Map showing hydrography of the Meduxnekeag River watershed in northeastern Maine. Colors are used only to differentiate individual subbasins. 

Contexts in source publication

Context 1

... of the area near Houlton, Maine, have observed seasonal episodic blooms of algae and documented elevated concentrations of fecal-coliform bacteria and inorganic nutrients and low dissolved oxygen concentrations in the Meduxnekeag River. Although point and nonpoint sources of urban and agricultural runoff likely contribute to water-quality impairment, the role of shallow groundwater inflows in delivering such contaminants to the Meduxnekeag River has not been well understood. To provide information about possible groundwater inflows to the river, airborne thermal infrared videography was evaluated as a means to identify and classify thermal anomalies in a 25-mile reach of the mainstem and tributaries of the Meduxnekeag River near Houlton, Maine. The U.S. Geological Survey, in cooperation with the Houlton Band of Maliseet Indians, collected thermal infrared images from a single-engine, fixed-wing aircraft during flights on December 3–4, 2003, and November 26, 2004. Eleven thermal anomalies were identified on the basis of data from the December 2003 flight and 17 from the November 2004 flight, which covered the same reaches of stream. Following image analysis, characterization, and prioritization, the georeferenced infrared images of the thermal anomalies were compared to features on topographic maps of the study area. The mapped anomalies were used to direct observations on the ground to confirm discharge locations and types of inflow. The variations in grayscale patterns on the images were thus confirmed as representing shallow groundwater-discharge zones (seeps), outfalls of treated wastewater, or ditches draining runoff from impervious surfaces. Groundwater discharge from shallow aquifers to surface waters represents a substantial environmental component of the flow in most rivers and streams, accounting for as much as 50 percent of average annual streamflow (Winter and others, 1998). Additionally, this discharge is important for maintaining stream base flow and temperature stability during summer dry periods, thereby sustaining critical habitat for native flora and fauna (Winter and others, 1998; Torgersen and others, 1999; Hayashi and Rosenberry, 2002). Groundwater discharge to streams and rivers varies broadly over spatial and temporal scales as a function of geologic and hydrologic setting (Sear and others, 1999). Discharge can occur as diffuse (nonpoint-source) flow or focused (point-source) flow depending on variables such as bedrock and sediment morphology, hydraulic conductivity of the sediment layers and plant communities, streambed topography and permeability, hydraulic gradients driven by precipitation and snowmelt, and stream stage (Fan and others, 2007; Winter and others, 1998; Alley and others, 2002; Harvey and Bencala, 1993; Rosenberry and others, 2000; Conant, 2004; Sear and others, 1999). Groundwater discharge may also be a significant vector for urban- and agricultural-contaminant inputs to surface-water systems (Focazio and others, 1998; Fryar and others, 2000). Contaminant loads delivered to surface waters through groundwater pathways can be a substantial proportion of the total annual load (Taniguchi and others, 1997; Focazio and others, 1998). In a study of the Chesapeake Bay watershed, groundwater nitrate load contributed about half of the total annual nitrogen load to streams in the watershed (Bachman and others, 1998). In addition to geologic setting, the nature and magnitude of these contaminant loads are affected by land use and daily-to-seasonal fluctuations in the hydrologic regime within a particular watershed (Hayashi and Rosenberry, 2002; Fryar and others, 2000; Wroblicky and others, 1998). Conventional methods that use instream data recordings of temperature and specific conductance to detect the source of groundwater provide information that is temporally continuous but constrained to reaches generally less than a kilometer long, with many potential discharge zones remaining undetected (Lowry and others, 2007; Torgersen and others, 2001). Other methods for characterizing groundwater discharge to surface waters have been employed, including seepage-meter measurements (Lee, 1977; Murdoch and Kelly, 2003; Paulsen and others, 2001; Rosenberry and Morin, 2004), piezometric measurements (Lee and Cherry, 1979; Winter and others, 1988; Kelly and Murdoch, 2003), temperature surveys (Lowry and others, 2007; Anderson, 2005; Conant, 2004; Constanz and Stonestrom, 2003; Constanz, 1998), studies with dyes and conservative tracers (Flury and Wai, 2003; Harvey and Bencala, 1993; Lee and others, 1980), and the assessment and modeling of hydrogeochemical data (Fryar and others, 2000; Harvey and Wagner, 2000; Harbaugh, 2005). Thermal infrared (TIR) remote sensing has been used to measure surface-water temperatures and circulation patterns in larger water bodies such as lakes (Anderson and others, 1995; Kay and others, 2005), measure sea-surface temperatures (Donlon and others, 2002), assess reach- and watershed-scale stream-temperature patterns (Atwell and others, 1971; Faux and others, 2001; Kay and others, 2001; Torgersen and others, 2001), identify important habitat-sustaining thermal refugia for fish in streams (Belknap and Naiman, 1998; Torgersen and others, 1999), and identify zones of groundwater discharge into rivers, streams and estuaries (Banks and others, 1996; Urish and Gomez, 2004; Loheide and Gorelick, 2006). Aerial TIR remote sensing has been used successfully in studies in the northeastern U.S. and elsewhere to identify zones of groundwater discharge into surface waters (Roseen and others, 2002; Ballestero and Roseen, 2003; Urish and Gomez, 2004; Mulligan and Charette, 2006; Loheide and Gorelick, 2006; Banks and others, 1996; Torgersen and others, 2001; Raabe and Bialkowska-Jelinska, 2010). Aerial TIR remote-sensing technology is particularly useful in that it provides a spatial context for evaluating relationships between land use and water quality in a watershed (Torgersen and others, 2001). Surface temperatures over long river reaches are easily surveyed at high resolution, and single images can be related to large areas of a water body. TIR sensors are capable of resolving temperature anomalies as small as 0.08 °C at resolutions of less than 1 meter (m) in surface waters; however, they cannot penetrate the water column below the upper 0.1 millimeter (mm) of the water surface (Torgersen and others, 2001; Kay and others, 2005). The detection of thermal anomalies in a water body is possible because of surficial thermal-energy variations caused by disruptions in the normal (in-situ) thermal flow in the stream. TIR imaging of potential groundwater-discharge zones relies on groundwater temperatures, which are thermally stable compared to seasonally variable surface-water temperatures (Banks and others, 1996). TIR imagery is typically displayed as grayscale with shades of gray representing differences in surface temperature. Generally, lighter shades of gray in IR images indicate warmer water, and darker gray cooler water (Banks and others, 1996; Torgersen and others, 2001; Whited and others, 2002). Although point and nonpoint sources of urban and agricultural runoff likely contribute to water- quality impairment, the role of groundwater in delivering these contaminants to the Meduxnekeag River is not well understood. In December 2003 and November 2004, the U.S. Geological Survey (USGS), in cooperation with the Natural Resources Department of the Houlton Band of Maliseet Indians (HBMI), used aerial TIR videography as a means to identify and map spatial distributions of thermal anomalies along a 25-mile (mi) reach of the mainstem and tributaries of the Meduxnekeag River, near Houlton, Maine (fig. 1). Thermal anomalies were characterized and ranked on the basis of grayscale tonal intensities on the TIR images and the locations of the anomalies relative to the main-stem of the river. Aerially imaged thermal anomalies that were evaluated as being potentially significant discharges to the river were confirmed by ground verification; however, this verification did not include in situ measurements of water-surface ...

Context 2

... Houlton Band of Maliseet Indians, Department of Natural Resources, has actively monitored the water quality of the river and its tributaries for more than two decades, observed seasonal episodic blooms of nuisance filamentous algae, and documented elevated concentrations of fecal- coliform bacteria and inorganic nutrients as well as low dissolved oxygen concentrations (unpublished data on file with the HBMI). Several organizations have previously documented water-quality problems in the Meduxnekeag River. The Maine Department of Environmental Protection (DEP) sampled for total phosphorus and other indicators of stream water-quality and identified point sources that could be contributing to impairment of the river (Maine Department of Environmental Protection, 2000). Independent investigators found that algal mats covered as much as 90 percent of the streambed during the summer at the sites they monitored (William Ball, Acheron Engineering, Environmental and Geologic Consultants, written commun., 2001). Fish-consumption advisories have been issued for the Meduxnekeag River because of elevated levels above Maine DEP standards of the pesticide dichloro- diphenyl-trichloroethane (DDT) in fish tissue (Maine Department of Environmental Protection, 2002) and for all Maine rivers because of elevated levels of mercury in fish tissue (Maine Department of Environmental Protection, 1998b). The river’s water quality is affected by sediments and pesticides in agricultural runoff, stormwater runoff and sewer overflow from urban commercial and residential areas, inflows from rural residential septic systems, and wastewater and thermal effects from industrial areas. Schalk and Tornes (2005) suggested that sediment, nutrients, and organic compounds from agricultural, urban, and industrial areas are likely mobilized during high seasonal periods of runoff. The Meduxnekeag River in southeastern Aroostook County in northern Maine drains 516 square miles (mi 2 ) (fig. 1) at its confluence with the St. John River in New Brunswick, Canada. The river begins at Meduxnekeag Lake, 8 mi west of the town of Houlton, Maine. The South Branch joins the Meduxnekeag River near Houlton where the river turns north-northeast, flowing for approximately 10 mi before it crosses the Canadian border to its confluence with the St. John River. Near Houlton, Maine, at the most downstream point where streamflow is measured by the USGS (station ID 01018035), the drainage area is 257 mi 2 (fig. 2). Upstream from the Canadian border, the Meduxnekeag River has a drainage area of 289 mi 2 (Fontaine and others, 1982). The watershed includes 67 mi of the main-stem river and 290 mi of tributaries. Many small lakes in the upstream parts of the watershed are at a higher elevation than the agricultural lands in the lower part of the watershed. Northeastern Maine is characterized by cold winters and short, warm summers. The growing season is 100 to 125 days. Average annual precipitation is about 39 inches (in.), which includes the water equivalent of 95 in. of snow. Average temperatures range from 12 °F in January to 68 °F in July (National Oceanic and Atmospheric Administration, 2002). Although precipitation is distributed fairly evenly throughout the year, most of the annual streamflow occurs during the spring snowmelt period and before evapotranspiration increases following leafout. Snowmelt runoff has been observed to cause severe erosion in late winter and early spring (Southern Aroostook County Soil and Water Conservation District, 1993). Occasional large summer and fall storms can also result in substantial amounts of runoff. The USGS maintains three streamflow gages (or streamgages) in the study area (fig. 2). Streamgage 01017960, the most upstream streamgage, was established in 2003 in cooperation with the Maine DEP, Town of Houlton, and Tate and Lyle Manufacturing above the confluence of the main stem of the Meduxnekeag River and the South Branch. This streamgage is about 2 mi upstream from the town of Houlton, and the watersheds of its major tributaries Mill Stream and Mill Brook are primarily forested (figs. 1, 3). Long-term streamflow statistics for this streamgage were not calculated for this upstream site because of the short period of record. Station 01018000, Meduxnekeag River near Houlton, was active from 1940 to 1982. During this time period, rating curves were established, and periodic measurements of water temperature, specific conductance, and streamflow were made. Station 01018000 was reactivated in 2003, in cooperation with HBMI, for additional streamflow measurements and water-quality monitoring. Of 56 measurements of streamflow on record for station 01018000, 35 were made during the spring months March to May; the median for these measurements was 1,760 cubic feet per second (ft 3 /s). Median measured streamflow during the rest of the year was 101 ft 3 /s. The peak recorded flow at station 01018000 was 6,010 ft 3 /s on April 4, 1976, probably in response to snowmelt runoff. Station 01018035, Meduxnekeag River at Lowery Road near Houlton, Maine, the most downstream main-stem station on the Meduxnekeag River, was established in July 2005, in cooperation with HBMI, for additional streamflow measurements and water-quality monitoring. One town (Houlton, population 5,270 in 2000) and one industry, a manufacturing plant, have permitted outfalls to the Meduxnekeag River (Maine Department of Environmental Protection, 1998a, 2003). Houlton’s municipal wastewater outfall is just downstream from the town limits. The manufacturing plant, which processes food starch (Town of Houlton, 2004), is just downstream from station 01017960 and upstream from the confluence of the South Branch with the main stem of the Meduxnekeag River. Most irrigation of agricultural fields is by withdrawals from the Meduxnekeag River (Matthew Williams, University of Maine Extension, written commun., 2004). The demand for irrigation water, however, puts stress on aquatic habitat during low-flow periods (Aroostook Water and Soil Management Board, 1996). Land cover in the Meduxnekeag River watershed is primarily forest with smaller amounts of agriculture (fig. 3). Forests cover about 79 percent of the watershed; agricultural lands, about 17 percent; and urban areas and open water, about 4 percent (Southern Aroostook County Soil and Water Conservation District, 1993). In 1993, agricultural lands included approximately 23,900 acres of active cropland, 3,900 acres of hay and pasture, and 3,000 acres of grassland. Some 393 farms of all sizes with 2,443 separate fields occupy 30,800 acres of agricultural land. About 20,000 acres of potatoes, most commonly in rotation with grain, are grown on 212 farms. Fifty-two livestock operations support 2,350 animals, mostly dairy or beef cattle. Most of the agricultural land is concentrated in the lower half of the watershed in and downstream of the Houlton area. The general ...