quality and land use. Because land use in the study area is heterogeneous, the effects of a particular land use on the quality of water at a particular well or group of wells are impos- sible to isolate. Most of the 52 wells sampled were in 42 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- the Cortland municipal area, which contains several land-use categories, including residential, industrial, commercial, recreation, major transportation routes, and other small categories associated with urban areas. The remaining 16 wells were grouped as agricultural. Water from these wells, which were in the southern end of the study area and at its edges, was readily identifiable by characteristics that are associated with agriculture. Thus, ground-water quality was compared simply between two broad categories of land use- urban and agricultural. Six wells that penetrated the confined aquifer were eliminated from the data set prior to analysis of because part of their recharge area is outside the study area. The distribution of chemical data for the two land- use groups were compared through use of boxplots (fig. 22B). Any statistically significant differences in concentrations reflecting land use were tested with the Kruskal-Wallis one-way ANOVA, which compares the distribution of ranks of concentrations between groups of data to determine whether any of the groups differ significantly from the others. Significance was chosen at 0.05 (95-percent confidence level). Table 9. Minimum, maximum, mean, median, and interquartile range of concentration or value for selected constituents or properties of ground water sampled during three studies in Cortland County, N.Y. study area [Values are in milligrams per liter unless otherwise noted, Ij,S/cm, microsiemens per centimeter; dashes indicate no data]. Constituent or property Specific conductance (uS/cm) Alkalinity (as CaCO3) Nitrate (total as N) Nitrite + nitrate (asN) Calcium Chloride Sulfate (as SO4) Sampling date1 1976 1980 April 1990 Sept. 1990 1976 1980 April 1990 Sept. 1990 1976 1980 April 1990 Sept. 1990 1976 1980 April 1990 Sept. 1990 1976 1980 April 1990 Sept. 1990 1976 1980 April 1990 Sept. 1990 1976 1980 April 1990 Sept. 1990 Number of samples 84 0 22 6 23 0 22 12 27 153 22 0 84 0 0 12 19 0 22 6 85 153 22 7 18 0 0 7 Interquartile range Minimum 235- 236 413 63- 62 151 1.6 .10 .80 - .55 - - 3.7 31 32 61 3.0 5.0 15 29 19 19 Maximum 700 797 631 231 239 300 4.9 5.0 12 5.6 - 9.9 80 97 87 78 72 110 79 26 25 Mean 451 474 545 168 180 209 3.4 2.4 4.6 3.2 6.0 64 78 76 24 24 45 46 22 23 Median 450- 470 556 167 191 210 3.4 2.7 4.6 3.4 - 5.8 63 80 77 18 21 36 38 21- 24 25th 405- 334 512 152 167 183 2.8 1.6 2.4 2.7 - 4.7 59 75 72 12 14 27 30 20 - 23 75th 499 542 584 182 219 224 4.1 3.3 6.1 - 4.0 7.3 76 - 87 81 28 30 61 72 24 25 1 1976 data from Duller and others (1978). 1980 data provided by Cortland County Department of Health. 1990 data from present study Inorganic Chemical Constituents and Physical Characteristics 43 ---PAGE BREAK--- A. Concentration by year £ 300 b LU 250 CO < 225 Oj 200 ' 1 1?5 I 15° § 125 DC O 0 75 O . Alkalinity T ! I r3^ T I I r~i nr - y 12 10 8 6 4 2 0 - Nitrite plus nitrate . (Apr 1 1990 is nitrate) - _ -T :M 100 80 60 40 20 0 . Dissolved chloride : T ! : - _ P - I I I V . 90 80 70 60 50 40 30 j Dissolved " calcium - T B o - T P e M ^ _ _ 1976 1990 1990 1976 1990 1990 1976 1980 1990 1990 1976 1990 1990 (Apr I) (Sept.) (April) (Sept.) (Apr l)(Sept.) (April) (Sept.) B. Concentration by land use oc LLJ b 300 DC LLJ Q_ ^ 250 ^ DC CD I-I 200 Z O 150 DCI- § 100 oz o o 50 Alkalinity " - 9.5 8.5 7.5 6.5 5.5 4.5 0 C o.o 2.5 1.5 0.5 - - Nitrite plusD nitrate as N ; 1 ; 250 200 150 100 50 0 Dissolved chloride . . i ~r I . ^ 1 . Q4UUU in ^3500 ^ DC LJJ 3000 1- LJJ^ P2500 Z LJJo 0-2000 LJJ 0. co-icnn \ \j\j\j LJJ^y 1000 COo DC O 500 Specific conductance -p I I I I "I Agricultural Urban Agricultural Urban Agricultural Urban Agricultural Urban EXPLANATION outliers values within _ 1.5 times interquartile range ^ 75"' quartile median ~^ri_ oc'*' m lartilp 1 i ^3 ujucu me values within 1 .5 times Figure 22. Concentration of alkalinity, dissolved calcium, nitrite plus nitrate, and interquartile in water frnm splerteri wplk in NY.. 1976. range 1980 and 1990. A. By year. B. By land use. outliers 44 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- Only two constituents specific conductance and chloride showed a significant difference between the two land-use categories; concentrations of both were significantly higher in the urban area than in the agricultural area. The higher chloride concentrations in the urban area may be attributable to road salt, indus- trial wastes, and sewage leaked from aging septic and sewer systems, and the significantly greater specific conductance values in the urban area than in the agricultural area are associated with the higher chloride concentrations. The median chloride concen- tration was 44 mg/L in the urban area and 30 mg/L in the agricultural area, and the median specific conduc- tance values were 610 jiS/cm and 516 jiS/cm, respec- tively. Median concentrations of sodium were considerably larger (although not statistically signifi- cant) in the urban area than the agricultural area (24 mg/L and 16 mg/L, respectively). Alkalinity was higher in the urban area, and median concen- trations of nitrate and calcium were nearly the same in both areas. Concentrations of nitrite plus nitrate were considerably higher in the agricultural area (although not a statistically significant difference) as a result of the agricultural sources of that constituent. Maximum, minimum, median, mean, and interquartile ranges of the selected constituents grouped by land-use categories are in table 10. SIMULATION OF GROUND-WATER FLOW A ground-water flow model was constructed to delineate areas that contribute ground water to municipal wells and the flowpaths of ground water migrating from two sources of contamination. Because annual and seasonal fluctuations in the amount of recharge affect flow conditions in the aquifer system, three steady-state ground-water recharge conditions- high, average, and low-were simulated. Description and Design of Numerical Model A quasi-three-dimensional, numerical three layer ground-water flow model was constructed to compute hydraulic head and flows in the glacial aquifer system in the study area. Heads in the confining unit were not calculated; instead, resistance to flow in the confining unit is included in terms of vertical conductance between the unconfined and confined aquifers. This approach to simulating flow through a confining layer is called the "quasi-three-dimensional" approach. The model was developed with the computer program MODFLOW (McDonald and Harbaugh, 1988), which is based on block-centered, finite-difference equations that describe the physics of water flowing through a porous medium. The equations relate water levels to geometry and hydraulic properties of the aquifer, such as hydraulic conductivity, and to stresses such as Table 10. Minimum, maximum, mean, median, and interquartile range of concentrations or values for selected constituents or properties of ground water in Cortland, N.Y. study area, by land use [Values are in milligrams per liter unless otherwise noted; Ij,S/cm = microsiemens per centimeter, * indicates significant difference at the 95-percent confidence level in constituent concentration between land uses.]. Constituent or property Alkalinity (as CaCO3) * Specific conductance (H,S/cm) Nitrate (as N) *Chloride (as Cl) Sodium (as Na) Calcium (as Ca) Nitrite + nitrate (as N) Land use Agricultural Urban Agricultural Urban Agricultural Urban Agricultural Urban Agricultural Urban Agricultural Urban Agricultural Urban Number of obser vations 20 76 12 43 12 45 17 50 14 70 14 70 6 31 Interquartile range Minimum 62 90 257 344 .1 .1 15 5 4.8 4.8 .10 47 .70 .30 Maximum 300 312 668 3,850 8 12 93 270 51 530 780 240 9.9 7.6 Mean 183 197 501 728 4.2 4.2 37 58 21 39 150 86 5.9 4.3 Median 180 197 516 610 4.2 4.5 30 44 16 24 78 79 6.4 4.6 25th 154 177 351 546 2.3 2.7 21 31 11 16 55 73 3.0 3.5 75th 218 221 634 802 6.0 5.5 42 67 30 39 88 89 8.7 5.4 Simulation of Ground-Water Flow 45 ---PAGE BREAK--- pumping and recharge. The aquifer system is repre- sented by a three-layer grid in which all characteristics of the system, including geometry, hydraulic proper- ties, and stresses, are defined by values specified at the centers of cells, and the model calculates the head at the center of each cell. The process of representing a continuous system with a specified number of discrete points is called discretization. In the finite-difference method, the discrete points are located along rows and columns, and each point is associated with a cell. Head and flows are calculated only for "active" cells, which represent the aquifers; "inactive" cells are those outside the aquifer boundary and are ignored in the calculations. The design of the numerical model is based on the conceptualization of the aquifer system as previously discussed in the hydrogeology section and shown schematically in figure 23. The aquifer system is simulated as a quasi-three-dimensional-flow system with three layers, and horizontal flow in the aquifers and vertical flow through the confining unit is assumed. The unconfined aquifer is divided into two layers (layers 1 and 2) to provide adequate vertical resolution and to simulate flowpaths and particles that could flow beneath cells that simulate streams and wells (such cells are known as "weak sinks" because not all flow into a cell is captured by the cell). The confined aquifer is represented by a single layer (layer 3) except where the confining unit is locally absent along the edges of the valley; in these locations an arbitrary elevation that approximates the elevation of the top of the adjacent confining unit was assigned to the top of the layer. In this quasi-three dimensional approach, the confining unit is represented by the vertical conductance between layers 2 and 3. Heads in the confining unit are not calculated, and vertical flow through it is the product of differences in head between layers 2 and 3 and the vertical conductance of the confining unit. The geometric and hydraulic values that were used as input to the model were based on available data. The values for areas with little or no data were estimated WEST Recharge from upland runoff EAST Recharge from precipitation Discharge to wells Discharge to large stream Recharge from tributary streams LAYER 1 - Upper part of unconfined aquifer LAYER 2 - Lower part of unconfined aquifei Confining unit (lacustrine fine sand, silt, and clay) Heads in this layer are not calculated. Resistance to flow in this layer is included in vertical conductance terms between layers 2 and 3 Confined aquifer Recharge to confined aquifer Impermeable (no-flow) boudary Bedrock Bedrock Not to scale Figure 23. Conceptual model of ground-water flow in glacial-aquifer system in Cortland, N.Y., study area. 46 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- and then adjusted, within reasonable limits, during model calibration until the simulated water levels and flows matched the measured values. Model Grid A rectangular, finite-difference grid with 66 rows and 119 columns was superimposed on a map of the aquifer system (fig. 24 and pi. 5) to discretize the hydrogeologic conditions of the conceptual model. A uniform cell size of 300 x 300 ft was used because municipal wells, streams, and sources of contamina- tion are distributed throughout the study area. (A grid with cells of uniform size simulates the hydrologic conditions throughout the study area more accurately than a grid with variably sized cells, although the latter provides increased resolution in areas represented by small cells, they may not accurately represent condi- tions in large-cell areas. Geometry of Model Layers A geologic section was drawn for every second row of the model, giving a total of 33 geologic sections. The surface elevations of the bottom of layer 1 and the top and bottom of layers 2 and 3 of the geologic sections were entered as arrays into the model. Corresponding data for intervening rows, for which no geologic sections were constructed, were interpolated, or were determined from records of wells corresponding to those rows. The elevation of the top of layer 1 (water table) was calculated by the model, and the elevation of the bottom of layer 1 (top of layer 2) was arbitrarily assigned to roughly equal to the middle of the unconfined aquifer, 76°15' 12'30" EXPLANATION I_I Active cell area Inactive cell area 0 1 2 3 KILOMETERS » c\ i - ik j^wv:^tJ X - Superfur TCE spill site _TOWN Fish T0WN OF VIRGIL hatchery 1 MILE ' 1 KILOMETER Base from U.S. Geological Survey 1:62,500 series: Gotland (1903) and Groton (1903) EXPLANATION STUDY AREA AQUIFER BOUNDARY MUNICIPAL WELL FIELD DRAINAGE DIVIDE BETWEEN - - OSWEGO RIVER BASIN AND SUSQUEHANNA RIVER BAISN DIRECTION OF STREAMFLOW Figure 24. Locations of active and inactive grid boundaries in layers 1, 2, and 3 of the three-dimensional ground-water flow model of Cortland, N.Y. study area. Simulation of Ground-Water Flow 47 ---PAGE BREAK--- except along the valley walls, where till or bedrock surface forms the bottom of layer 1. The bottom of layer 2 is either the top surface of the lacustrine confining unit (fig. 12) in the central parts of the valley or the top of layer 3 along some reaches of the valley walls (fig. 23). The top of layer 3 is the bottom of the confining unit or, where the confining unit is absent, the bottom of layer 2. The bottom of layer 3 is the till or bedrock surface in the valley (fig. 23). Hydraulic Conductivity Horizontal hydraulic conductivity of the aquifers was estimated to be 10 times the vertical hydraulic conductivity throughout the modeled area (anisotropy 10:1). Vertical hydraulic conductivity of stratified drift tends to be less than horizontal hydraulic conductivity because, at a small scale, the stratified drift consists of many layers of sediment particles, some of which are plate shaped and tend to settle horizontally, thereby impeding the vertical flow of ground water. Unconfined Aquifer Horizontal hydraulic conductivity values ranging from 1 to 1,200 ft/d were assigned to layers 1 and 2, on the basis of eight aquifer tests that used large pumping wells, such as municipal and industrial wells (table and several additional hydraulic conduc- tivity measurements made by Blasland, Bouck and Lee Engineers (1992) during slug tests of test wells at the Rosen Superfund site in the southeastern part of the aquifer (pi. Hydraulic conductivity values of less than 10 ft/d were assigned to the unconfined aquifer on the back (west) side of the Valley Heads moraine in the western part of the study area to represent the poorly sorted, silty sand and gravel and the fine-grained deposits of till and lacustrine fine sand, silt, and clay. Hydraulic conductivity values of 10 to 100 ft/d were assigned to the poorly to moderately sorted silty sand and gravel that form kame deposits at the crest of the Valley Heads moraine, and silty gravel outwash along parts of the valley walls. Blasland, Bouck and Lee Engineers (1992) calculated an average hydraulic conductivity of about 10 ft/d for kame deposits and 30 ft/d for outwash, and a range of 3.4 to 136 ft/d in 20 slug tests of test wells at the Rosen"Superfund" site. Hydraulic conductivity values ranging from 100 to 300 ft/d were assigned to areas along the edges of the vaHey, where alluvial inwash and colluvium that consist of silty sand and gravel deposited by upland tributaries and by runoff are mixed with well-sorted outwash sand and gravel along the sides of the valley. This zone is well developed where large tributaries flow onto the aquifer and is weakly devel- oped where small tributaries flow onto the aquifer and where slope erosion has occurred along the valley walls. Values ranging from 300 to 1,200 ft/d were assigned to areas containing well-sorted outwash sand and gravel in the central parts of the valleys. In general, hydraulic conductivity values northeast of the moraine progressively decrease from more than 1,000 ft/d to less than 400 ft/d to reflect the decreasing grain size of the outwash deposits with increasing distance from the moraine. Confined Aquifer Little information is available on hydraulic properties of the confined aquifer (layer there- fore, model values were estimated from two aquifer tests in the western part of the study area, six slug tests at the Rosen "Superfund" site, several qualitative estimates of the water-producing capacity of wells installed in the confined aquifer, and distribution of grain size of sediments collected from the confined aquifer during test drill- ing. The estimated values range from 10 to 150 ft/d and average 60 ft/d. Stream-Aquifer Interaction The Streamflow-Routing Package developed by Prudic (1989) was used to simulate the interaction between the water table and the streams and springs in the study area. Streams are divided into reaches and segments; each reach corresponds to an individual cell in the model grid, and a segment consists of a group of connected reaches in order. Streams in the study area are represented by 425 reaches (cells) grouped into 52 segments. Leakage between the stream and aquifer is subtracted from or added to the amount of streamflow in each reach, depending on the head difference between the stream surface and water table, and is adjusted according to a streambed-conductance term. The amount of leakage is computed by Darcy's Law as follows: 48 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- Q = C ^ (H -H ) ^ s a' where Q = leakage to or from the aquifer through the streambed (L3/T); Hs = head in stream Ha = head in aquifer side of streambed and r\ Cstr = streambed conductance (L defined as the vertical hydraulic con- ductivity of the streambed, times width of the stream reach times stream length of the stream, divided by the streambed thickness. Recharge to the aquifer from streams ceases when all streamflow leaks into the aquifer, leaving the stream dry. The stream can flow again in reaches, however, wherever the head in the aquifer is above the streambed elevation. All streams are in layer 1 of the model, and each stream cell is assigned values for the following: the layer, row, and column number of the cell representing that reach, the average water-surface elevation in the cell, streambed conductance, elevations of bottom and top of the streambed, and stream discharge, in cubic feet per day, for the first reach of each segment. Water-surface and streambed elevations were determined by continuous runs of levels made in the channels of Dry Creek, West Branch Tiough- nioga and Tioughnioga Rivers, spot measurements made with levels in the channels of Otter Creek and several other small tributaries, and estimates from l:24,000-scale USGS topographic maps with 20-ft contour intervals. Water levels and streambed eleva- tions are accurate to several tenths of a foot where levels were run and are accurate to several feet where they were estimated from topographic contours. A streambed thickness of 1.5 ft, which was used in a model of the Cortland aquifer built by Cosner and Harsh (1978), was also used in this model. Conductance of streambeds within the study area differs from reach to reach. Tributaries that erode fine- grained matrix of till in the uplands transport this material to streams that eventually flow over the aquifer, where they deposit some of it; thus, the streambed in the study area consists of both coarse and fine-grained sediments and has a lower hydraulic conductivity than the aquifer. Estimating the conduc- tance of the streambed entailed adjusting the hydraulic conductivity value in the model until seepage to and from the stream approximated the gains and losses measured at 33 streamflow-measurement sites during each of the low-, average- and high-flow periods. Estimated hydraulic conductivity values ranged from 0.5 to 7.5 ft/d and typically averaged about 3.5 ft/d for streambeds in highly permeable outwash deposits and from 0.1 to 0.5 ft/d for streambeds over less permeable deposits, such as kames. Stream width was measured at the 33 streamflow-measuring sites and was estimated in many other places; the range was from less than 2 ft in small tributaries to 120 ft in the Tioughnioga River. Boundary Conditions Several types of boundaries were specified in the model to represent the aquifer system. The types used in layer 1 are indicated on plate 5. Natural boundaries were used where possible, but arbitrary boundaries were used to limit the modeled area because the aquifer system extends many miles beyond the study area. The arbitrary boundaries were placed far enough from municipal well fields and sources of chemical contamination that their effect on model results in these areas of concern would be negligible. Specified-Flux boundaries refer to those bound- aries where a volume of water per unit of time crosses a unit cross-sectional area (such as recharge from pre- cipitation over the aquifer). A specified-flux boundary, represented by recharge wells, was used along the val- ley walls (pi. 5) to simulate the seepage of surface run- off and ground water from bordering unchanneled uplands into the unconfined aquifer (layer The amount of inflow was determined by the size of the drainage area of the upland bordering the modeled area. Drainage areas of unchanneled uplands were delineated on maps, and their size measured by a digi- tizer.. Then the drainage areas were divided by the number of bordering cells, and each of the resulting areas was multiplied by the recharge rate from uplands (See eq. A specified-flux boundary also was used to simulate ground-water flow into and out of the confined aquifer (layer 3) along all four major stream valleys (Fall Creek valley, West Branch and East Branch Tioughnioga Rivers, and Tioughnioga River). Ground water in layer 3 flows into the modeled area in the West and East Branches of the Tioughnioga River valleys and flows out of the modeled area in the Fall Creek and Tioughnioga River valleys. No-flow boundaries represent geologic units or streamlines through which no flux occurs. A stream- Simulation of Ground-Water Flow 49 ---PAGE BREAK--- line is a curve that is tangent to the direction of ground-water flow; thus no flow crosses a streamline. A no-flow boundary was applied at the bottom of the confined aquifer (layer 3) to represent the contact between the aquifer and the underlying till or shale along the bottom of the valley because till and shale have extremely low hydraulic conductivity, and any flow across that boundary would be negligible com- pared to the amount of flow through the sand and gravel aquifer. Arbitrary streamline (no-flow) boundaries were used where West Branch and East Branch Tioughnioga Rivers and the Tioughnioga River enter the modeled area and flow against the side of a valley; this results in ground-water flow from across the valley to the stream. These streamline boundaries were placed far from pumping wells and chemical plumes to avoid affecting model results. A free-surface recharge boundary was used to represent the water table (top of the unconfined aquifer, layer where recharge from precipitation is applied uniformly. The water table can rise or fall, depending on the balance of stresses in the aquifer sys- tem, such as pumping, recharge, and gain or loss of water from streams that flow over the aquifer. A specified-vertical leakance boundary was used to simulate ground-water movement through the lacus- trine confining unit between layer 2 and layer 3 (quasi- three-dimensional approach). Resistance to flow through the confining unit is included in the vertical leakance calculations and is defined as the vertical hydraulic conductivity of the confining unit, divided by its thickness. The volume of water that passes through a model cell (vertical conductance) between layers 2 and 3 is equal to the product of vertical leakance and model cell area. Ground-Water Withdrawals and Recharge Basins Large amounts of water are pumped from several municipal, industrial, and institutional wells completed in the unconfined aquifer (layers 1 and but no large pumping wells tap the confined aquifer. Ground-water withdrawals by large pumping wells for the three simulated conditions are given in table 5. Withdrawals from pumping wells with a fully penetrating screen (extending from the top of layer 1 to the bottom of layer 2, such as the municipal wells for the City of Cortland) were distributed equally in layers 1 and 2. Withdrawals from wells with short screens (10 to 20 ft long) were assigned to the layer that contained the screen. Water pumped from the recovery well at the typewriter plant is passed through an air stripper to remove VOCs, then routed through a pipe to the recharge basins in the northern part of the property, where it seeps back into the aquifer. Infiltration of water from the basins to layer 1 was represented by recharge wells in the model. Model Calibration The three steady-state models were calibrated to represent high-, average-, and low-recharge condi- tions. Calibration entailed matching measured and simulated hydraulic heads in the unconfined and confined aquifers, and simulated and measured gains and losses of streamflow along individual reaches of the modeled area. Steady-state conditions were assumed for the three recharge conditions. The water- level measurements used for calibration were made on March 28-29,1990, during average high-recharge conditions, on May 28-June 4,1991 during average- recharge conditions, and on October 7-9,1991, during average low-recharge conditions. The long-term water-level monitoring well (well 102, pi. 1) at the City of Cortland well field and the USGS stream gage on the Tioughnioga River at Cortland (station number 01509000, pi. 2) were used as guides to determine when the three conditions occurred and the time when synoptic ground-water and streamflow measurements were to be made. The model was calibrated by a trial-and error procedure, whereby the model is first run with the initial input values; if significant differences between measured and simulated water levels and streamflow are noted, the input values are adjusted, and the model is run again. The process is repeated until simulated values are close to measured values. The calibration process resulted in changes of horizontal hydraulic conductivity, vertical conductance of the streambed, and rate of areal recharge. Calibration was considered satisfactory when: simulated water levels were within 3 ft of the measured water levels, and simulated streamflow displayed the gain-and-loss patterns indicated by measured and estimated values. Differences between simulated and measured ground-water levels are shown in table 11, and differences between simulated and measured streamflow gains and losses are shown 50 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- in table 12. The resulting water budgets for each simulated condition are given in table 13. In all three simulations, the largest source of recharge to the aquifer system (55 to 58 percent of total recharge) is from the uplands-seepage losses from upland streams that flow onto the aquifer plus unchanneled runoff and ground-water inflow from the uplands (table 13A). The second-largest source of recharge is precipitation (33 to 39 percent of total recharge) that directly falls over the aquifer. The largest discharge from the aquifer system (57 to 71 percent of total discharge) is leakage of ground water to streams (table 13B); the second largest (26 to 40 percent of total discharge) is to pumping wells. Simulated heads in layers 1,2, and 3 for high-, average-, and low-recharge conditions are shown in figures 25A, B, and C, respectively. Simulated heads in layer 2 are similar to those in layer 1 because both layers represent the same hydrogeologic unit. Parts of the aquifer that have low hydraulic conductivity, such as the morainal deposits in the western part of the aquifer and kame deposits along the valley edges, have relatively steep hydraulic gradients, and parts that have high hydraulic conductivity, such as outwash deposits, have relatively low hydraulic gradients. Simulated ground-water levels in areas of low hydraulic conductivity differed from measured values more than in other areas, but the differences were less than 3 ft. Simulated water levels in areas with high hydraulic conductivity (outwash deposits) and near perennial streams in the northern and eastern parts of the aquifer were typically within 1 ft of measured ground-water levels. Simulation of large pumping wells in the study area resulted in losses of streamflow in model stream Table 11.- Difference between measured and simulated heads at 49 selected wells in Cortland, N.Y., study area for high-, average-, and low-recharge conditions [Values are observed head minus simulated head, in feet. Dashes indicate no water-level measurement made during calibration period because well was dry or not yet installed. Well and cell locations shown in pi. Location Model row 5 5 12 13 14 14 14 16 18 20 20 20 21 21 21 22 23 23 23 24 24 24 25 26 27 Model column 91 95 77 67 53 67 96 91 51 60 70 74 30 64 69 49 39 43 73 33 47 68 54 53 33 Well number 13 114 404 369 106 330 446 373 5 204 337 47 110 358 357 340 306 121 102 320 105 356 368 4 349 Recharge conditions High 1.7 0.4 -0.7 1.3 -1.7 0.8 0.1 0.8 0.3 1.8 ~ -1.8 -3.1 1.2. 0.8 - 1.4 -0.6 0.2 -1.2 0.4 -0.5 0.4 2.9 -0.2 Average -0.2 0.3 -1.3 -0.9 0.2 1.0 -0.4 0.4 2.0 -0.1 -2.0 -3.0 0.9 0.4 -2.0 -1.3 -0.5 -0.6 -0.8 0.4 1.4 0.3 2.4 -0.8 Low 2.7 1.1 0.1 -0.1 0.0 -0.1 -0.9 0.1 1.0 1.1 -0.9 -2.7 0.5 0.1 -0.4 0.2 2.8 2.3 -1.8 4.3 1.2 -0.6 1.8 -0.1 Model row 27 28 30 30 30 31 31 32 32 36 36 36 37 37 37 38 38 38 41 41 42 47 48 50 Location Model column 53 40 11 25 70 45 112 45 112 18 77 103 96 101 105 36 81 83 7 95 87 101 102 98 ROOT MEAN SQUARED Well number 203 22 327 348 11 401 420 402 419 33 303 280 365 335 279 307 359 360 108 332 321 281 282 346 Recharge conditions High 2.5 -0.6 2.5 1.8 1.9 - 1.4 1.2 3.6 -0.5 0.3 2.8 - 2.3 1.3 4.1 -1.0 -4.1 - -5.1 1.5 1.6 -0.8 1.92 Average 1.1 -0.1 2.4 2.0 1.0 2.2 0.8 2.4 0.5 0.2 -0.9 -0.8 0.9 -0.3 1.0 -1.6 1.8 -1.7 -0.8 -0.6 -1.4 0.3 0.5 -1.9 1.73 Low -0.5 1.9 0.4 - 0.4 2.2 1.0 0.9 -5.3 -1.0 -0.5 0.0 -0.3 1.1 -7.5 - - -1.0 -0.1 1.0 0.5 0.6 -2.2 1.96 Model Calibration 51 ---PAGE BREAK--- Table 12.- Measured and simulated streamflow gains and losses, for selected stream reaches during high-, average-, and low-recharge conditions in Cortland, N.Y. study area. [Values are in cubic feet per second (ft Positive numbers indicate gains; negative numbers indicate losses. Dashes indicate no measurement was made in that reach. Reach locations shown in figure 14. ] Recharge conditions High Average Measured Measured March 29, May 30-June 4, Stream Reach 1990 Simulated 1991 Simulated West Branch Tioughnioga River A Tioughnioga River A B Otter Creek A -1.2 B C 0.9 D -0.8 E 1.3 F -1.6 G 0.4 Tributary to Otter Creek A -1.8 Dry Creek A -1.0 B -0.6 C -0.6 D -0.6 E 2.0 Perplexity Creek A -0.6 Tributary to Perplexity Creek A -0.4 B -0.8 9.3 3.5 4.1 -1.7 0.0 0.8 -0.9 -0.2 -0.6 -0.6 -0.7 -0.6 -0.4 -0.3 -0.3 0.1 -0.6 -0.3 -0.1 Table 13. Steady-state water budgets for the glacial aquifer system low-recharge conditions. [Rates are in cubic feet per second, (ft3/s)] 9.2 1.9 5.7 -0.4 -0.2 0.1 0.0 0.1 -0.3 -1.2 -0.2 -0.4 0.0 -0.4 0.0 1.0 -0.2 -0.1 Dry at Cortland, N.Y 8.4 3.2 4.0 -0.4 -0.6 -0.1 -0.7 -0.1 -0.4 -0.7 -0.4 -0.4 -0.1 -0.3 0.0 -0.1 -0.1 -0.2 -0.1 for high-, Low Measured October 9, 1991 Simulated 5.9 3.8 -1.3 Dry Dry Dry Dry Dry Dry Dry Dry -0.2 -0.2 Dry Dry Dry Dry -0.05 Dry average- 5.4 2.5 2.5 -0.2 -0.1 -0.1 , and Recharge conditions Budget component A. Recharge to the aquifer system Precipitation on the aquifer Upland sources Seepage losses from tributary streams Unchanneled runoff and ground- water inflow from uplands Ground-water inflow from valleys to model area Infiltration at recharge basins TOTAL B. Discharge from aquifer system Pumping wells Discharge from aquifer to streams Ground- water outflow from model area TOTAL High March 28-29, 1990 Percent of Amount total 16.2 39 15.2 36 9.2 22 1.0 3 (LQ Q 41.6 100 10.7 26 29.5 71 L4 3 41.6 100 Average May28-June4, 1991 Percent of Amount total 14.4 12.2 8.8 .9 12 38.0 10.7 26.1 L2 38.0 38 32 23 3 4 100 28 69 3 100 Low October 7-9, 1991 Amount 9.3 10.4 5.4 .5 22 21. S 11.0 15.9 £ 27.8 Percent of total 33 38 19 2 8 100 40 57 3 100 52 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- Layer 1, (upper part of uricohtoed aejtJifff)! 1100 1280 1300 1160 1300 Layer 2 (lower part of unconfirmed aquifer) EXPLANATION Q ACTIVE MODEL AREA Q INACTIVE MODEL AREA -1160 POTENTIOMETRIC CONTOUR- shows simulated altitude at which water level would have stood in cased wells. Contour interval 10 feet unless otherwise noted. Datum is sea level. O PUBLIC-SUPPLY WELL ( ) INDUSTRIAL WELL 1 MILE 1 KILOMETER Figure 25A. Simulated head in model layers 1, 2, and 3 during steady-state high-recharge conditions in Cortland, N.Y. study area. (Locations and vertical positions of layers are shown in fig. 23.) Model Calibration 53 ---PAGE BREAK--- Layer » (upper gaftiM uncferifiOjecl aquifer) 1280 1300 LayerI8 (Ifiwerpartiof EXPLANATION Q ACTIVE MODEL AREA INACTIVE MODEL AREA POTENTIOMETRIC CONTOUR- shows simulated altitude at which water level would have stood in 1160 cased wells. Contour interval 10 feet unless otherwise noted. Datum is sea level. O PUBLIC-SUPPLY WELL ( ) INDUSTRIAL WELL 1260 1300 layer 3 1 MILE 1 KILOMETER Figure 25B. Simulated head in model layers 1, 2, and 3 during steady-state average-recharge conditions in Cortland, N.Y. study area. (Locations and vertical positions of layers are shown in fig. 23.) 54 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- 1120 1100- 1280 Layer 1 (upperpartof unconfined aquifer) 1140-- 1240 1260 1280J Layer 2 (lower part of unconfined aquifer) 1260- Layer 3 {confined aquifer) EXPLANATION Q ACTIVE MODEL AREA INACTIVE MODEL AREA -1160- POTENTIOMETRIC CONTOUR- shows simulated altitude at which water level would have stood in cased wells. Contour interval 10 feet unless otherwise noted. Datum is sea level. O PUBLIC-SUPPLY WELL ( ) INDUSTRIAL WELL 1 MILE I 1 KILOMETER Figure 25C. Simulated head in model layers 1, 2, and 3 during steady-state low-recharge conditions in Cortland, N.Y. study area. (Locations and vertical positions of layers are shown in fig. 23.) Model Calibration 55 ---PAGE BREAK--- reaches that are near the major pumping wells, such as where Otter Creek flows by the City of Cortland well field. Model Sensitivity Sensitivity analyses of the model simulating average-recharge conditions were conducted to assess which model parameters resulted in large changes, and which ones resulted in small changes, in the simulated water levels (heads) and in the streamflow gains or losses. Future data-collection efforts can be directed to those aquifer properties to which the model is most sensitive. Recharge, horizontal hydraulic conductivity of the unconfined aquifer, vertical hydraulic conductivity of the streambed, and conductance between the uncon- fined aquifer (layer 2) and the confined aquifer (layer 3) were varied one at a time, and the effect on calculated heads (fig. 26) and on gains or losses in streamflow in reaches of the West Branch Tioughnioga River and Tioughnioga River (fig. 27) were noted. The vertical axis in figure 26 shows the root mean square of the difference between the computed and observed heads at 48 observation wells. The root mean square of the difference between calculated and measured head was 1.29 for the final calculated model (multiplication factor equal to 1 in the graph); all sensitivity analyses for multiplication factors other than 1 had root mean squares greater than 1.29. Results of these analyses indicate that the model is relatively sensitive to recharge and horizontal hydraulic conductivity of the unconfined aquifer, moderately sensitive to vertical hydraulic conductivity of the streambed, and insensitive to vertical conduc- tance between the unconfined and confined aquifers i ' r ~i I i r i i I r~ =J DC Z> ill O u. LU LU HI OQ O LU O UJ Q U. < LI; UJ Q X U- O O =i 3 < CO 111 ^ O O DC EXPLANATION Recharge Hydraulic conductivity of the unconfined aquifer Vertical-hydraulic conductivity of the streambed Vertical conductance between unconfined and confined aquifers 0.1 0.2 0.3 0.4 0.5 0.7 1 MULTIPLICATION FACTOR 10 Figure 26. Results of sensitivity analyses for hydraulic head in unconfined aquifer in Cortland, N.Y. study area. (Root-mean squared calculated at 48 model cells containing observation wells.) 56 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- O m O 1U LLI CO DC LU 0_ LU 9 ? 6 <0 A. WEST BRANCH TIOUGHNIOGA RIVER EXPLANATION Recharge Vertical-hydraulic conductivity of the streambed Hydraulic conductivity of the unconfined aquifer Vertical conductance between unconfined_ and confined aquifers Measured gain in streamflow 0.1 10 O O LU q CO y DC LLI CL LLI LU LL g 8 CO o 87 6 0.2 0.3 0.4 0.5 0.7 1 2 MULTIPLICATION FACTOR 10 EXPLANATION Recharge Vertical-hydraulic conductivity of the streambed Hydraulic conductivity of the unconfined aquifer Vertical conductance between unconfined and confined aquifers Measured gain in streamflow B. TIOUGHNIOGA RIVER 5 0.1 0.2 0.3 0.4 0.5 0.7 1 2 MULTIPLICATION FACTOR 10 Figure 27. Results of sensitivity analyses for ground-water discharge to major streams during average- recharge conditions in the glacial-drift aquifer in the Cortland, N.Y. study area: A. Reach A of West Branch Tioughnioga River. B. Reaches A and B of Tioughnioga River. (Reach locations are shown in fig. 14.) Model Sensitivity 57 ---PAGE BREAK--- (fig. 26). The most sensitive areas in the model are those with low hydraulic conductivity (kame deposits). Neither the vertical hydraulic conductivity of the streambed, nor the vertical leakance between the unconfined and confined aquifers could be measured directly, and both can vary over a wide range of values; therefore, the values of these parameters were tested over a greater range than the others. Results of the sensitivity analyses for streams (fig. 27) indicated that the ground-water discharge to West Branch Tioughnioga River and to Tioughnioga River was relatively sensitive to changes in recharge, verti- cal-hydraulic conductivity of the streambed, and horizontal hydraulic conductivity of the unconfined aquifer. The simulated heads and streamflow were relatively insensitive to vertical conductance between the unconfined and confined aquifer. Model Applications Delineation of contributing areas to wells by use of numerical flow models has a high potential for accuracy because models can incorporate many of the hydrogeologic factors that affect ground-water flow, such as aquifer geometry, hydraulic conductivity, recharge rate, and pumping rate. Particle tracking provides a simple means of evaluating the advective- transport characteristics of ground-water systems, including computation of the flowpath and travel times of the advective phase of contaminants. Areas Contributing Recharge to Municipal Wells Pumping large quantities of ground water from an aquifer system causes a drawdown within the aquifer, and the drawdown decreases with increasing distance from the pumping well. The lowering of head that results from pumping causes ground water to flow to the well, and the flowpaths to the well depend on the hydrogeologic characteristics of the flow system, the well location and pumping rate, the system boundaries, and the rate and distribution of recharge to the aquifer system. Factors that affect the areas contributing recharge to wells are described in detail in Reilly and Pollock (1993). Areas contributing recharge to major pumping wells in the glacial-aquifer system in the study area were delineated for the periods of high, average, and low recharge described previously. The ground-water flowpaths for the three conditions were calculated by the MODPATH program (Pollock, 1989) from the output from MODFLOW (McDonald and Harbaugh, 1988). The simulated areas contributing water to municipal wells and to the recovery well at the typewriter plant are delineated in plates 2, 3, and 4. A porosity of 0.3 for the aquifers and confining unit was used for MODPATH calculations. The contributing-area analyses indicate that the contributing areas to pumping wells are U-shaped (open end facing upgradient of the well) and extend over most of the unconfined aquifer upgradient from the wells. In general, the contributing areas resulting from the low-recharge simulation were the largest, and those resulting from the high-recharge simulation were the smallest (pi. 2, and The City of Cortland munic- ipal well had the largest contributing area in all three simulations. Recharge from tributaries such as Otter Creek and the unnamed tributary north of Otter Creek were important sources of water to the city well; thus, evaluations of the quality of water pumped by the well need to consider water quality in these two streams. At present, land use in the proximal parts of the contribut- ing area to the city well field is mostly residential and forest, and land use in the distal parts is mostly residential and commercial, with some industry and agriculture. The typewriter plant in the western part of the aquifer was not part of the contributing area to the city well field in any simulations when the purge well at the plant was pumping (pis. 2,3, and but when the purge well was turned off, the industrial site was part of the contributing area in the high- and average- recharge simulations. The contributing area to the Town of Cortlandville well field at Terrace Road is relatively small. This well field also receives recharge from a reach of Otter Creek (pis. 2, 3, and Land use within the contribut- ing area is mostly commercial. Chemical spills near the contributing area have come close to contaminat- ing this well field, and, as a result, the Town has installed another municipal well at Lime Hollow Road, which is upgradient of most of the commercial and industrial areas. The municipal well for the Town of Cortlandville at Lime Hollow Road began pumping in 1991; there- fore, the contributing area for this well during high- recharge condition of 1990 was not simulated. The contributing areas for this well under average- and low-recharge conditions extend to the ground-water divide in the southwestern part of the study area and contain several ponds (pis. 3 and Land use within the contributing area is mostly agricultural and forest, 58 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- with some residential areas and one industry. This well's contributing area is in the least developed part of the study area and, therefore, should be the least threatened by urban sources of contamination. Flowpaths to Wells and Streams from Sources of Contamination The flowpaths and traveltimes of the advective phase of contaminants in the study area were deter- mined through the MODPATH program, which computes paths (tracks) of imaginary "particles" of water moving through a simulated ground-water system and keeps track of their traveltime. Particle tracking provides a simple means of evaluating the advective-transport characteristics of ground-water systems. Advective-flow models cannot be used to compute solute concentrations in ground water because they do not account for the effects of disper- sion, adsorption, chemical reactions, or other transport phenomena, but they are useful intermediate step between ground-water flow models and solute-trans- port models. Chemical dispersion typically cause contaminant plumes to be larger than indicated by advective-flow models. Particles were tracked from the contaminated sites at the typewriter plant in the western part of the study area and at the Superfund site in the southeastern part; and flow velocities were calculated from an average porosity of 0.3. Typewriter Plant Movement of particles from the typewriter plant to discharge points was tracked in simulations in which the purge well was pumping to represent conditions during remediation work; the resulting flowpath analy- ses indicated that the TCE spill is within the contribut- ing area of the purge well. Particles were then tracked in simulations in which the purge well was not pumping, to represent the flowpath of the advective movement of TCE during the several years before the remedial pumping began. Particles were applied to the water table (top face of cell in layer 1) at the cell repre- senting the TCE spill; the resulting flowpaths in the unconfined aquifer during high-, average-, and low- recharge conditions are shown in figures 28A, B, and C. Ground water flows northeastward from the spill area to the center of the aquifer. Flowpaths shifted progressively southward from a northeastward route during high-recharge conditions to more eastward paths during average- and low-recharge conditions, respectively, as a result of differences in the distribu- tion of recharge during those conditions. The discharge areas of ground water that flows from the TCE spill vary according to recharge condi- tions. During high- and average-recharge conditions, ground water flows 2.25 mi northeastward, then discharges to the City of Cortland municipal well (fig. 28A, B) and to a small pond near the municipal well. The flowpaths shift to the south during average-recharge conditions, and the traveltime of ground water flowing from the typewriter plant to the municipal well is about 4 years. Simulated ground-water flowpaths from the TCE spill site shift even farther southward during low- recharge conditions and do not end at the City's municipal well or at the pond, but discharge into the Tioughnioga River (fig. 28C), 3.7 mi northeast of the spill site. Traveltime is about 7 years from the spill site to the Tioughnioga River. Superfund Site The movement of particles from the Rosen Super- fund site also were tracked from the water table to delineate the flowpath and traveltime of ground water migrating from the site to the discharge area (fig. 28A, B, Ground water flows about 1 mi northeastward from the site to the central part of the valley, where it bends to the southeast and discharges into the Tiough- nioga River about 1 mi east of the site. Traveltime of ground water from the site to the Tioughnioga River ranges from 3 years during high-recharge conditions to just under 4 years during low-recharge conditions. The flowpaths shift progressively from northeastward paths during high-recharge conditions to more south- ward traces during average- and low-recharge condi- tions (figs. 28A, B, The particle-tracking analyses indicate that some ground water may also discharge to an industrial well owned by ETL, Inc. on the east side of the river (fig. 28A, B, Trichloroethane (TCA) is one of the contaminants found in ground water at, and migrating from, the Superfund site (Blasland, Bouck and Lee, Engineers, 1992); it also was found at the ETL well. A detailed study would be needed to deter- mine whether the TCA at the ETL well comes from the "Superfund" site, the ETL property, or some other source. Flowpaths that originated in the unconfined aquifer at the Superfund site under the three simulated conditions extended through the model cell that contains USGS test well 332 (local well number 90- Model Applications 59 ---PAGE BREAK--- EXPLANATION D ACTIVE MODEL AREA Q INACTIVE MODEL AREA D STREAM CELL O PUBLIC SUPPLY WELL ® INDUSTRY WELL FLOW PATH IN UNCONFINED AQUIFER-boxes correspond 2 to travel time of one year, numbers are cumulative FLOW PATH IN CONFINED ' AQUIFER-boxes correspond 20 to traveltime of ten years, numbers are cumulative 1MILE I \ 1 KILOMETER Figure 28. Flowpaths and traveltime of ground water moving from two contaminant sources in Cortland, N.Y. study area: A. Under high-recharge conditions. B. Under average-recharge conditions. C. Under low-recharge conditions. (Location is shown in fig.24.) 60 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- IS, pi. in which TCA, TCE, and trans-DCE were detected in concentrations of 107,15.2, and 28.9 [ig/L, respectively (appendix Although the particle-track- ing analyses indicate that those contaminants have migrated to well 332 and beyond, their origin at the Superfund site is uncertain because the area between well 332 and the Superfund site contains other poten- tial sources of these contaminants. Particles were then tracked from the bottom of layer 3 at the Superfund site to simulate conditions where the confining layer is absent along the valley wall (layer 3 is in hydraulic connection to layers 1 and 2) and where a DNAPL could readily sink from land surface to the bottom of the layer 3, releasing a dissolved-phase contaminant to ground water. Whether the Superfund site contains a DNAPL is uncertain; therefore this scenario is hypothetical. Parti- cles were not placed in cells of layer 3 where it is confined because contaminants would not move readily through the confining layer. Flowpaths in the confined aquifer initially trend northeastward from the "Superfund" site, then bend to the southeast, where they exit the modeled area as underflow through the Tioughnioga River valley (fig. 28A, B, Traveltime from the Superfund site to the eastern edge of the modeled area ranged from 30 years under high- recharge conditions and to more than 70 years under low-recharge conditions. SUMMARY Glacial aquifers in the Otter Creek-Dry Creek Valley and in parts of the adjacent West Branch, East Branch, and Tioughnioga River Valleys are the sole source of water for the City of Cortland and surround- ing communities. Several parts of the aquifer system have been contaminated by: solvents and degreas- ers, including trichloroethylene (TCE), trichloroethane (TCA), and dichloroethene (DCE), gasoline from leaking storage tanks at least at two service stations, bacteria from failing septic systems, and leachate from a "Superfund" site. The USGS, in cooperation with the Cortland County Departments of Planning and Health, studied the hydrogeology and water quality of a glacial-aquifer system in Cortland County during 1989-93 and simulated ground-water flow to delineate areas contributing recharge to munic- ipal wells and to the ground-water flowpaths from two chemical-spill sites. The unconsolidated deposits in the study area were deposited between 10,000 and 23,000 years ago during Late Wisconsinan glaciation. The central parts of the valley contain kames deposits that form a confined aquifer where they are overlain by fine- grained lacustrine sediments. The kame deposits are typically 60 to 170 ft thick in the western and eastern parts of the valley. Kames are overlain by outwash where the confining layer is absent in some places along the edges of the valley. A large moraine system (Valley Heads Moraine) formed in central and western New York valleys during a major standstill of the ice front 14,000 and 14,900 years ago. The western part of the study area contains a Valley Heads moraine in the Otter Creek- Dry Creek valley; the moraine is a heterogeneous deposit consisting of coarse sand and gravel in the upper part, and till and lacustrine deposits in the lower part. A proglacial lake formed in valleys of the Tiough- nioga River basin during deglaciation. Fine-grained sediments that were deposited within this lake range from 90 ft thick in the eastern part of the study area to 150 ft in the northern and central parts, and to 170 ft in the southwest part. This unit extends throughout the study area except where it pinches out along the edges of the valley. After the proglacial lake drained from the study area, glacial meltwaters from the Valley Heads ice deposited large amounts of outwash atop the lacustrine unit. The outwash grades from coarse boulder gravel near the ice-front location in the southwestern part of the study area to coarse cobble-and-pebbly sand and gravel in distal reaches in the central and eastern parts. The glacial-aquifer system consists of 40 to 80 ft of unconfined sand and gravel (mostly outwash) that overlies a confining layer (lacustrine deposits and till), that ranges from 1 to 155 ft thick. This unit, in turn, overlies confined sand and gravel (kame deposit) that ranges from 0 to 170 ft thick. The confining unit impedes ground-water movement between the uncon- fined and confined aquifers in the middle of the valley, but the two aquifers are hydraulically connected wherever the confining layer is absent along the valley walls. Hydraulic conductivity of the unconfined aquifer, as determined from aquifer tests at large pumping wells, ranges from 85 to 1,150 ft/d, and hydraulic conductivity of the confined aquifer, as determined from aquifer tests where range from 60 to 65 ft/d at a fish hatchery in the western part of the Summary 61 ---PAGE BREAK--- study area and from 3 to 140 ft/d at the Superfund site in the eastern part. Water levels were measured in about 100 wells during three periods-early spring (March 28-29, 1990), late spring and early summer (May 28 through June 4,1991), and fall (October 7-9,1991) to document ground-water levels during high-, average-, and low- recharge periods, respectively. Data from these three periods were used to calibrate the ground- water-flow model. The unconfined aquifer supplies four major pumping centers- the City of Cortland well field in the central part of the study area, the Town of Cortlandville well field at Terrace Road, the Town of Cortlandville municipal well at Lime Hollow Road, and a purge well at a typewriter production plant in the western part. Water in the unconfined aquifer generally moves from the edges of the valley toward the center, then northeastward along the axis of the Otter Creek-Dry Creek valley, where it discharges to pumping wells, West Branch Tiough- nioga, and Tioughnioga Rivers. Water in the confined aquifer generally follows the direction of flow in the unconfined aquifer. The aquifer system receives recharge from three sources under natural (nonpumping) conditions: direct infiltration of precipitation on the aquifer, runoff from unchanneled hillsides that border the aquifer, and seepage from tributary streams that flow over the aquifer. The unconfined aquifer receives additional recharge from two sources under pumping conditions- treated pumped water from the typewriter plant that infiltrates to the aquifer at recharge basins, and induced infiltration from streams and ponds near the major pumping wells. Average annual recharge to the aquifer system from precipitation is 23.0 in. Most tributary streams in the study area lose water to the aquifer where they flow into the main valley, except for some reaches in Otter Creek. Only part of the streamflow in upland tributaries seeps into the aquifer during high-flow conditions during the spring and during large storms throughout the year, however; the rest flows over the aquifer and discharges into the West Branch Tioughnioga and Tioughnioga Rivers. Most tributary streams lose all their water to the aquifer during low-flow conditions in the summer, fall, and winter, except for Dry Creek, which goes dry only during exceptionally dry periods. Losing streams typically dry up in the upstream direction, whereas gaining streams typically dry up starting at the headwaters. Ground water discharges from the unconfined aquifer system by seeping into major streams and seasonally into some reaches of Otter Creek; flowing to pumping wells; moving as underflow along the Tioughnioga River valley; and seeping to springs that form the headwaters to the Fall Creek Valley in the western part of the study area. Most water in the unconfined aquifer either discharges into streams or is pumped from municipal wells, and most of the water pumped from municipal wells is eventu- ally treated at the sewage-treatment plant, then discharged into the Tioughnioga River. Ground-water withdrawals from the unconfined aquifer through municipal and industrial wells range from 6.76 to 7.20 Mgal/d. The largest ground-water user is the City of Cortland which, during 1984-92, pumped from 3.9 to 4.3 Mgal/d. The direction of ground-water flow in the confined aquifer generally is similar to that in the unconfined aquifer. The extent of a TCE plume in the unconfined aquifer was determined for three sampling periods (April 4-5,1990, September 17-20,1990, and April 27,1993); results indicate that TCE has migrated 1.25 mi northeastward from its source at the typewriter plant in the western part of the study area. The extent of the plume was similar during all three sampling periods, indicating that steady-state conditions have been reached. Little or no TCE was detected in the confined aquifer. TCE concentrations were highest during 1987-88; then decreased from 1989 through mid-1992, and dropped significantly at the end of the 1992, probably in response to removal of contaminated soil and pumping of a recovery well at the source. The "pump and treat" remedial program at the typewriter plant was expected to significantly lower the TCE concentrations throughout the aquifer by April 1993, but the results of the April 1993 sampling showed the concentrations at most wells in the middle and distal parts of the plume had increased signifi- cantly since the preceding fall and were nearly as high as the maximum values in 1987-89. The increase in April 1993 is attributed to desorption of TCE from sediments that had been unsaturated for 10 years, when the water table rose to its highest levels in 10 years in response to above-normal recharge in March and April, 1993. TCE concentrations at wells within 1,500 ft of the source fluctuated seasonally; they were highest during the spring and lowest during the summer and fall. TCE 62 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- concentrations at wells more than 1,500 ft from the source showed little or no seasonal trend. Detection of a TCE degradation product (c/s-1,2- DCE) in most wells in the plume indicates biologi- cally mediated reductive dehalogenation. The fate of TCE after biodegradation to cw-l,2-DCE is uncer- tain; other degradation compounds either were not detected (vinyl chloride) or were detected only on rare occasion in trace amounts (1,2-DCA). The end degradation products of TCE are carbon dioxide, water, and chloride. Ground-water samples were collected from both aquifers during April and September 1990 and analyzed for inorganic and organic chemical constitu- ents; results indicate that the quality of water generally meets New York State drinking-water standards, except for part of the unconfined aquifer that is contaminated by TCE. Elevated concentrations of chloride are generally found in ground water near major roads that are heavily salted. Median concen- trations of chloride during the April and September 1990 samplings, were 37 and 39 mg/L, respectively. The trend of increasing chloride concentrations since the early 1940's at the City of Cortland well field, noted by Buller (1978), has continued. Nitrate concentrations in ground water ranged from 0.1 mg/L to 12.0 mg/L, with a median concentra- tion of 4.3 mg/L. The concentrations were higher in the agricultural areas than elsewhere. A quasi-three-dimensional, digital ground-water flow model was constructed to simulate hydraulic head, estimate the contributing areas for municipal wells, and delineate the flowpaths of ground water migrating from two sources of contamination. The aquifer system was represented as three layers-the upper two represent the unconfined aquifer, and the third represents the confined aquifer. The resulting areas contributing recharge to wells are U-shaped and extend over most the aquifer upgradient from the pumping wells. The largest contributing areas resulted from low-recharge simulations, and the smallest resulted from high-recharge simulations. The City of Cortland municipal well had the largest contributing area in all three simulations. Otter Creek and an unnamed tributary north of Otter Creek are within the contributing area to the city's wells; thus, their water can affect the municipal ground-water supply. The contributing area for the Town of Cortland- ville well field at Terrace Road is relatively small. The well receives recharge from a reach of Otter Creek. The contributing area for the Town of Cortlandville municipal well on Lime Hollow Road extends to the ground-water divide in the southwestern part of the study area. The contributing area for the well on Lime Hollow Road is in the least developed part of the study area and, therefore, is expected to be only minimally affected by contamination from urban sources. The flowpaths and traveltimes of ground water moving from two major chemical-spill sites (the typewriter plant in the western part of the study area and the Superfund Site in the southeastern part) were determined through MODPATH, a particle-tracking program. Results indicate that, during high- and average-recharge conditions, ground water from the typewriter plant discharges to the City of Cortland municipal well, 2.25 mi to the northeast. No signifi- cant concentrations of TCE (above 5 Iig/L) were detected at the municipal well because degradation and volatilization had reduced them below this level within 1.25 mi of their source. Traveltime from the source to the municipal well is about 4 years. During low-recharge conditions, flowpaths shift southward, bypass the municipal well, and end at the Tioughnioga River. Traveltime from the source to the river is about 7 years. Ground water from the Superfund site flows to the central part of the unconfined aquifer and discharges into the Tioughnioga River about 1 mi to the east. Traveltime from the source to the Tioughnioga River ranged from 3 to 4 years. Summary 63 ---PAGE BREAK--- REFERENCES CITED Apfel, E. 1967, Monarch-Edlund water wells as of December 1,1967, South Cortland, Cortland County, New York: Apfel and Associates, Inc., Syracuse, N.Y., 20 p. Asselstine, E.S., 1946, Progress report on the ground-water conditions in the Cortland quadrangle, New York: New York State Water Power and Control Commis- sion Bulletin GW-16,49 p. Barbee, G.C., 1994, Fate of chlorinated aliphatic hydrocar- bons in the vadose zone and ground water: Ground Water Monitoring Review, v.14, no. 1, p. 129-140. Barrio-Lage, Parsons, F.Z., Nasser, R.S., and Lorenzo, P.A., 1986, Sequential dehalogenation of chlorinated ethenes: Environmental Science and Technology, v. 20, no. 1, p. 96-99. Behnke, J.J., 1974, A summary of nitrogen compounds in ground water: Jour, of Hydrology, v. 27, p. 155-167. Blasland, Bouck and Lee, Engineers, 1992, Remedial inves-tigation report, Rosen Site, Cortland, New York: Syracuse, N.Y., Blasland, Bouck and Lee Engineers, 210 p. ____1995, Natural attenuation at the Rosen Brothers Superfund Site, Cortland, New York: Syracuse, N.Y., Blasland, Bouck and Lee Engineers, 19 p. Buchanan, T.J., and Somers, W.P., 1969, Discharge measurements at gaging stations: U.S. Geological Survey Techniques of Water-Resources Investiga- tions, book 3, chap. A8,65 p. Buller, William, Nichols, W.J., and Harsh, J.F., 1978, Quality and movement of ground water in the Otter Creek-Dry Creek basin, Cortland County, New York: U.S. Geological Survey Water-Resources Investiga- tions, Open-File Report 78-3, 63 p. Campbell, J.B., Szabo, C.O., Sherwood, D.A., and Deloff, D.D., 1992, Water Resources Data, New York, Water Year 1991, v. 3, Western New York: U.S. Geological Survey Water Data Report NY-91-3,200 p. Cho, H.J., Jaffe, P.R., and Smith, J.A., 1991, Design of a field experiment to study the dynamics of trichloro- ethylene vapors in the unsaturated zone during an infiltration event, in Mallard, G.E. and Aronson, D.A., (eds.), 1991, U.S. Geological Survey toxic Substances Hydrology Program Proceedings of the Technical Meeting, Monterey, California, March 11- 15,1991: U.S. Geological Survey Water-Resources Investigations Report 91-4034, p. 721-725. Conover, W.J., and Iman, R.L., 1981, Rank transformations as a bridge between parametric and nonparametric statistics: American Statistician, v. 35, p. 124-129. Cosner, O.J. and Harsh, J.F., 1978, Digital-model simu- lation of the glacial-outwash aquifer, Otter Creek- Dry Creek basin, Cortland County, New York: U.S. Geological Survey Water-Resources Investigations- Open-File Report 78-71, 34 p. Cressey, G.B., 1966, Land Forms, in Thompson, J.H., Geography of New York State: Syracuse University Press, fig. 9, p. 26. Davis, Andy, and Olsen, R.L, 1990, Predicting the fate and transport of organic compounds in ground water (Part Hazardous Material Control, v. 3, no. 3, p. 39-64. Dethier, B. 1966, Precipitation in New York State: New York State College of Agriculture at Cornell Univer- sity, Ithaca, N.Y, Bulletin 1009, 78 p. Ehlke, T.A., Imbrigiotta, T.E., Wilson, B.H., and Wilson, J.T., 1991, Biotransformation of cis-l ,2-dichloro- ethylene in aquifer material from Picattiny Arsenal, New Jersey, in Mallard, G.E. and Aronson, D.A. (eds.), 1991, U.S. Geological Survey Toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Monterey, California, March 11- 15,1991: U.S. Geological Survey Water-Resources Investigations Report 91-4034, p. 689-697. Farrell, James, and Reinhard, Martin, 1994, Desorption of halogenated organics from solids, sediments, and soil under unsaturated conditions: Environment Science and Technology, v. 28, no. 1, p. 53-62. Fullerton, D.S., 1980, Preliminary correlation of post-Erie interstadial events (16,000-10,000 radiocarbon years before present), Central and Eastern Great Lakes Region, and Hudson, Champlain, and St. Lawerence Lowlands, United States and Canada: U. S. Geological Survey Professional Paper 1089,52 p. Galson Technical Services, Inc., 1988, Subsurface investi- gation for Monarch Cortland facility, South Cortland, New York: East Syracuse, N.Y, Galson Technical Services, Inc., 15 p. Haeni, P.P., 1988, Application of seismic-refraction techniques to hydrologic studies: U.S. Geological Survey Techniques of Water-Resources Investiga- tions, book 2, chap. D2, 86 p. Heath, R.C., 1983, Basic ground-water hydrology: U.S. Geological Survey Water-Supply Paper 2220, p. 13. Imbrigiotta, T.E., Ehlke, T.A., and Martin, Mary, 1991, Chemical evidence affecting the fate and transport of chlorinated solvents in ground water at Picatinny Arsenal, New Jersey in Mallard, G.E. and Aronson, D.A. (eds.), 1991, U.S. Geological Survey toxic Substances Hydrology Program-Proceedings of the Technical Meeting, Monterey, California, March 11- 15, 1991: U.S. Geological Survey Water-Resources Investigations Report 91-4034, p. 681-688. Isachsen,Y.W., Landing Ed, Lauber, J.M., Rickard, L.V., and Rogers, W.B., 1991, Geology of New York-a simplified account: New York State Museum-Geolog- ical Survey Educational Leaflet no. 28,284 p. Jeffers, P.M., Ward, L.M., Woytowitch, L.M., and Wolfe, N.L., 1989, Homogeneous hydrolysis rate constants for selected chlorinated methanes, ethanes, ethenes, and propanes: Environmental Science and Technol- ogy v. 23, no. 8, p. 965-969. 64 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- Kennedy, E.J., 1990, Levels at streamflow gaging stations: U.S. Geological Survey Techniques of Water- Resources Investigations, book 3, chapter A19, 27 p. MacKay, D.M., Roberts, P.V., and Cherry, J.A., 1985, Transport of organic contaminants in ground water: Environmental Science and Technology, v. 19, no. 5, p. 384-392. McCarty, 1991, Engineering concepts for in-situ bioreme- diation, in Bioremediation, Fundamentals and effec- tive applications-Third annual symposium of the Gulf Coast Hazardous Substance Research Center, Lamar University, Beaumont, Texas: February 21-22, 1991 p. 3-13. McDonald, M.G., and Harbaugh, A.W., 1988, A modular three-dimensional finite-difference ground-water flow model: Techniques of Water-Resources Investi- gations of the U.S. Geological Survey, book 6, chap. Al,586p. Miller, T.S., Brooks, T.D., and Stelz, W.G., 1981, Geo- hydrology of the valley-fill aquifer in the Cortland- Homer-Preble area, Cortland and Onondaga Counties, New York; U.S. Geological Survey Open- File Report 81-1022, 7 sheets, scale 1:24,000. Morrissey, D.J., Randall, A.D., and Williams, J.H., 1987, Upland runoff as a major source of recharge to strati- fied drift in the glaciated northeast, in Regional aquifer systems of the United States-The northeast glacial aquifers: American Water Resources Associa- tion, AWRA monograph series no. 11, p. 17-36. Muller, E.H., 1970, Glacial geology and geomorphology between Cortland and Syracuse, in New York State Geological Association, 42nd annual meeting, Syracuse, N.Y, p. 1-15. Muller, E.H., Braun, D.D., Young, R.A., and Wilson, M.P., 1988, Morphogenesis of the Genesee Valley, in Northeastern Geology, v. 10, no. 2, p. 112-133. Muller, E.H., and Calkin, P.E., 1993, Timing of Pleistocene glacial events in New York State, in Canadian Journal of Earth Science, v. 30, p. 1829-1845. National Academy of Sciences, 1977, Drinking Water and Health: National Research Council, Washington D.C., 939 p. National Oceanic and Atmospheric Administration, 1990, Climatological Data-Annual Summary, New York, 1990: Asheville, N.C., National Climate Data Center, v. 102, no. 13, 36 p. 1991, Climatological Data-Annual Summary, New York, 1991: Asheville, N.C., National Climate Data Center, v. 103, no. 13, 36 p. 1992, Climatological Data-Annual Summary, New York, 1992: Asheville, N.C., National Climate Data Center, v. 104, no. 13, 36 p. 1993, Climatological Data-Annual Summary, New York, 1993: Asheville, N.C., National Climate Data Center, v. 105, no. 13, 36 p. Neuman, S.P., 1974, Effect of partial penetration on flow in unconfined aquifers considering delayed gravity response: Water Resources Research, v. 10, no. 2, p. 303-312. O'Brien and Gere Engineers, Inc., 1987, Site investigation and interim remedial action plan: Syracuse, N.Y, O'Brien and Gere, Engineers, Inc., 28 p. 1990, Supplemental site investigation, Smith Corona Corporation: O'Brien and Gere, Engineers, Inc., Syracuse, New York, 27 p. Olsen, R.L. and Davis, Andy, 1990, Predicting the fate and transport of organic compounds in ground water (pt. Hazardous Material Control, v. 3, no. 4, p. 18- 37. Plumb, R.H., 1991, The occurrence of Appendix IX organic constituents in disposal site ground water in Ground Water Monitoring Review, v. 11, no. 2, p. 157-164. Pollock, D.W., 1989, Documentation of computer programs to compute and display pathlines using results from the U.S. Geological Survey modular three-dimensional finite-difference ground-water flow model: U.S. Geological Survey Open-File Report 89-622, 49 p. , 1994, Users guide for MODPATH/ MODPLOT, Version 3 A particle tracking post-processing package for MODFLOW, the U.S. Geological Survey finite-difference ground-water flow model: U.S. Geological Survey Open-File Report 94-464, 246 p. Prudic, D.E., 1989, Documentation of a computer program to simulate stream-aquifer relations using a modular, finite-difference, ground-water flow model: U.S. Geological Survey Open-File Report 88-729, 113 p. Randall, A.D., 1972, Records of wells and test borings in the Susquehanna River basin, New York: New York State Department of Environmental Conservation Bulletin 69, 92 p. Lyford, P.P. and Cohen, A.J., 1987, Estimation of water available for recharge to sand and gravel aquifers in the glaciated northeastern United States, in Randall, A.D., and Johnson, A.I., (eds.) Regional aquifer systems of the United States- The northeast glacial aquifers: American Water Resources Association, AWRA monograph series no. 11, p. 37-42. Randall, A.D., Snavely, D.S., Holecek, T.J., and Waller, R.M., 1988, Alternative sources of large seasonal ground-water supplies in the headwaters of the Susquehanna River basin, New York: U.S. Geological Survey Water-Resources Investigations Report 85- 4127, 121 p. Reilly, T.E. and Pollock, D.W., 1993, Factors affecting areas contributing recharge to wells in shallow aquifers: U.S. Geological Survey Water-Supply Paper 2412, 21 p. Resource Engineering, 1987, Site Investigation Report of J. M. Murray Center for the handicapped, Inc: Cortland, N.Y, Resource Engineering, 16 p. References Cited 65 ---PAGE BREAK--- Reynolds, R.J., 1985, Hydrogeology of the surficial out wash aquifer at Cortland, Cortland County, New York: U.S. Geological Survey Water-Resources Investigations Report 85-4090,43 p. 1987, Diffusivity of a glacial-outwash aquifer by the floodwave-response technique: Ground Water, v. 25, no. 3, p. 290-299. Roffner, J.A., 1985, Climate of the States: Gale Research Company, Detroit, Michigan, 1572 p. Scott, J.H., Tibbetts, B.L., and Burdick, R.G., 1972, Computer analysis of seismic refraction data: U.S. Bureau of Mines Report of Investigation 7595, 95 p. Tepper, D.H., Morrissey, D.J., Johnson, C.D., and Maloney, T.J., 1990, Hydrogeology, water quality, and effects of increased pumpage of the Saco River Valley Glacial aquifer- Bartlett, New Hampshire to Fryeburg, Maine: U.S. Geological Survey Water- Resources Investigations Report 88-4179,113 p. Thomas, M.P., 1966, Effect of glacial geology upon the time distribution of streamflow in eastern and south- ern Connecticut: U.S. Geological Survey Profes- sional Paper 550-B, p. B209-B212. U.S. Environmental Protection Agency, 199la, Final rule, National primary drinking-water regulations-- organic chemicals and inorganic chemicals (sections 141.11, 141.12,141.32,141.50, 141.51, 141.61, and 141.62 of part 141 143.3 of part 143): U.S. Federal Register, v. 56, no. 20, January 30, 1991,p.3526-3597. , 1979, Methods of chemical analysis of water and wastes: Environmental Monitoring and Support Laboratory, Cincinnati, Ohio, EPA-600/4-79-020, 430 p. U.S. National Oceanic and Atmospheric Administration, 1990, Climatological data - annual summary, New York, 1991: U.S. Department of Commerce, National Climatic Center, 28 p. 1991, Climatological data - annual summary, New York, 1991: U.S. Department of Commerce, National Climatic Center, 28 p. 1992, Climatological data - annual summary, New York, 1991: U.S. Department of Commerce, National Climatic Center, 28 p. 1993, Climatological data - annual summary, New York, 1990: U.S. Department of Commerce, National Climatic Center, 28 p. Wilson, B.H., Ehkle, T.A., Imbrigiotta, T.E., and Wilson, J.T., 1991, Reductive dechlorination in anoxic aquifer material from Picatinny Arsenal, New Jersey, in Mallard, G.E. and Aronson, D.A., 1991, U.S. Geological Survey Toxic Substances Hydrology Program Proceedings of the Technical Meeting, Monterey, California, March 11-15, 1991: U.S. Geological Survey Water-Resources Investigations Report 91-4034, p. 704-707. Woodrow, D.L., and Sevon,W.D., 1985, The Catskill Delta: Geological Society of America Special Paper 201, 246 p. 66 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- Appendix 1. Records of wells and test borings in the Cortland, N.Y., study area. [Well locations shown on pi. Appendix USGS well no. 3 4 5 6 8 9 10 11 13 15 20 22 23 24 33 39 40 47 48 102 105 Local identi- fiier TH3 CP4 CP5 TH6 CP8 CP9 CP10 CT 11 CP13 CP15 CT20 CT22 CT23 CT24 CP1 CT ID CT1S CT2D CT2S C-102 CT5D Lati- tude 423548 423454 423507 423545 423526 423527 423535 423518 423650 423704 423414 423429 423346 423314 423323 423558 423558 423548 423548 423541 423447 Longi- tude 0761155 0761242 0761309 0761152 0761131 0761132 0761140 0761141 0761140 0761108 0761428 0761317 0761332 0761415 0761403 0761215 0761215 0761153 0761153 0761147 0761306 Owner City of Cortland Town of Cortlandville Town of Cortlandville City of Cortland Leonard Barker Leonard Barker Curtis Cortland County Cortland County Cortland County Cortland County Cortland County Cortland County Cortland County Monarch Tool Corp. Cortland County Cortland County Cortland County Cortland County City of Cortland Gutchess Lumber Date drilled (mo d yr) 08-16-79 12-29-75 03-04-76 08-29-79 1974 1976 1974 12-04-75 11-01-73 02-27-76 09-10-76 09-15-76 09-20-76 09-28-76 11-10-75 09-29-75 09-29-75 10-01-75 10-01-75 10-09-75 Elevation land- surface (feet) 1,136 1,173 1,172 1,136 1,146 1,146 1,149 1,154 1,116 1,114 1,247 1,186 1,228 1,270 1,259 1,155 1,155 1,138 1,138 1,137 1,170 Well depth (feet) 55 63 40 55 20 20 20 60 25 ~ 160 45 85 105 209 143 44 49 25 45 59 Elevation Depth to of top of bedrock casing (feet) (feet) 1,137.67 1,173.45 1,172.64 1,137.24 1,146.2 1,146.2 106 1,148.6 1,156.5 1,116.6 1,114.3 124 1,247.11 1,186.38 - ~ 207 1,154.96 1,155.39 1,137.38 1,137.03 1,138.59 1,170.89 Water level below land surface Feet 11.95 19.50 6.55 11.60 12.97 12.13 14.06 20.90 7.80 5.57 68.20 22.91 58.30 58.27 17.57 15.06 7.55 7.15 7.68 10.62 Date 08-16-79 12-29-75 03-04-76 08-29-79 02-27-76 02-27-76 02-27-76 12-15-75 11-01-73 02-27-76 09-10-76 09-15-76 09-20-76 11-10-75 11-24-75 11-24-75 10-07-75 11-24-75 08-30-76 12-01-75 Reported - yield Remarks (gal/min) (Numbers refer to depths below land surface.) Test well. 0-25 fine sand, 25-45 gravel, 45-58 sand and gravel, 58-67 ft 75 silty sand and clay. Test well. Well abandoned. 75 Test well. Test well. Test well. - 0-15 clay and sand, 15-50 pebbly sand, 50-55 sand and gravel, 55-60 gravelly sand, 60-65 clayey sand, 65-75 silty gravel, 75-85 sand and gravel, 85-100 pebbly fine to coarse sand, 100-104 ft gray clay. Test well. Test well. 0-11 sand and gravel, 11-16 clay with some gravel, 16-48 sand and gravel, 48-64 clay with some embedded gravel (till?), 64-65 silt and fine sand, 65-72 sand and gravel, 72-85 clay with some gravel, 85-87 sand and gravel, 87-89 clay, 89-94 sand and gravel, 94-97 silty clay with some gravel , 97- 1 1 8 clay , 1 1 8- 1 24 clay with some shale pebbles (till) , 1 24- 1 60 ft shale. 0-45 ft sand and gravel. Well destroyed. 0-13 sand and gravel, 13-28 silt with some gravel, 28-85 ft silty sand and gravel. Well destroyed. 0-100 dirty gravel, 100-105 ft clay. No water encoun- tered during drilling. 25 0-83 sand and gravel, 83-130 clay and gravel, 130-178 sand and gravel, 178-183 fine sand, 183-187 fine sand and gravel, 187-207 sand and gravel, 207-209 ft shale. - - 0-58 sand and gravel, 58-70 ft fine sand and clay. 0-25 ft sand and gravel. USGS water-level observation well. 0-15 sand and gravel, 15-20 clay and gravel (till?), 20-35 sand and gravel, 35-50 sand, 50-65 sand and gravel, 65-75 sandy clay and gravel, 75-80 sand and gravel, 80-82 fine sand, 82 ft clay. 106 CP6 423522 0761315 Cortland County Airport 1970 1,198 75 1,198.43 34.50 02-26-76 ---PAGE BREAK--- CD ZJC CO CO CO T3 4 CO r z. -o c. 03 e o0 CD -i ^ COen o "co J) TJC _co "CD ^o CO i_oo CDrr T- _X C Q. Q. Remarks (Numbers refer to depths below land surface.) 0 CD t; ~o 'E O CD -S 1* Is 1 Q CD 5 CD m ~ T3 ^ CO I^ 5 C O "5 S 0 O ^ ^ R c- i. 2 ! CD CD ^ Q -Q _ *CD "CD 5 .g ^ § , 8 _ § ^ "t CD J5 D & lil C- tt T3 <0 = "DC CD 'OJ « = 3 ~ ° CD - 15 § - ' -2 "co i i- 0 C CD -IS"" (8 = to > ? 3 Test well. 1 vo Si cso (N 00 q rt ! 0 fN 5:^t ON ' o c o 4 VH~ ^ CQ 00 ^o o 00Nn 9 t^ X,U t^ 3 o § -a S e 0-25 clay hardpan, 25-37 fine sandy clayey gravel (no water), 37-40 1 and gravel, 40-67 f sand and gravel, 67-92 silty clay, 92-95 ft sand a gravel. ! vo fOo g> g> g> ^ g> ^ g> OOOO O O O O KKKK ffi ffi ffi ffi § ^ ^ ^ I c c c a c a c a & & & & & & & & 'a 'a 'a '3 '8 '3 '8 '3 £ £ £ £ £ £ £ £ o ' * in in o in ininin^ ^ in in oooo o o o o OOOOOOO\ O O C^ O _H_H,^rt (NI fN (NI (N c^c^c^cn (N (N (N fN HHa!iy5 ^ 0-30 clay and gravel, 30-32 gravel, 32-52 clay and gravel, 52-62 san and gravel, 62-66 sand, 66-71 clay and gravel, 71-84 sand and grave] 109 clay and gravel, 109-113 sand and gravel, 113-133 clay and gra 133-134 gravel, 134-140 ft clay and gravel. Screen silted up and well abandoned. o o (N vo vo9o\0 60e 1 1 ^ O ^ r4 voo\ ' ^ -g S ffi ^ 13 £ 00in * vo o M § S 1 rt m m c^ * ^ i (N 1 9 oo 0 i 00 1 rtm 1 ' G\ O r^ *o vri O VO 2" 2" 'O IO ^o ^o o\ o\ 'o ^ ^ s K e 1 1 1 in c*~i m vo vo t r- o o ON (N m § § : i fN m m m 68 Hydrogeology, Water Quality, and Simulation of Ground-Water Flow in a Glacial-Aquifer System, Cortland County, New York ---PAGE BREAK--- Appendix 1 . Records of wells and test borings in the Cortland, N.Y., study area (continued) > o o (O USGS well no. 163 167 168 170 171 172 173 174 176 177 178 179 180 181 183 184 186 187 188 203 204 205 207 210 214 279 280 281 Local identi- fiier - - PW-4 ~ Welll - - CP3 CT4D CT5S CT7D CT10S CT14 ELM A ELMB ETLA Lati- tude 423428 423516 423520 423528 423533 423542 423548 423549 423602 423604 423606 423610 423610 423619 423631 423638 423657 423709 423710 423451 423522 423447 423353 423422 423632 423609 423609 423542 Longi- tude 0761452 0761152 0761115 0761130 0760824 0761154 0761015 0761143 0761019 0761039 0761233 0761209 0761209 0760946 0761115 0761006 0761107 0761119 0761127 0761244 0761239 0761306 0761303 0761407 0761051 0760936 0760942 0760922 Owner Roger Abdallah Creamery Lehigh Railroad Mobil Oil Polkville Agway City of Cortland Brewer Tichener City of Cortland Rubbermaid Cortland County Ames, Bob Murray Center Murray Center Brewer Tichener Cortland Hospital Cortland Ready Mix Gates, Al Briggs, Lynn Shultz,Al Town of Cortlandville Cortland County Cortland County Cortland County Cortland County Cortland County Cortland County Cortland County ETL Date drilled (mo d yr) 1976 1945 1932 1974 - 03-01-57 1944 1917 1923 07-09-79 07-24-77 1943 1943 1962 05-01-60 08-01-63 05-01-66 02-22-66 1966 02-27-76 10-07-75 10-09-75 10-24-75 11-28-75 01-27-76 11-09-79 11-15-79 05-05-80 Elevation land- surface (feet) 1,290 1,155 1,136 1,147 1,096 1,138 1,115 1,138 1,110 1,117 1,180 1,150 1,150 1,105 1,121 1,100 1,115 1,120 1,125 1,183 1,149 1,170 1232 1,202 1,112 1,100 1,102 1,100 Well [ depth 1 (feet) 155 69 28 24 34 68 155 16 185 101 158 65 125 47 44 49 45 195 156 70 46.5 26 46 42 24 29 34 45 Elevation Depth to of top of oedrock casing - (feet) (feet) 30 - - - - - 1,111.88 75 - 75 - 1,120.84 - 188 148 1,183.44 1,152.15 1,173.14 1,233.09 116 1,201.75 1,114.5 44 1,102.63 1,103.87 1,102.16 Water level below land surface Feet 50 3 16.70 8 5 8 9 3 15.50 34 20.0 - 15.6 11 16 14 - 10 19.08 5.21 10.56 26.60 40.5 9.23 10.4 9.4 12.3 Date 1976 07-25-75 - 03-01-57 - 08-01-65 08-16-79 07-24-77 04-23-87 - 02-18-72 05-01-60 08-01-63 05-01-66 - - 02-27-76 10-10-75 12-01-75 10-08-91 10-08-91 02-04-76 10-08-91 10-08-91 05-28-91 Reported yield Remarks (gal/min) (Numbers refer to depths below land surface.) 4 40 75 35 4^00 75 4,000 22 70 10 150 60 310 250 100 35 30 40 - - - - _ 0-15 gravel, 15-30 till, 30-155 ft bedrock. - - - 0-37 ft sand and gravel. Public water-supply well. 0-68 sand and gravel, 68-77 ft sand and clay. - Dug well. Former public water-supply well. Currently used for water- level observation well by city of Cortland. Formerly owned by Brockway Motors. 0-35 sand and gravel, 35-130 clay, 130-185 ft sand and gravel. 0-55 sand and gravel, 55-85 clay, 85-102 ft sand and gravel. 0-5 gravel, 5-15 till, 15-158 ft bedrock. 0-75 sand and gravel, 75-125 ft shale. - - 0-20 sand and gravel, 20-75 sand and clay, 75-177 clay, 177-188 sand and gravel, 188-195 ft shale. Test well. 0-60 sand and gravel, 60-70 pebbly sand, 70-75 ft clay. 0-15 sand and gravel, 15-20 clay and gravel (till?), 20-35 ft sand and gravel. 0-5 gravel, 5-15 clayey gravel (till?), 15-32 sand and gravel, 32-37 fine sand, 37-41 medium sand, 41-50 fine sand and gravel, 50-54 till. 0-20 clayey sand and gravel, 20-25 sand and pebbles, 25-30 fine to medium sand, 30-65 clayey silty sand and gravel (till?), 65-70 fine to medium sand, 70-100 sand and gravel, 100-106 till, 106 ft shale. Well removed. 0-7 fill, 7-14 silt, 14-23 sand and gravel, 23-24 ft fine sand. 0-5 fill, 5-30 sand and gravel, 30-37 medium to coarse sand, 37-44 till, 44-45 ft shale. 0-45 sand and gravel, 45-50 ft fine to medium sand with some gravel 0-5 sand and gravel, 5-47 gravelly medium to coarse sand, 47-60 ft silty clay. ---PAGE BREAK--- Appendix 1 . Records of wells and test borings in the Cortland, N.Y., study area (continued) Hydrogeo o < f CD O 8L 0)3 Q. 3C D>5' 3ao 5c3Q. 0) 0 n 15' 0) 0 0)o 2! £1 w aar ooa. 0)3Q. OoC Z a. V) 3c SI5' I0 oc3Q. ^ I n 15' 0) O D)0 si! .a c_ w a CO 3 Oo3. D) Q. O 0C1 Z I O ^ A. April 1990 (continued) Well no. 343 346 347 348 349 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 368 375 376 403 406 420 449 450 Well depth (feet) _ 24 73 65 55 55 63 57 48 35 23 _ 25 67 63 35 14 _ - 54 26 _ pH, Date field (m/d) Alkalinity (units) 04/05 04/06 04/05 04/03 04/04 04/04 04/03 04/06 04/06 04/03 04/05 04/05 04/05 04/03 04/03 04/05 04/05 04/03 04/04 04/03 04/05 04/03 04/05 04/03 04/06 04/03 04/06 04/04 04/03 195 253 7.60 157 176 210 146 154 186 160 163 196 7.63 196 7.57 203 122 90 194 7.61 203 198 152 248 227 265 187 137 192 7.71 128 7.66 238 7.55 152 139 Specific conduct- ance, Phos- field phorus, (fiS/cm) total 0.65 814 1.26 0.71 0.15 0.64 0.28 0.33 0.33 0.56 0.01 589 0.48 533 0.12 0.40 0.05 0.08 608 0.92 1.09 0.85 0.44 0.12 0.68 0.09 0.47 0.50 638 0.29 498 0.05 616 0.18 0.56 0.40 Total recoverable (ng/L) Nitrate, total 5.6 2.2 0. 8.1 7.4 1.5 2.5 4.0 3.4 2.8 4.5 5.1 4.5 3.2 0.26 4.1 4.3 5.3 0.24 0.44 5.1 3.2 7.2 0. 4.9 2.3 5.5 3.7 1.8 Chlor- ide, total 37 61 34 19 24 61 107 38 48 45 44 26 31 66 5 49 36 56 42 90 101 232 23 19 68 19 33 16 61 Sodium, total 16 46 22 28 15 28 37 23 26 27 29 13 17 39 7 32 22 24 20 46 46 144 11 51 35 10 16 7 48 Potas- sium, total l.l 3.1 1.5 8.3 2.9 1.9 1.7 1.2 1.4 1.4 1.3 0.90 1.1 1.9 1.3 1.3 1.3 1.6 1.5 2.7 1.6 4.7 0.90 0.15 2.6 1.0 1.3 0.70 2.1 Magne- sium, total 14 41 13 124 138 13 12 14 11 10 15 15 14 7.8 3.2 13 15 15 14 26 15 25 12 1.8 11 20 16 13 14 Calcium, total 76 241 52 780 572 63 70 72 68 69 72 79 79 53 47 78 81 75 74 122 88 141 79 0.1 83 70 91 61 64 Lead <500 1,400 <500 <500 <500 <500 <500 <500 <500 <500 <500 <500 <500 <500 1,200 <500 <500 <500 <500 <500 <500 <500 <500 <500 <500 <500 <500 <500 <500 Iron <200 46,500 1,100 2,200 21,400 2,100 400 8,200 <200 <200 <200 <200 <200 25,700 60,500 <200 2200 <200 3600 29,200 <200 19,400 <200 <200 <200 6,800 5,100 <200 19,500 Copper <50 150 <50 <50 <00 70 <50 <50 <50 <50 <50 <50 <50 70 70 <50 80 <50 <50 130 <50 80 <50 <50 <50 100 <50 <50 <50 Man- ganese <30 3,400 <30 5,700 4,,100 90 <30 <30 <30 <30 <30 <30 <30 400 1,000 <30 <30 <30 <30 400 <30 500 <30 <30 <30 1,000 <30 <30 1,000 Zinc 0 200 200 100 200 400 100 500 20 0 20 0 50 4,400 3,200 10 400 40 10 300 20 200 20 0 300 7,800 20 30 100 ---PAGE BREAK--- Appendix 2.--lnorganic chemical analyses of ground water from selected wells in the glacial-aquifer system in the Cortland, New York study area, September 1990. [ , indicates no data available; less than; all values are in milligrams per liter unless otherwise noted; ^iS/cm, microsiemens per centimeter; ^ig/L, micrograms per liter; well locations shown on pi. B. SEPTEMBER 1990 (Analyses by the U.S. Geological Survey National Water Quality Laboratory) c*3Q. S' IO Well number 5 11 22 48 105 106 108 121 172 177 204 207 210 280 303 304 317 320 321 330 Well depth (feet) 40 81 45 25 59 75 38 32 69 92 46 53 45 34 28 23 68 90 80 15 Date m/d 09/19 09/19 09/20 09/17 09/17 09/19 09/17 09/20 09/18 09/19 09/17 09/17 09/17 09/19 09/17 09/17 09/18 09/20 09/17 09/17 Specific conduct- ance (nS/cm) 631 568 554 546 559 1,390 - 413 807 - 558 363 3,850 651 pH - 7.8 7.5 7.8 7.7 7.6 7.8 - 7.7 7.7 7.6 8.0 7.2 7.4 Alkalinity 224 223 189 220 175 166 300 181 206 200 197 151 204 214 226 205 140 271 218 Nitrite plus nitrate as N 9.9 5.0 7.6 6.4 4.9 6.9 6.0 5.9 5.3 4.1 4.6 8.3 4.7 3.7 4.0 5.7 1.7 3.0 6.5 Phos- phorus, ortho as P <0.0l <0.0l <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 - <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 Dissolved, Cyanide Calcium 79 <0.01 87 76 75 79 74 61 87 78 53 230 <0.01 89 in milligrams per liter Magne- sium - 13 14 14 14 15 14 - 14 17 15 13 39 12 Sodium Chloride Sulfate Fluoride 40 19 0.1 79 24 0.1 34 16 18 21 17 0.1 38 25 0.1 20 14 170 30 24 0.1 29 23 0.1 4.8 56 72 25 0.1 34 23 0.1 14 4.8 530 31 Silica - 6.3 7.4 7.3 - 7.2 7.4 7.5 7.5 7.8 - 7.6 8.7 10 6.8 ---PAGE BREAK--- Appendix 2. Inorganic chemical analyses of ground water from selected wells in the glacial-aquifer system in Cortland, N. Y. study area (continued) < Q. 1 (D O 0 (Q < 15oc 2L £ D)3Q. 3c_ 5' 3 0 0 3c3Q. 1 5 n 15' D) O D)0. B. SEPTEMBER 1990 continued. Dissolved, in micrograms per liter Well number Arsenic 5 11 22 48 <1 105 106 108 121 172 177 204 207 210 280 303 304 317 320 321 330 <1 Barium 58 29 49 66 61 57 - 73 87 _ _ 63 71 290 35 Beryl- Cadm- ium ium <0.5 <1 <0.5 <1 <0.5 <1 - <0.5 <1 0.8 <1 <0.5 <1 <0.5 <1 <0.5 <1 _ _ <0.5 <1 <0.5 <1 2.0 3 <0.5 <1 Chrom- ium <5 <5 <5 - <5 <5 <5 <5 <5 _ <5 <5 15 <5 Cobalt Copper <3 <10 <3 <10 <3 <10 <3 <10 <3 <10 <3 20 <3 <10 <3 <10 _ _ <3 50 <3 <10 9 30 <3 <10 Iron - 33 6 3 13 23 12 - 12 37 _ _ 14 4 11 10 Mang- Molyb- lead anese denum Nickel Silver <10 17 <10 <10 <1 <10 1 <10 <10 <1 <10 2 <10 <10 <1 <10 12 <10 <10 <1 <10 15 <10 <10 <1 <10 6 <10 <10 <1 <10 42 <10 <10 <1 <10 2 <10 <10 <1 _ <10 11 <10 <10 <1 <10 4 <10 <10 <1 30 4 30 30 4 <10 2 <10 <10 <1 Stron- tium - 110 110 110 120 110 140 92 140 _ _ 100 91 400 120 Vanad- ium <6 <6 <6 - <6 <6 <6 - <6 <6 _ _ <6 <6 18 <6 Zinc 4 24 62 - - 15 36 35 430 7 _ _ 4 5 9 28 Lithium 7 7 8 - 13 9 14 7 10 _ _ 9 10 23 6 > .a I Oo3 D) a. Oo ---PAGE BREAK--- Appendix 2. Inorganic chemical analyses of ground water from selected wells in the glacial-aquifer system in Cortland, N.Y. study area (continued). B. SEPTEMBER 1990 continued. TO a (D a.S1 10 Well number 332 333 335 337 338 339 340 341 342 343 346 347 348 349 351 352 355 356 357 358 362 363 364 365 366 367 368 Well depth (feet) 34 55 34 42 217 35 35 137 52 24 24 ~ 73 65 55 55 63 58 48 35 67 63 35 14 Date mo/d 09/18 09/18 09/18 09/19 09/19 09/18 09/19 09/19 09/17 09/20 09/19 09/20 09/18 09/18 09/20 09/20 09/19 09/18 09/18 09/18 09/20 09/20 09/20 09/17 09/20 09/20 09/17 Specific conduct- ance (ys/cm) 815 565 854 533 288 584 763 716 610 662 823 492 571 594 613 505 614 [PHONE REDACTED] PH (units) 7.7 7.7 7.7 7.6 8.9 7.7 7.8 8.0 7.8 7.8 7.5 7.9 7.7 7.8 7.9 7.9 7.6 7.7 6.9 7.2 Alkalinity 230 211 255 201 104 200 224 111 176 252 225 193 174 211 - 158 199 197 231 194 199 312 209 200 Nitrite plus nitrate as N 0.7 4.2 5.1 6.5 0.1 5.4 3.6 0.1 3.7 0.7 1.6 2.8 3.3 4.0 5.3 4.4 4.6 4.7 0.3 3.5 5.5 - Phor- phous, ortho as P <0.0l <0.0l <0.0l <0.0l <0.01 <0.01 <0.01 <0.01 0.02 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 - Dissolved, Cyanide Calcium 100 77 91 <0.01 81 13 83 98 32 72 <0.01 89 92 77 56 61 79 83 110 100 in milligrams per liter Magne- sium. 18 16 18 14 6 14 17 14 14 19 16 13 13 10 - 15 16 17 ~ 14 Sodium Chloride Sulfate Fluoride 35 16 59 11 42 20 34 86 33 22 52 23 44 24 21 57 23 0.1 23 45 70 22 0.1 110 Silica 8.3 7.8 8.5 7.4 0.55 6.5 8.7 1.4 7.6 7.8 8.5 ~ 6.4 3.9 5.7 7.3 7.2 14 6.9 ---PAGE BREAK--- Appendix 2. Inorganic chemical analyses of ground water from selected wells in the glacial-aquifer system in Cortland, N.Y. study area (continued). B. SEPTEMBER 1990, continued. 5 (Q (D O O f O SL ~r a>3Q. V) 3c D)5' 3ao5c3Q. k n o 5' a> O 0)o. > c_ 0) ai o 0a. D)3 Q. O 0Ca v« Dissolved, in micrograms per liter Well number Arsenic 332 333 335 337 <1 338 339 340 341 342 343 <1 346 347 348 349 351 352 355 356 357 358 362 363 364 365 366 367 368 Barium 75 50 69 53 120 38 72 210 42 62 87 _ _ 51 47 43 _ 59 61 150 71 Beryl- ium <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 _ <0.5 0.6 <0.5 _ <0.5 <0.5 <0.5 ~ - 0.5 Cadm- Chrom- ium ium <1 <5 <1 <5 <1 <5 <1 <5 <1 <5 <1 <5 <1 <5 <1 <5 <1 <5 <1 <5 <1 <5 _ _ <1 <5 <1 <5 <1 <5 _ <1 <5 <1 <5 <1 <5 2 <5 Cobalt <3 <3 <3 <3 <3 <3 <3 <3 <3 <3 <3 _ _ _ <3 <3 <3 _ _ <3 <3 <5 ~ <3 Copper Iron <10 10 <10 32 <10 95 <10 19 <10 41 <10 15 <10 9 <10 62 <10 280 <10 5 <10 17 _ _ <10 6 <10 18 <10 7 _ <10 580 <10 730 40 7900 <10 17 Mang- Molyb- lead anese denum <10 790 <10 <10 250 <10 <10 270 <10 <10 26 <10 <10 33 <10 <10 12 <10 <10 64 <10 <10 65 <10 <10 200 <10 <10 580 <10 <10 94 <10 _ _ <10 3 <10 <10 27 <10 <10 1 <10 _ <10 14 <10 <10 20 <10 30 3600 <10 ~ <10 4 <10 Stron- Nickel Silver tium <10 <1 200 <10 <1 110 <10 <1 170 <10 <1 110 <10 <1 400 <10 <1 120 <10 <1 150 <10 <1 390 <10 <1 110 <10 <1 150 <10 <1 130 _ _ <10 <1 110 <10 <1 100 <10 <1 85 _ _ <10 <1 110 <10 <1 120 <10 <1 190 <10 <1 150 Vanad- ium <6 <6 <6 <6 <6 <6 <6 <6 <6 <6 <6 _ _ <6 <6 <6 _ <6 <6 <6 - <6 Zinc 9 3 13 3 5 6 7 9 15 16 8 _ _ 72 91 11 _ _ 120 4 210 8 Lithium 16 9 11 7 29 5 10 240 9 8 10 _ _ 8 8 7 - _ _ 11 9 12 9 ---PAGE BREAK--- Appendix 3.--Organic chemical analyses of ground water from selected wells in the glacial-aquifer system in the Cortland, N.Y. study area. [Analyses by the U.S. Geological Survey National Water Quality Laboratory; no data available; less than; Well locations shown on pi 1] Appendix 3 00 Well number l 11 22 105 121 172 177 204 205 280 304 317 320 330 332 Well depth (feet) 81.0 45.0 59.0 32.0 77.0 92.5 46.5 26.0 34.0 23.0 68.0 90.0 15.0 34.0 Sample Sample Tetra- Tri- date depth chloro- chloro- yr/mo/d (feet) ethylene ethylene 90/04/05 90/04/05 90/09/19 90/04/03 90/09/20 93/04/27 90/04/05 90/09/17 93/04/27 90/04/04 90/09/20 93/04/27 90/04/05 90/10/02 90/10/15 90/04/06 90/04/04 90/09/17 93/04/27 90/04/05 90/09/17 93/04/27 90/04/06 90/09/19 90/04/04 90/09/17 93/04/27 93/04/27 90/09/20 90/04/05 90/04/05 90/09/18 <0.20 0.40 0.04 <0.20 11.0 14.6 6.2 82.8 53.9 54.0 101 66.8 75.0 0.02 - 11 .4 0.25 2.30 1.90 3.4 7.00 24.2 4.80 2.70 2.60 2.3 0.40 0.10 15.2 Concentration (total), in micrograms per liter 1,1,1- Tri- 1,1 -Di- Dichlor- chloro- chloro- bromo- Methyl- ethane ethylene Benzene Toluene Xylene methane chloride 0.25 2.0 <0.20 <0.2 <0.2 <0.2 <0.2 <0.2 1.4 0.20 <0.20 <0.2 <0.2 <0.2 <0.2 <0.2 0.05 0.01 0.20 0.60 1.2 0.10 0.60 - 0.2 0.40 0.30 0.60 0.06 1.6 0.20 0.4 1.5 0.20 1.0 0.40 0.30 0.10 2.4 0.8 0.10 2.4 - 0.8 107 frans-1 c/s-1 dichloro- dichloro- ethene ethene <0.2 5.9 3.50 23.8 16.8 25.0 34.5 21.0 32.0 0.20 0.60 0.3 1.10 0.80 2.10 1.80 10.0 1.00 0.10 0.40 0.2 0.50 0.50 0.95 1 .20 1.0 1.0 28.9 ---PAGE BREAK--- Appendix 3.-Organic chemical analyses of ground water from selected wells in the glacial-aquifer system in the Cortland, N.Y. study area, continued Hydrogeology, ^ so Q. cn I fto3 0 3c Q. 1 £ Oo (D O ---PAGE BREAK--- ---PAGE BREAK--- ---PAGE BREAK--- ---PAGE BREAK---