Search USGSSearch

SEARCH · Search USGS

Results for “Scientific Investigations Report”

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

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

At least 1,207 records · Page 67Linked to original sources

Groundwater discharge by evapotranspiration, flow of water in unsaturated soil, and stable isotope water sourcing in areas of sparse vegetation, Amargosa Desert, Nye County, Nevada

This report documents methodology and results of a study to evaluate groundwater discharge by evapotranspiration (GWET) in sparsely vegetated areas of Amargosa Desert and improve understanding of hydrologic-continuum processes controlling groundwater discharge. Evapotranspiration and GWET rates were computed and characterized at three sites over 2 years using a combination of micrometeorological, unsaturated zone, and stable-isotope measurements. One site (Amargosa Flat Shallow [AFS]) was in a sparse and isolated area of saltgrass ( Distichlis spicata ) where the depth to groundwater was 3.8 meters (m). The second site (Amargosa Flat Deep [AFD]) was in a sparse cover of predominantly shadscale ( Atriplex confertifolia ) where the depth to groundwater was 5.3 m. The third site (Amargosa Desert Research Site [ADRS]), selected as a control site where GWET is assumed to be zero, was located in sparse vegetation dominated by creosote bush ( Larrea tridentata ) where the depth to groundwater was 110 m. Results indicated that capillary rise brought groundwater to within 0.9 m (at AFS) and 3 m (at AFD) of land surface, and that GWET rates were largely controlled by the slow but relatively persistent upward flow of water through the unsaturated zone in response to atmospheric-evaporative demands. Greater GWET at AFS (50 ± 20 millimeters per year [mm/yr]) than at AFD (16 ± 15 mm/yr) corresponded with its shallower depth to the capillary fringe and constantly higher soil-water content. The stable-isotope dataset for hydrogen (δ 2 H) and oxygen (δ 18 O) illustrated a broad range of plant-water-uptake scenarios. The AFS saltgrass and AFD shadscale responded to changing environmental conditions and their opportunistic water use included the time- and depth-variable uptake of unsaturated-zone water derived from a combination of groundwater and precipitation. These results can be used to estimate GWET in other areas of Amargosa Desert where hydrologic conditions are similar.

Nevada

Temporal changes in nutrient concentrations in the Lower Grand River and selected drainage basins, Missouri and Iowa, during the Mississippi River Basin Healthy Watersheds Initiative (2010–23)

This report describes a cooperative study by the U.S. Geological Survey and Missouri Department of Natural Resources that evaluated temporal changes in total nitrogen (TN) and total phosphorus (TP) concentrations in the Lower Grand River hydrologic unit. The study focused on trends since 2010, when the basin was designated as a priority drainage basin of the Mississippi River Basin Healthy Watersheds Initiative (MRBI). At three local drainage basins within the Lower Grand hydrological unit (MRBI sites), stream nutrient trends were evaluated using flow-adjusted (FA) TN and TP concentrations for water years 2011 through 2023. FATN concentration trends were not statistically significant for any MRBI site. One site (site 2) showed a statistically significant increasing trend in FATP concentration, indicating a possible increase in phosphorus sources in parts of the basin. Overall, streamflow variability appeared to be the dominant factor affecting nutrient concentrations at MRBI sites. At five regional drainage basins, including the Grand River and nearby rivers with data from 1994 through 2023 (long-term sites), annual flow-normalized (FN) TN and TP concentrations were evaluated for trends before (water years 2000–10) and during (water years 2010–23) the MRBI. For water years 2010 through 2023, annual FNTN and FNTP concentrations decreased in the Grand River, as well as in the Nodaway and Chariton Rivers, which were not targeted by the MRBI. The Grand River (site 9) reversed from increasing to decreasing FNTP concentrations after 2010, with a 26-percent reduction. Annual FNTN and FNTP concentrations also decreased at the Missouri River sites. While nutrient reductions in the Grand River may reflect the effects of implemented conservation practices, similar trends in nearby, nontargeted rivers and the absence of strong decreasing trends at MRBI sites suggest that broader regional factors, instead of or in addition to MRBI efforts, may have contributed to nutrient reductions in the Grand River.

Iowa, Missouri

2009 Spring floods in North Dakota, western Minnesota, and northeastern South Dakota

In 2009, record-breaking snowfalls and additional spring moisture caused severe flooding in parts of the Missouri River and Red River of the North (Red River) Basins in North Dakota, Minnesota, and South Dakota. There were 48 peak of record stages and 36 discharges recorded at U.S. Geological Survey streamgages located in both basins between March 20 and May 15, 2009. High water continued to affect many communities up and down the rivers' main stems and tributaries for nearly 2 months. Record snowfall for single-day totals, as well as monthly totals, occurred throughout the Missouri River and Red River of the North Basins. Additional moisture in the spring as well as the timing of warmer temperatures caused record flooding in many places in both basins with many locations reporting two flood crests. Ice jams on the Missouri River, located north and south of Bismarck, N. Dak., caused flooding. Southwest Bismarck was evacuated as rising waters first began inundating homes in low-lying areas along the river and then continued flowing into the city's lower south side. On March 24, 2009, the peak stage of the Missouri River at Bismarck, N. Dak. streamgage was 16.11 feet, which was the highest recorded stage since the completion of Garrison Dam in 1954. South of Bismarck, the Missouri River near Schmidt, N. Dak. streamgage recorded a peak stage of 24.24 feet on March 25, 2009, which surpassed the peak of record of 23.56 feet that occurred on December 9, 1976. While peak stage reached record levels at these streamgages, the discharge through the river at these locations did not reach record levels. The record high stages resulted from ice jams occurring on the Missouri River north and south of the cities of Bismarck and Mandan. At the Red River of the North at Fargo, N. Dak. streamgage, the Red River reached a record stage of 40.84 feet surpassing the previous peak of record stage of 39.72 feet set in 1997. The associated peak streamflow of 29,500 cubic feet per second exceeded the previous peak of record set in 1997 by 1,500 cubic feet per second. For the cities of Fargo, and Moorhead, Minn., and the surrounding area, the stage of the Red River remained above flood stage for nearly 2 months. In addition to high stage and flow on the main-stem Missouri and Red Rivers, peak of record stage and discharge were recorded at many U.S. Geological Survey streamgages in the Missouri River and Red River Basins. Several reservoirs and lakes in the region also experienced record stage elevations from the high flows during the 2009 spring snowmelt floods.

Scientific Investigations Report

Ambient and potential denitrification rates in marsh soils of Northeast Creek and Bass Harbor Marsh watersheds, Mount Desert Island, Maine

Nutrient enrichment from atmospheric deposition, agricultural activities, wildlife, and domestic sources is a concern at Acadia National Park on Mount Desert Island, Maine, because of the potential problems of degradation of water quality and eutrophication in estuaries. Degradation of water quality has been observed at Bass Harbor Marsh estuary in the park but only minimally in Northeast Creek estuary. Previous studies at Acadia National Park have estimated nutrient inputs to estuaries from atmospheric deposition and surface-water runoff, and have identified shallow groundwater as an additional potential source of nutrients. Previous studies at Acadia National Park have assumed that a certain fraction of the nitrogen input was removed through microbial denitrification, but rates of denitrification (natural or maximum potential) in marsh soils have not been determined. The U.S. Geological Survey, in cooperation with Acadia National Park, measured in-place denitrification rates in marsh soils in Northeast Creek and in Bass Harbor Marsh watersheds during summer 2008 and summer 2009. Denitrification was measured under ambient conditions as well as after additions of inorganic nitrogen and glucose. In-place denitrification rates under ambient conditions were similar to those reported for other coastal wetlands, although they were generally lower than those reported for salt marshes having high ambient concentrations of nitrate (NO 3 ). Denitrification rates generally increased by at least an order of magnitude following NO 3 additions, with or without glucose (as the carbohydrate) additions, compared with the ambient treatments that received no nutrient additions. The treatment that added both glucose and NO 3 resulted in a variety of denitrification responses when compared with the addition of NO 3 alone. In most cases, the addition of glucose to a given rate of NO3 addition resulted in higher rates of denitrification. These variable responses indicate that the amount of labile carbohydrates can limit denitrification even if NO 3 is present. For most sites in both watersheds, the maximum denitrification rates ranged from of 150 to 900 micromoles of nitrous oxide per square meter per hour. These rates were equivalent to the release of 37 to 221 grams of nitrogen per square meter per year. Weak positive correlations were observed for soil temperature and for measured ammonium concentration in groundwater. Weak negative correlations were observed between denitrification rate and water level and specific conductance. The rates of denitrification in Bass Harbor Marsh and Northeast Creek under ambient conditions, both of which were relatively low, indicate that NO 3 availability is low in both systems. It is evident from the addition of combined treatments of NO 3 and glucose that these marsh soils are capable of comparatively high rates of denitrification, therefore, estuarine eutrophication is not a result of nitrogen inputs to marsh soils that are in excess of the denitrification capacity in these systems. If terrestrial inputs to the estuary are the cause of the observed eutrophic condition in Bass Harbor Marsh, then these inputs to the estuary must bypass the marsh in channelized surface flow, or perhaps they circumvent the marsh in shallow groundwater seepage along subsurface pathways that enter the estuary directly.

Maine

Methods for and estimates of 2003 and projected water use in the Seacoast Region, Southeastern New Hampshire

New methods were developed to estimate water use in 2003 and future water demand in 2017 and 2025 in the Seacoast region in southeastern New Hampshire, which has experienced a 37-percent population increase during 1980 to 2000. Water-use activities for which estimates were developed include water withdrawal, delivery, demand, consumptive use, release, return flow, and transfer by registered and aggregated unregistered (less than 20,000 gallons per day (gal/d)) users at the census-block and town scales. Estimates of water use rely on understanding what influences water demand and its associated consumptive use, because changes in demand and consumptive use affect withdrawal and return flow. Domestic water demand was estimated using a per capita water demand model that related metered deliveries to domestic users with census block and block-group data. The model was used to predict annual, summer, and winter per capita water-demand coefficients for each census block. Significant predictors of domestic water demand include population per housing unit, median value of owner-occupied single family homes, median year of housing construction (with 1900 as the base value), population density, housing unit density, and proportion of housing units that are in urban areas. Mean annual domestic per capita water-demand coefficient in the Seacoast region was 75 gal/d; the coefficient increased to 91 gal/d during the summer and decreased to 65 gal/d during the winter. Domestic consumptive use was estimated as the difference between annual and winter domestic water demand. Estimates of commercial and industrial water demand were based on coefficients derived from reported use and metered deliveries. Projections of water demand in 2017 and 2025 were determined by using the housing and employee projections for those years developed through a Transportation Demand Model and applying current domestic and non-domestic coefficients. Water demand in 2003 was estimated as 25.8 million gallons per day (Mgal/d), 35 percent of which was during the summer months of June, July, and August. Domestic water demand was 18.6 Mgal/d (72 percent), commercial water demand was 3.7 Mgal/d (14 percent), industrial water demand was 2.9 Mgal/d (11 percent), irrigation water demand was 0.3 Mgal/d (1 percent), and mining and aquaculture water demand was 0.2 Mgal/d (1 percent). Domestic consumptive use for the Seacoast region was 16 percent of domestic water demand, which translates to a loss of 3 Mgal/d over the entire Seacoast region. In 2003, water withdrawal was 771.2 Mgal/d, of which 742.2 Mgal/d was instream use for hydroelectric power generation and thermoelectric power cooling. The remaining 29.0 Mgal/d was withdrawn by community water systems (22.6 Mgal/d; 72 percent), domestic users (6.4 Mgal/d; 21 percent), commercial users (1.0 Mgal/d; 3 percent), industrial users (1.0 Mgal/d; 3 percent), irrigation (0.2 Mgal/d; 1 percent) and other users (less than 0.1 Mgal/d). Return flow for 2003 was 772.2 Mgal/d, of which 742.0 Mgal/d was returned following use for hydroelectric power generation and thermoelectric plant cooling. The remaining 30.2 Mgal/d was returned by community wastewater systems (20.2 Mgal/d; 68 percent), domestic users (7.8 Mgal/d; 26 percent), commercial users (1.2 Mgal/d; 3 percent), industrial users (0.8 Mgal/d; 3 percent), and other users (0.1 Mgal/d). Domestic water demand is projected to increase by 54 percent to 28.7 Mgal/d from 2003 to 2025 based on projection of future population growth. Non-domestic (commercial, industrial, irrigation, and mining) water demand is projected to increase by 66 percent to 11.8 Mgal/d from 2003 to 2025.

Scientific Investigations Report

Assessment of artificial recharge at Sand Hollow Reservoir, Washington County, Utah, updated to conditions through 2006

Sand Hollow, Utah, is the site of a surface-water reservoir completed in March 2002 and operated by the Washington County Water Conservancy District (WCWCD) primarily as an aquifer storage and recovery project. The reservoir is an off-channel facility that receives water from the Virgin River, diverted near the town of Virgin, Utah. Hydrologic data collected are described and listed in this report, including ground-water levels, reservoir stage, reservoir-water temperature, meteorology, evaporation, and estimated ground-water recharge. Since the construction of the reservoir in 2002, diversions from the Virgin River have resulted in generally rising stage and surface area. Large spring run-off volumes during 2005-06 allowed the WCWCD to fill the reservoir to near capacity, with a surface area of about 1,300 acres in 2006. Reservoir stage reached a record altitude of about 3,060 feet in May 2006, resulting in a depth of nearly 90 feet and a reservoir storage of about 51,000 acre-feet. Water temperature in the reservoir shows large seasonal variation and has ranged from about 5 to 32?C. Estimated ground-water recharge rates have ranged from 0.01 to 0.43 feet per day. Estimated recharge volumes have ranged from about 200 to about 3,500 acre-feet per month. Total ground-water recharge from March 2002 through August 2006 is estimated to be about 51,000 acre-feet. Estimated evaporation rates have varied from 0.05 to 0.97 feet per month, resulting in evaporation losses of 20 to 1,200 acre-feet per month. Total evaporation from March 2002 through August 2006 is estimated to be about 17,000 acre-feet. The combination of generally declining recharge rates and increasing reservoir altitude and area explains the trend of an increasing ratio of evaporation to recharge volume over time, with the total volume of water lost through evaporation nearly as large as the volume of ground-water recharge during the first 8 months of 2006. With removal of the viscosity effects (caused by seasonal water temperature variations), the intrinsic permeability indicates a large seasonal variation in clogging, with large winter increases likely caused by a combination of both decreased biofilms and the reduced volume of trapped gas bubbles.

Utah

Questa baseline and pre-mining ground-water quality investigation. 14. Interpretation of ground-water geochemistry in catchments other than the Straight Creek catchment, Red River Valley, Taos County, New Mexico, 2002-2003

The U.S. Geological Survey, in cooperation with the New Mexico Environment Department, is investigating the pre-mining ground-water chemistry at the Molycorp molybdenum mine in the Red River Valley, New Mexico. The primary approach is to determine the processes controlling ground-water chemistry at an unmined, off-site but proximal analog. The Straight Creek catchment, chosen for this purpose, consists of the same Tertiary-age quartz-sericite-pyrite altered andesite and rhyolitic volcanics as the mine site. Straight Creek is about 5 kilometers east of the eastern boundary of the mine site. Both Straight Creek and the mine site are at approximately the same altitude, face south, and have the same climatic conditions. Thirteen wells in the proximal analog drainage catchment were sampled for ground-water chemistry. Eleven wells were installed for this study and two existing wells at the Advanced Waste-Water Treatment (AWWT) facility were included in this study. Eight wells were sampled outside the Straight Creek catchment: one each in the Hansen, Hottentot, and La Bobita debris fans, four in a well cluster in upper Capulin Canyon (three in alluvial deposits and one in bedrock), and an existing well at the U.S. Forest Service Questa Ranger Station in Red River alluvial deposits. Two surface waters from the Hansen Creek catchment and two from the Hottentot drainage catchment also were sampled for comparison to ground-water compositions. In this report, these samples are evaluated to determine if the geochemical interpretations from the Straight Creek ground-water geochemistry could be extended to other ground waters in the Red River Valley , including the mine site. Total-recoverable major cations and trace metals and dissolved major cations, selected trace metals, anions, alkalinity; and iron-redox species were determined for all surface- and ground-water samples. Rare-earth elements and low-level As, Bi, Mo, Rb, Re, Sb, Se, Te, Th, U, Tl, V, W, Y, and Zr were determined on selected samples. Dissolved organic carbon (DOC), mercury, sulfate stable isotope composition (δ 34 S and δ 18 O of sulfate), stable isotope composition of water (δ 2 H and δ 18 O of water) were measured for selected samples. Chlorofluorocarbons (CFC) and 3 He and 3 H were measured for age dating on selected samples. Linear regressions from the Straight Creek ground-water data were used to compare ground-water chemistry trends in non-Straight Creek ground waters with Straight Creek alluvial ground-water chemistry dilution trends. Most of the solute trends for the ground waters are similar to those for Straight Creek but there are some notable exceptions. In lithologies that contain substantial pyrite mineralization, acid waters form with similar chemistries to those in Straight Creek and all the waters tend to be calcium-sulfate type. Hottentot ground waters contain substantially lower calcium concentrations relative to those in Straight Creek. This anomaly results from the exposure of rhyolite porphyry in the Hottentot scar and weathering zone. The rhyolite contains less calcium than the altered andesites and tuffs in the Straight Creek catchment and probably does not have the abundant gypsum and calcite. The Hansen ground waters have reached gypsum saturation and have similar calcium, magnesium, and beryllium concentrations as Straight Creek ground waters but have lower concentrations of fluoride, manganese, zinc, cobalt, nickel, copper, and lithium. Lower concentrations of elements related to mineralization at Hansen likely reflect the more distal location of Hansen with respect to intrusive centers that provided the heat source for hydrothermal alteration. The other ground water with water chemistry trends that are outside the Straight Creek trends was from an alluvial well from Capulin Canyon (CC2A). Although it had pH values near 6.0 and most major ions similar to the other Capulin Canyon ground waters, it contained high concentrations of fluoride, manganese, aluminum, iron, beryllium, and zinc similar to a mineralized zone and had low alkalinity. Saturation indices indicate that solubility constraints continue to provide upper limits on some solute concentrations. Siderite, ferrihydrite, calcite, gypsum, rhodochrosite, and barite provide limits for concentrations of Fe(II), Fe(III), Ca, Mn, and Ba, respectively. Beryllium concentrations may be subject to an upper concentration limit by the solubility of Be(OH) 2 but these concentrations probably are not reached in the ground waters. Ground-water isotopic data were consistent with the meteoric water line estimated for precipitation in the Red River Valley, indicating that all the ground waters examined in this study were meteoric, recent in origin, and showed no substantial indication of evaporation. Tritium-helium-3 and chlorofluorocarbon (CFC) age dating were partially successful. Generally, dates were consistent with location and depth of wells. Two samples had good agreement between CFC dates and tritium-helium dates, whereas a third reflected either substantial mixing with younger or older waters or complications arising from excess helium-4. The well at La Bobita appeared to contain a large component of modern water, most likely as a result of mixing with water from Red River alluvial deposits.

New Mexico

Distribution of indoor radon concentrations in Pennsylvania, 1990-2007

Results from 548,507 indoor radon tests from a database compiled by the Pennsylvania Department of Environmental Protection, Bureau of Radiation Protection, Radon Division, are evaluated in this report in an effort to determine areas where concentrations of radon are highest. Indoor radon concentrations were aggregated according to geologic unit and hydrogeologic setting for spatial analysis. Indoor radon concentrations greater than or equal to the U.S. Environmental Protection Agency (USEPA) action level of 4 picocuries per liter (pCi/L) were observed for 39 percent of the test results; the highest concentration was 1,866.4 pCi/L. When analyzed according to Pennsylvania’s geologic units, 93 of the 188 (49.5 percent) geologic units with indoor radon concentrations had median concentrations greater than the USEPA action level of 4 pCi/L; most of these geologic units are located in the eastern part of the State and include metamorphic rocks, limestones, sandstones, shales, and glacial deposits. When analyzed according to Pennsylvania’s hydrogeologic settings, 5 of the 20 (25 percent) settings had median indoor radon concentrations greater than the USEPA action level of 4 pCi/L; these settings are located mostly in the south-central part of the State. Median indoor radon concentrations aggregated according to geologic units and hydrogeologic settings are useful for drawing general conclusions about the occurrence of indoor radon in specific geologic units and hydrogeologic settings, but the associated data and maps have limitations. The aggregated indoor radon data have testing and spatial accuracy limitations due to lack of available information regarding testing conditions and the imprecision of geocoded test locations. In addition, the associated data describing geologic units and hydrogeologic settings have spatial and interpretation accuracy limitations, which are a result of using statewide data to define conditions at test locations and geologic data that represent a broad interpretation of geologic units across the State. As a result, indoor air radon concentration distributions are not proposed for use in predicting individual concentrations at specific sites nor for use as a decision-making tool for property owners to decide whether to test for indoor radon concentrations at specific property locations.

Pennsylvania

Baseline assessment of groundwater quality in Wayne County, Pennsylvania, 2014

The Devonian-age Marcellus Shale and the Ordovician-age Utica Shale, geologic formations which have potential for natural gas development, underlie Wayne County and neighboring counties in northeastern Pennsylvania. In 2014, the U.S. Geological Survey, in cooperation with the Wayne Conservation District, conducted a study to assess baseline shallow groundwater quality in bedrock aquifers in Wayne County prior to potential extensive shale-gas development. The 2014 study expanded on previous, more limited studies that included sampling of groundwater from 2 wells in 2011 and 32 wells in 2013 in Wayne County. Eighty-nine water wells were sampled in summer 2014 to provide data on the presence of methane and other aspects of existing groundwater quality throughout the county, including concentrations of inorganic constituents commonly present at low levels in shallow, fresh groundwater but elevated in brines associated with fluids extracted from geologic formations during shale-gas development. Depths of sampled wells ranged from 85 to 1,300 feet (ft) with a median of 291 ft. All of the groundwater samples collected in 2014 were analyzed for bacteria, major ions, nutrients, selected inorganic trace constituents (including metals and other elements), radon-222, gross alpha- and gross beta-particle activity, selected man-made organic compounds (including volatile organic compounds and glycols), dissolved gases (methane, ethane, and propane), and, if sufficient methane was present, the isotopic composition of methane. Results of the 2014 study show that groundwater quality generally met most drinking-water standards, but some well-water samples had one or more constituents or properties, including arsenic, iron, pH, bacteria, and radon-222, that exceeded primary or secondary maximum contaminant levels (MCLs). Arsenic concentrations were higher than the MCL of 10 micrograms per liter (µg/L) in 4 of 89 samples (4.5 percent) with concentrations as high as 20 µg/L; arsenic concentrations were higher than the Health Advisory level of 2 µg/L in 27 of 89 samples (30 percent). Total iron concentrations exceeded the secondary maximum contaminant level (SMCL) of 300 µg/L in 9 of 89 samples (10 percent). The pH ranged from 5.4 to 9.3 and did not meet the SMCL range of greater than 6.5 to less than 8.5 in 27 samples (30 percent); 22 samples had pH values less than 6.5, and 5 samples had pH values greater than 8.5. Total coliform bacteria were detected in 22 of 89 samples (25 percent); Escherichia coli were detected in only 2 of those 22 samples. Radon-222 activities ranged from 25 to 7,400 picocuries per liter (pCi/L), with a median of 2,120 pCi/L, and exceeded the proposed drinking-water standard of 300 pCi/L in 86 of 89 samples (97 percent); radon-222 activities were higher than the alternative proposed standard of 4,000 pCi/L in 12 of 89 samples (13.5 percent). Water from 8 of the 89 wells (9 percent) had concentrations of methane greater than the reporting level of 0.24 milligrams per liter (mg/L) with the detectable methane concentrations ranging from 0.74 to 9.6 mg/L. Of 16 replicate samples submitted to another laboratory with a lower reporting level of 0.0002 mg/L, 15 samples had detectable methane concentrations that ranged from 0.0011 to 9.7 mg/L. Of these 15 samples, low levels of ethane (0.00032 to 0.0017 mg/L) were detected in 6 of 7 samples with methane concentrations greater than 0.75 mg/L. The isotopic composition of methane in 6 of 8 samples with sufficient dissolved methane (about 1 mg/L) for isotopic analysis is consistent with a predominantly thermogenic methane source (sample carbon isotopic ratio δ13C CH4 values ranging from -56.36 to -45.97 parts per thousand (‰) and hydrogen isotopic ratio δD CH4 values ranging from -233.1 to -141.1 ‰). However, the low levels of ethane relative to methane indicate that the methane may be of microbial origin and subsequently underwent oxidation. Isotopic compositions indicated a possibly mixed thermogenic and microbial source (carbon dioxide reduction process) for the methane in 1 of the 8 samples (δ 13 C CH4 of -63.72 and δD CH4 of -192.3 ‰) and potential oxidation of microbial and (or) thermogenic methane in the remaining sample (δ 13 C CH4 of -46.56 and δD CH4 of -79.7 ‰). Groundwater samples with relatively elevated methane concentrations (near or greater than 1 mg/L) had a chemical composition that differed in some respects (pH, selected major ions, and inorganic trace constituents) from groundwater with relatively low methane concentrations (less than 0.75 mg/L). The seven well-water samples with the highest methane concentrations (from about 1 to 9.6 mg/L) also had among the highest pH values (8.1 to 9.3, respectively) and the highest concentrations of sodium, lithium, boron, fluoride, arsenic, and bromide. Relatively elevated concentrations of some other constituents, such as barium, strontium, and chloride, commonly were present in, but not limited to, those well-water samples with elevated methane. Groundwater samples with the highest methane concentrations had chloride/bromide ratios that indicate mixing with a small amount of brine (0.02 percent or less, by volume) similar in composition to that reported for gas and oil well brines in Pennsylvania. Most other samples with low methane concentrations (less than about 1 mg/L) had chloride/bromide ratios that indicate predominantly man-made sources of chloride, such as road salt, septic systems, and (or) animal waste. Although naturally occurring brines may originate from deeper parts of the aquifer system, the man-made sources are likely to affect shallow groundwater. Geochemical modeling showed that the water chemistry of samples with elevated pH, sodium, lithium, bromide, and alkalinity could result from dissolution of calcite (calcium carbonate) combined with cation exchange and mixing with a small amount of brine. Through cation exchange reactions (which are equivalent to processes in a water softener) calcium ions released by calcite dissolution are exchanged for sodium ions on clay minerals. The spatial distribution of groundwater compositions generally shows that (1) relatively dilute, slightly acidic, oxygenated, calcium-carbonate type waters tend to occur in the uplands along the western border of Wayne County; (2) waters of near neutral pH with the highest amounts of hardness (calcium and magnesium) generally occur in areas of intermediate altitudes; and (3) waters with pH values greater than 8, low oxygen concentrations, and the highest arsenic, sodium, lithium, bromide, and methane concentrations can occur in deep wells in uplands but most frequently occur in stream valleys, especially at low elevations (less than about 1,200 ft above North American Vertical Datum of 1988) where groundwater may be discharging regionally, such as to the Delaware River. Thus, the baseline assessment of groundwater quality in Wayne County prior to gas-well development shows that shallow (less than about 1,000 ft deep) groundwater is generally of good quality, but methane and some constituents present in high concentrations in brine (and produced waters from gas and oil wells) may be present at low to moderate concentrations in some parts of Wayne County.

Pennsylvania

Assessment of well yield, dominant fractures, and groundwater recharge in Wake County, North Carolina

A cooperative study led by the U.S. Geological Survey and Wake County Environmental Services was initiated to characterize the fractured-rock aquifer system and assess the sustainability of groundwater resources in and around Wake County. This report contributes to the development of a comprehensive groundwater budget for the study area, thereby helping to enable resource managers to make sound and sustainable water-supply and water-use decisions. Construction information was used to analyze the well depth, casing depth, and reported well yield of more than 7,500 inventoried wells. The median well depth and casing depth were 265 feet (ft) below land surface (bls) and 68 ft bls, respectively, and the median well yield was 10 gallons per minute. Generally, well yield increased with depth to around 200 ft bls and then began to decrease with depth within the fractured-rock aquifer. Borehole geophysical logging methods were used to characterize the fractured-rock aquifer by mapping the orientation of geologic structures within the subsurface. Structure measurements were made on resulting log data and mapped to observed general spatial trends within the regional groundwater system and more distinct hydrogeologic units. Many of the fractures observed within the borehole logs are steeply dipping across Wake County, although open fractures with shallow dip angles were also observed in most rock classes. Regional geologic structural trends were observed in proximity to the Jonesboro Fault. Potential groundwater recharge in the study area was estimated using a Soil-Water-Balance (SWB) model, as well as using base flow hydrograph separation. The SWB model calculated net infiltration below the root zone for 1981 through 2019 for a 5,402-square-mile area that extends to the counties surrounding Wake County. The mean annual net infiltration rate for the 39-year period was about 8.6 inches per year for the study area. The mean annual net infiltration results from the SWB model were comparable to annual base flow estimates using the PART hydrograph-separation method at six U.S. Geological Survey streamgages within the study area. Mean annual base flow for all six drainage basins was near 7.5 inches per year and estimates ranged from 2.9 to 8.9 inches. Comparisons of mean annual potential recharge from the SWB model and base flow estimates were generally within 2 inches, except during high flows for most of the drainage basins in the study area.

North Carolina

Multiple-source tracking: Investigating sources of pathogens, nutrients, and sediment in the Upper Little River Basin, Kentucky, water years 2013–14

The South Fork Little River (SFLR) and the North Fork Little River (NFLR) are two major headwater tributaries that flow into the Little River just south of Hopkinsville, Kentucky. Both tributaries are included in those water bodies in Kentucky and across the Nation that have been reported with declining water quality. Each tributary has been listed by the Kentucky Energy and Environment Cabinet—Kentucky Division of Water in the 303(d) List of Waters for Kentucky Report to Congress as impaired by nutrients, pathogens, and sediment for contact recreation from point and nonpoint sources since 2002. In 2009, the Kentucky Energy and Environment Cabinet—Kentucky Division of Water developed a pathogen total maximum daily load (TMDL) for the Little River Basin including the SFLR and NFLR Basins. Future nutrient and suspended-sediment TMDLs are planned once nutrient criteria and suspended-sediment protocols have been developed for Kentucky. In this study, different approaches were used to identify potential sources of fecal-indicator bacteria (FIB), nitrate, and suspended sediment; to inform the TMDL process; and to aid in the implementation of effective watershed-management activities. The main focus of source identification was in the SFLR Basin. To begin understanding the potential sources of fecal contamination, samples were collected at 19 sites for densities of FIB ( E. coli ) in water and fluvial sediment and at 11 sites for Bacteroidales genetic markers (General AllBac, human HF183, ruminant BoBac, canid BacCan, and waterfowl GFD) during the recreational season (May through October) in 2013 and 2014. Results indicated 34 percent of all E. coli water samples ( n =227 samples) did not meet the U.S. Environmental Protection Agency 2012 recommended national criteria for primary recreational waters. No criterion currently exists for E. coli in fluvial sediment. By use of the Spearman’s rank correlation test, densities of FIB in fluvial sediments were observed to have a statistically significant positive correlation with drainage area. As drainage area increased, so did the densities of FIB in the fluvial sediments. There was no statistically significant correlation between drainage area and FIB in water. The human-associated marker (HF183) was found above the detection limit in 26 percent of the samples ( n =120 samples); a higher proportion of positive samples was in the NFLR Basin. The ruminant-associated marker (BoBac) was above the detection limit in 65 percent of samples; a higher proportion of positive samples was in the headwaters of the SFLR Basin. Nutrient yields differed between the SFLR and NFLR Basins. Comparatively, the SFLR Basin produced the largest estimated mean yields of total nitrogen (16,000 pounds per year per square mile (lb/yr/mi 2 ) and nitrite plus nitrate nitrogen (12,500 lb/yr/mi 2 ), and the NFLR Basin produced the largest estimated mean yields of ammonia plus organic nitrogen (4,700 lb/yr/mi 2 ), total phosphorus (1,100 lb/yr/mi 2 ), and orthophosphorus (590 lb/yr/mi 2 ). Nitrate sources in surface water were assessed in both basins using dual-nitrate isotope (nitrogen and oxygen) ratios. Data from the different land uses in the SFLR Basin showed differences in nitrate concentrations and overlapping, but moderately distinct, isotopic signatures. Predominantly forested sites consistently had low nitrate concentrations (median = 0.233 milligrams per liter) with minimal variability, and agricultural sites had the highest nitrate concentrations (median = 7.55 milligrams per liter) with the greatest variability. The median nitrate concentration for sites with mixed land use was 2.66 milligrams per liter. Dual-isotope data for forested sites plotted within ranges characteristic of soil-derived nitrate with possible but minimal influence from recycled atmospheric nitrate. Ranges of dual-isotope data for sites with agricultural and mixed land uses were characteristic of possible mixtures of chemical fertilizer, soil-derived nitrate, and manure and septic wastes. In the NFLR Basin, a positive linear relation was observed between nitrate concentrations and nitrogen isotope ratios (δ 15 N NO3 ) (R 2 =0.56; p -value <0.001) that potentially suggests the NFLR Basin has a higher proportion of δ 15 N NO3 -enriched sources, such as manure and sewage. However, mixing of other nitrate-derived sources cannot be excluded, because many values of δ 15 N NO3 and concentrations of nitrate showed minimal variation and plotted within dual-nitrate isotope ranges characteristic of fertilizer and soil-derived nitrate sources. A sediment-fingerprinting approach was used to quantify the relative contribution of four upland sources in the SFLR Basin (agricultural, pasture, riparian/forest, and streambank) to understand how land management affects suspended-sediment concentration. Carbon isotope ratios (δ 13 C), together with calcium and carbon concentrations, were the best indicators of sediment source; the uncertainty was less than 11 percent. Fine-sediment samples collected at the SFLR Basin outlet indicated streambanks as the largest source of the fine sediment to the stream followed by cropland and riparian/forest-source areas, respectively; pasture was a minor contributing source. Streambanks and cropland were essentially equal contributors of fine sediment at the NFLR Basin outlet.

Kentucky

A national streamflow network gap analysis

The U.S. Geological Survey (USGS) conducted a gap analysis to evaluate how well the USGS streamgage network meets a variety of needs, focusing on the ability to calculate various statistics at locations that have streamgages (gaged) and that do not have streamgages (ungaged). This report presents the results of analysis to determine where there are gaps in the network of gaged locations, how accurately desired statistics can be calculated with a given length of record, and whether the current network allows for estimation of these statistics at ungaged locations. The analysis indicated that there is variability across the Nation’s streamflow data-collection network in terms of the spatial and temporal coverage of streamgages. In general, the Eastern United States has better coverage than the Western United States. The arid Southwestern United States, Alaska, and Hawaii were observed to have the poorest spatial coverage, using the dataset assembled for this study. Except in Hawaii, these areas also tended to have short streamflow records. Differences in hydrology lead to differences in the uncertainty of statistics calculated in different regions of the country. Arid and semiarid areas of the Central and Southwestern United States generally exhibited the highest levels of interannual variability in flow, leading to larger uncertainty in flow statistics. At ungaged locations, information can be transferred from nearby streamgages if there is sufficient similarity between the gaged watersheds and the ungaged watersheds of interest. Areas where streamgages exhibit high correlation are most likely to be suitable for this type of information transfer. The areas with the most highly correlated streamgages appear to coincide with mountainous areas of the United States. Lower correlations are found in the Central United States and coastal areas of the Southeastern United States. Information transfer from gaged basins to ungaged basins is also most likely to be successful when basin attributes show high similarity. At the scale of the analysis completed in this study, the attributes of basins upstream of USGS streamgages cover the full range of basin attributes observed at potential locations of interest fairly well. Some exceptions included very high or very low elevation areas and very arid areas.

Puerto Rico

Hydrogeology of the Susquehanna River valley-fill aquifer system in the Endicott-Vestal area of southwestern Broome County, New York

The village of Endicott, New York, and the adjacent town of Vestal have historically used groundwater from the Susquehanna River valley-fill aquifer system for municipal water supply, but parts of some aquifers in this urban area suffer from legacy contamination from varied sources. Endicott would like to identify sites distant from known contamination where productive aquifers could supply municipal wells with water that would not require intensive treatment. The distribution or geometry of aquifers within the Susquehanna River valley fill in western Endicott and northwestern Vestal are delineated in this report largely on the basis of abundant borehole data that have been compiled in a table of well records. Early in deglaciation, meltwater deposited sand and gravel in channels within or beneath the decaying ice and as narrow terraces along the valley walls. These ice-contact deposits vary widely over short distances from clean (free of silt) and highly permeable to clogged with silt and poorly permeable, but collectively constitute the principal aquifers in Endicott and Vestal. Some ice-contact deposits form a buried ridge, deposited in a meltwater channel within the ice sheet, that approximately underlies the Susquehanna River and (or) its north bank from Endwell westward to Nanticoke Creek and has been tapped by several municipal and industrial wells. Similar but thinner ice-contact deposits discontinuously underlie the valley floor to the south in Vestal, and a smaller buried ridge of ice-contact deposits is likely beneath or west of Nanticoke Creek south of West Corners. As deglaciation continued, a large lake developed; thick deposits of gray silt with red clay layers are continuous north of the Susquehanna River from Endwell to West Endicott, and similar deposits are present discontinuously elsewhere. Late in deglaciation, meltwater deposited highly permeable pebbly sand atop the valley fill, generally atop lacustrine silt. The saturated thickness of this surficial sand is seldom great enough to support large-capacity wells, but where it directly overlies ice-contact deposits it facilitates recharge from precipitation and infiltration of river water to the deeper aquifers. Three localities in Endicott were identified where thick ice-contact deposits capable of supporting municipal supply wells are documented by test wells or extrapolated to be present from nearby data and depositional history. Chemical analyses of water samples disclosed no contaminants in these localities when sampled, but the presence of contaminants or natural high iron a few thousand feet away from each locality is documented.

New York

Estimates of groundwater discharge by evapotranspiration, Stump Spring and Hiko Springs, Clark County, southern Nevada, 2016–18

This report documents methodology and results of a study that estimated groundwater discharge by evapotranspiration ( GWET ) from phreatophytic vegetation in two desert riparian areas with ephemeral spring discharge in Clark County, southern Nevada. The phreatophytes consisted primarily of western honey mesquite [ Prosopis glandulosa var. torreyana (L.D. Benson) M.C. Johnst. ] at Stump Spring and mixed shrubs at Hiko Springs. An eddy-covariance station and precipitation gage were established to concurrently measure actual evapotranspiration (AET) and precipitation. Site-scale GWET rates—computed by subtracting measured precipitation from AET—were 239 ±45 millimeters per year (mm/yr) based on measurements over one growing season at Stump Spring and 109 ±27 mm/yr averaged over two growing seasons at Hiko Springs. The volume of GWET for each groundwater discharge area (GDA) was estimated by developing relations between site-scale computed GWET rates and phreatophytic vegetation represented by a Normalized Difference Vegetation Index (NDVI). A GDA was delineated for the natural drainage in each area by mapping the extent of phreatophytes using high-resolution imagery. A second GDA was delineated at Stump Spring by mapping the extent of phreatophytes in the Area of Critical Environmental Concern (ACEC). Site-scale GWET rates were scaled up by applying the site-based GWET -NDVI relations to NDVI distributions in each GDA. The areas of phreatophytic vegetation within each GDA, area-weighted mean GWET rates, and GWET volumes were as follows: (1) Stump Spring—59 hectares (ha), 126 mm/yr, 7.4 ± 1.4 ×104 cubic meters per year (m 3 /yr) (60 ± 11 acre-feet/yr); Stump Spring ACEC—49 ha, 98 mm/yr, 4.9 ± 0.9 × 104 m 3 /yr (39 ± 7 acre-feet/yr); and (2) Hiko Springs—7.2 ha, 112 mm/yr, 0.8 ±0.2 × 104 m 3 /yr (6.6 ±1.6 acre-feet/yr). The GWET rate computed at Stump Spring compared favorably with published GWET rates for mesquite.

Nevada

Assessment of conservation management practices on water quality and observed trends in the Plum Creek Basin, 2010–20

The U.S. Geological Survey and University of Wisconsin–Green Bay collected hydrologic and water-quality data to assess the effectiveness of agricultural conservation management practice (CMP) implementation at mainstem Plum Creek and west Plum Creek in northeastern Wisconsin. These two subbasins cover 88 percent of the Plum Creek Basin (Hydrologic Unit Code 12), which is a subbasin of the lower Fox River Basin. A published total maximum daily load report for the lower Fox River Basin rated Plum Creek as one of the greatest contributors of total suspended solids (TSS) and total phosphorus (TP) draining into the lower Fox River. To reduce TSS and TP exports from Plum Creek, additional cropland conservation practices and watercourse protections were applied between 2012 and 2020. To detect water-quality trends, data were collected during 2010 to 2020 at mainstem Plum Creek and 2013 to 2020 at west Plum Creek. The project used two methods to evaluate CMP effectiveness. The first method focused on evaluating water-quality changes between initial and post-CMP implementation periods during rain- or snowmelt-induced runoff events (hereafter referred to as “events”). In this approach random-forest models were developed to account for environmental factors which influence water quality. Model residuals from the two time periods were compared to determine the significance of water-quality changes associated with CMP implementation for mainstem and west Plum Creek Basins. The second method used a Weighted Regressions on Time, Discharge, and Season time-series approach to examine changes in water quality during the entire study period in mainstem Plum Creek. Results from both methods indicated there were minimal water-quality changes in TSS concentrations and flow-normalized delivery during runoff events during the 10-year period from 2010 to 2020; however, TP concentrations during low streamflow (less than 3 cubic feet per second [ft 3 /s]) may have decreased. The lack of observed improvement may be attributable to any of the following: variability in weather and hydrologic conditions, insufficient post-treatment data, additional cropland being converted to corn production, above average rainfall, streambank degradation, acute and legacy sources of phosphorus from farm fields, excessive/vulnerable manure applications and spills, and point-source discharges.

Wisconsin

Landscape geochemistry near mineralized areas of eastern Alaska: Chapter H in Recent U.S. Geological Survey studies in the Tintina Gold Province, Alaska, United States, and Yukon, Canada--results of a 5-year project

The Pogo lode gold deposit was discovered in eastern Alaska in the early 1990s and provided the opportunity to study elemental distribution and mobility in the natural environment prior to mine development. Studying mineralized systems prior to mining allows us to compare the natural biogeochemical signature in mineralized versus nonmineralized areas. The resultant data and interpretation also provide a baseline for evaluating what, if any, changes in elemental distribution result from development. This report investigates the chemistry of stream water, streambed sediment, and soil in the context of regional bedrock geology. The major-ion chemistry of the waters reflects a rock-dominated aqueous system, and the waters are classified as Ca 2+ and Mg 2+ - HCO 3 - to Ca 2+ and Mg 2+ - SO 4 -2 waters. Creeks draining the gneissic lithologies tend to be more sulfate dominated than those draining the intrusive units. Sulfate also dominated creeks draining mineralized areas; however, the underlying paragneiss unit could be contributing substantially to the sulfate concentration, and the sulfate concentration in these creeks may reflect a complex baltholith-paragneiss boundary rather than mineralization. Arsenic concentrations in bed sediments were elevated in mineralized areas relative to nonmineralized areas. Elevated concentrations of nickel, chromium, iron, manganese, and cobalt appear to reflect the presence of ultramafic rocks in the drainage. In general, aqueous metal concentrations were below the State of Alaska’s Aquatic Life Criteria and Drinking Water Standards, with the exception of arsenic in stream water, which ranged in concentration from less than 1 to 14 micrograms per liter (&mu;g/L) and exceeded the drinking water standard at one site. The arsenic and antimony concentration in the A, B, and C soil horizons ranged from 3 to 410 milligrams per kilogram (mg/kg), 6.1 to 440 mg/kg, and 2 to 300 mg/kg, respectively, for arsenic and 0.4 to 24 mg/kg, 0.6 to 25 mg/kg, and 0.2 to 16 mg/kg, respectively, for antimony. The arsenic and antimony concentrations in stream waters correlate well with the concentrations in soils. However, significantly less arsenic and antimony was extracted from C horizon soils in water leaching experiments, indicating that the arsenic and antimony in the C horizon is present in a less available form than in the A or B horizons. Arsenic and antimony uptake by grayleaf willow (Salix glauca L.) appears minimal, with arsenic concentrations ranging from less than 0.01 to 0.14 mg/kg and antimony concentrations ranging from less than 0.003 to 0.23 mg/kg in willow leaves. In general, the highest concentrations of both arsenic and antimony in water and soils were found near mineralized areas. Elevated arsenic concentrations were also found in bed sediments from mineralized areas. In these sample matrices, the presence of arsenic and (or) antimony was a good indicator of contact with mineralized rock units.

Alaska;Yukon

Climate simulation and flood risk analysis for 2008-40 for Devils Lake, North Dakota

Devils Lake and Stump Lake in northeastern North Dakota receive surface runoff from a 3,810-square-mile drainage basin, and evaporation provides the only major water loss unless the lakes are above their natural spill elevation to the Sheyenne River. In September 2007, flow from Devils Lake to Stump Lake had filled Stump Lake and the two lakes consisted of essentially one water body with an elevation of 1,447.1 feet, about 3 feet below the existing base flood elevation (1,450 feet) and about 12 feet below the natural outlet elevation to the Sheyenne River (1,459 feet). Devils Lake could continue to rise, causing extensive additional flood damages in the basin and, in the event of an uncontrolled natural spill, downstream in the Red River of the North Basin. This report describes the results of a study conducted by the U.S. Geological Survey, in cooperation with the Federal Emergency Management Agency, to evaluate future flood risk for Devils Lake and provide information for developing updated flood-insurance rate maps and planning flood-mitigation activities such as raising levees or roads. In about 1980, a large, abrupt, and highly significant increase in precipitation occurred in the Devils Lake Basin and elsewhere in the Northern Great Plains, and wetter-than-normal conditions have persisted through the present (2007). Although future precipitation is impossible to predict, paleoclimatic evidence and recent research on climate dynamics indicate the current wet conditions are not likely to end anytime soon. For example, there is about a 72-percent chance wet conditions will last at least 10 more years and about a 37-percent chance wet conditions will last at least 30 more years. A stochastic simulation model for Devils Lake and Stump Lake developed in a previous study was updated and used to generate 10,000 potential future realizations, or traces, of precipitation, evaporation, inflow, and lake levels given existing conditions on September 30, 2007, and randomly generated future duration of the current wet period. On the basis of the simulations, and assuming ice-free conditions and calm wind, the Devils Lake flood elevation for an annualized flood risk of 1 percent (analogous to a “100-year” riverine flood) was estimated to be 1,454.6 feet for a 10-year time horizon (2008­­­–17). Therefore, without adjusting for wind or ice, a residence near Devils Lake at elevation 1,454.6 feet has the same chance of being flooded sometime during the next 10 years as a residence at the edge of the 100-year flood plain along a river. Adjusting for the effects of wind or ice, which will increase the flood elevations for many locations near the lakes, was not within the scope of this study.

North Dakota

Basin-scale simulation of current and potential climate changed hydrologic conditions in the Lake Michigan Basin, United States

The Great Lakes Restoration Initiative (GLRI) is the largest public investment in the Great Lakes in two decades. A task force of 11 Federal agencies developed an action plan to implement the initiative. The U.S. Department of the Interior was one of the 11 agencies that entered into an interagency agreement with the U.S. Environmental Protection Agency as part of the GLRI to complete scientific projects throughout the Great Lakes basin. The U.S. Geological Survey, a bureau within the Department of the Interior, is involved in the GLRI to provide scientific support to management decisions as well as measure progress of the Great Lakes basin restoration efforts. This report presents basin-scale simulated current and forecast climatic and hydrologic conditions in the Lake Michigan Basin. The forecasts were obtained by constructing and calibrating a Precipitation-Runoff Modeling System (PRMS) model of the Lake Michigan Basin; the PRMS model was calibrated using the parameter estimation and uncertainty analysis (PEST) software suite. The calibrated model was used to evaluate potential responses to climate change by using four simulated carbon emission scenarios from eight general circulation models released by the World Climate Research Programme&rsquo;s Coupled Model Intercomparison Project phase 3. Statistically downscaled datasets of these scenarios were used to project hydrologic response for the Lake Michigan Basin. In general, most of the observation sites in the Lake Michigan Basin indicated slight increases in annual streamflow in response to future climate change scenarios. Monthly streamflows indicated a general shift from the current (2014) winter-storage/snowmelt-pulse system to a system with a more equally distributed hydrograph throughout the year. Simulated soil moisture within the basin illustrates that conditions within the basin are also expected to change on a monthly timescale. One effect of increasing air temperature as a result of the changing climate was the appreciable increase in the length of the growing season in the Lake Michigan Basin. The increase in growing season will cause an increase in evapotranspiration across the Lake Michigan Basin, which will directly affect soil moisture and late growing season streamflows. Output from the Lake Michigan Basin PRMS model is available through an online dynamic web mapping service available at (http://pubs.usgs.gov/sir/2014/5175/). The map service includes layers for the each of the 8 global climate models and 4 carbon emission scenarios combinations for 12 hydrologic model state variables. The layers are pre-rendered maps of annual hydrologic response from 1977 through 2099 that provide an easily accessible online method to examine climate change effects across the Lake Michigan Basin.

Indiana, Illinois, Michigan, Wisconsin