National Water-Quality Assessment (NAWQA) Program Home Page

National Map with
Links to Study Units

Project Publications

Project Main Page

USGS National Water Quality Assessment Data Warehouse

MA-100 --- Abstract

Natural and Human Factors Affecting Shallow Water Quality in Surficial Aquifers in the Connecticut, Housatonic, and Thames River Basins

Water-Resources Investigations Report 98-4042

By Stephen J. Grady and John R. Mullaney

The quality of shallow ground water in the Connecticut, Housatonic, and Thames River Basins in New England and eastern New York is affected by natural factors related to regional variations in the lithologic composition of surficial aquifers, and by human activities related to land use. Ground-water samples from 120 monitoring wells screened in the uppermost parts of surficial aquifers were collected from 1992 to 1995 and analyzed for 177 water-quality variables. Nonparametric statistical procedures were used to compare the concentrations or frequency of detections of 22 continuous (or slightly censored) and 54 censored water-quality variables. These procedures indicate significant differences in the sample populations of 32 of the 76 variables compared for four aquifer-lithologic-composition and three land-use categories.

Surficial aquifers in the study area were classified into crystalline, arkosic, calcareous, or carbonate categories based on the areal distribution of bedrock units of similar mineralogical composition and the direction and extent of glacial and meltwater transport of sediments composing the surficial aquifers. The effects of aquifer-lithologic composition on ground-water quality were evidenced by significant differences in the concentrations or frequency of detection for 18 of 27 naturally occurring water-quality variables in water samples from the four surficial aquifers.

In general, water type (very fresh, slightly acidic) in the crystalline aquifers differs from water types (fresh, slightly alkaline) in the carbonate, calcareous, and arkosic surficial aquifers. Calcium and bicarbonate were the dominant ions in most waters from the surficial aquifers, but their prevalence diminishes from 100 percent of all water samples from carbonate aquifers to 60 percent of water samples from crystalline aquifers. Sulfate and chloride replaced bicarbonate as the principal anion, and sodium was more prevalent than calcium in some water samples from crystalline aquifers, even in natural, undeveloped settings.

Natural water-quality problems associated with some or all surficial aquifers include elevated concentrations of iron and manganese that exceed the respective U.S. Environmental Protection Agency’s Secondary Maximum Contaminant Levels (SMCL) of 300 and 50 micrograms per liter (mg/L), pH values less than the 6.5 lower limit of the SMCL, and radon-222 activities exceeding the 300 picocuries per liter proposed Maximum Contaminant Level (MCL).

Effects of human activities on shallow ground-water quality are related to undeveloped (forested), agricultural, and urban land use, are widespread, and are largely consistent for similar land uses across the study area. Eighteen continuous (or slightly censored) and 8 censored water-quality variables show statistically significant differences for comparisons of sample populations by land use.

Median concentrations of 15 of the 18 water-quality variables with significant land-use effects were lowest in water samples from monitoring wells in undeveloped areas and highest in samples from urban wells. The median dissolved-solids and chloride concentrations for samples from urban areas were significantly higher than respective medians for agricultural areas and undeveloped areas. Fifteen percent of the samples from urban monitoring wells exceeded the 500 milligrams per liter (mg/L) SMCL for dissolved solids. The median chloride concentration for urban areas was more than twice to nearly ten times higher than medians for agricultural and undeveloped areas, respectively. Calcium, magnesium, hardness as CaCO₃, sodium, bicarbonate, alkalinity, DO, sulfate, silica, bromide, and manganese concentrations also were strongly related to land use.

Nitrite plus nitrate concentrations were less than 1.0 mg/L as nitrogen for 95 percent of the samples from undeveloped wells. The median nitrite plus nitrate concentration for ground-water samples from urban areas was 1.1 mg/L and the maximum concentrations was 9.7 mg/L. The median nitrite plus nitrate concentration for samples from agricultural areas was 3.8 mg/L and nearly 13 percent of the samples from monitoring wells in agricultural areas exceeded the 10 mg/L MCL.

Detections of 24 of 85 pesticide compounds (or their metabolites) in samples from one or more of 103 monitoring wells varied significantly by land use. Pesticides were detected more frequently in ground-water samples from agricultural areas than in samples from wells in undeveloped and urban areas. Ninety percent of all pesticide detections were herbicides, with triazine herbicides detected most commonly. Detections of the triazine herbicide atrazine and its metabolite desethylatrazine account for 53 percent of all pesticide detections. Carbamate insecticides comprised more than one-half of the limited insecticide detections, with carbaryl detections most frequent among the carbamates. Ninety-six percent of pesticide concentrations were less than 1.0 mg/L and 41 percent were less than 0.01 mg/L. Only one pesticide detection, atrazine at 3.6 mg/L in one sample from an agricultural well, exceeded a MCL.

The frequency of atrazine, desethylatrazine, metolachlor, and prometon detections varied significantly by land use. Detections of atrazine, desethylatrazine, and metolachlor were significantly more frequent in water samples from agricultural wells than in samples from undeveloped and urban areas. These three herbicide compounds have been used widely in agricultural areas for pre-emergent control of broadleaf weeds in corn fields. Atrazine and metolachlor were the most and the third-most heavily used pesticides in the study area during 1982-85. Undetected or rarely detected in ground-water samples from undeveloped areas and agricultural areas, prometon detections were common (32 percent) in wells from urban areas. Prometon is a nonselective herbicide effectively used for either pre- or post- emergent, perennial broadleaf weed and grass control, and is commonly used on railroad, powerline, pipeline, and roadway right-of-ways.

Detections of 25 of 60 volatile organic compounds (VOCs) in one or more of 86 monitoring wells sampled for these compounds vary by land use. Overall, the frequency of VOC detections was significantly higher in water samples from urban areas and agricultural areas than from undeveloped areas. VOCs were detected in 62 percent of wells in urban areas, 38 percent of wells in agricultural areas, and in only 9 percent of wells in undeveloped areas. Concentrations of VOCs detected in the study area range from less than 0.2 to 300 mg/L and 5 of the 25 compounds detected—benzene, 1,2-dibromoethane, naphthalene, tetrachloroethene, and trichloroethene—exceed MCLs or HAs (health advisories). Eight of the 25 compounds detected are either known, probable, or possible human carcinogens.

Detections of the VOCs methyl tert-butyl ether (MTBE), chloroform, tetrachloroethene, and trichloroethene differ significantly by land use. MTBE, the most frequently detected VOC, was found in 40 percent of wells in urban areas, which was significantly more frequent than in undeveloped and agricultural areas. Nearly all MTBE detections in ground water were in Connecticut and Massachusetts where its use as a gasoline additive is required to reduce atmospheric concentrations of carbon monoxide and ozone from automobile emissions. MTBE concentrations ranged from 0.2 to 2.4  mg/L, well below the 20 mg/L lower limit of the draft lifetime HA. Chloroform was the second most frequently detected VOC. Chloroform concentrations of 0.3 to 3.1 mg/L were measured in 32 percent of wells in urban areas. Tetrachloroethene and trichloroethene were detected only in ground-water samples collected from urban areas.

For additional information contact: