USGS Science⌕ Search

SEARCH · USGS Science

Results for “Soil and Sediment Contamination”

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.

258 records · Page 15Linked to original sources
USGS2001

Relation of Mercury to Other Chemical Constituents in Ground Water in the Kirkwood-Cohansey Aquifer System, New Jersey Coastal Plain, and Mechanisms for Mobilization of Mercury from Sediments to Ground Water

Water from 265 domestic wells that tap the unconfined Kirkwood-Cohansey aquifer system in the Coastal Plain of New Jersey contained concentrations of mercury that are equal to or exceed the U.S. Environmental Protection Agency maximum contaminant level (MCL) of 2 ug/L (micrograms per liter). The wells range in depth from 50 to 200 feet, and are located in 32 discrete, mostly residential, areas that were developed primarily on former agricultural land during the 1950?s through the 1970?s. Concentrations in two other areas exceeded 1 ug/L. Naturally occurring mercury concentrations in ground water from the Kirkwood-Cohansey aquifer system typically are less than 0.01 ug/L, but concentrations in water from some wells were as much as 42 ug/L. No evidence currently exists that conclusively links known point sources such as landfills, industrial operations, or commercial enterprises to most of the elevated concentrations of mercury in ground water in the residential areas. Possible sources of the mercury include pesticides and atmospheric deposition. Analysis of water from wells in 6 of the 34 areas for other constituents indicates that nitrate concentrations also commonly are elevated above background levels (which typically are undetectable at 0.01 milligrams per liter), and exceed the MCL of 10 milligrams per liter in some samples. Several volatile organic compounds (VOCs), including chloroform, also have been measured in water from wells at many of the 34 sites. Analytical results for water samples collected at several depths from boreholes at 2 of the 34 sites indicate elevated concentrations of calcium, magnesium, barium, strontium, nitrate, and chloride, which may be related to both agricultural chemical applications and septic-system effluent. Determinations of tritium and helium concentrations indicate that water containing elevated concentrations of mercury recharged the aquifer between 9.4 and 79 years ago, which includes the period during which many of the 34 sites were undergoing a change from agricultural or undeveloped to residential land use. Batch equilibrium experiments were used to measure adsorption of dissolved mercury, mercuric chloride, and phenylmercuric acetate by aquifer sediments at pH 3.5-4.0, 4.5-5.0, and 5.5-6.0. In nearly all cases, 55 to 95 percent of the mercury added to the sediments was adsorbed. Mercury mobilization from aquifer sediments inoculated with mercury was investigated by leaching the sediments with two concentrations of nitric acid (a component of acid rain), a sodium chloride solution (simulating road salt), and three fertilizer solutions. A solution of 20-20-20 (nitrogenphosphorous-potassium) fertilizer removed virtually all of the mercury with which the sediments had been inoculated. The sodium chloride solution was moderately effective in removing applied mercury from the sediments, as was a solution of nitric acid. A more dilute nitric acid solution and two sodium nitrate fertilizer solutions were less effective. Results of these experiments indicate that mercury adsorbs to aquifer sediments, but that varying amounts can be removed by infiltrating solutions, some of which can be related to specific land uses. Land-use history at the 34 sites generally indicates a change from agricultural or undeveloped settings to residential settings. Whatever the source of mercury to these sites, a change in the geochemical environment of the soil and aquifer brought about by land-use change probably provides mechanisms for mobilizing the mercury from soils and sediments to ground water.

Water-Resources Investigations Report↗
USGS2001

Water-quality assessment of the eastern Iowa basins– Nitrogen, phosphorus, suspended sediment, and organic carbon in surface water, 1996–98

Twelve sites on streams and rivers in the Eastern Iowa Basins study unit were sampled monthly and during selected storm events from March 1996 through September 1998 to assess the occurrence, distribution, and transport of nitrogen, phosphorus, suspended sediment, and organic carbon as part of the U.S. Geological Survey’s National Water-Quality Assessment Program. One site was dropped from monthly sampling after 1996. Dissolved nitrogen and phosphorus were detected in every water sample collected. Nitrate accounted for 92 percent of the total dissolved nitrogen. About 22 percent of the samples had nitrate concentrations that exceeded the U.S. Environmental Protection Agency’s maximum contaminant level of 10 milligrams per liter as nitrogen for drinking-water regulations. The median concentration of total dissolved nitrogen for surface water in the study unit was 7.2 milligrams per liter. The median total phosphorus concentration for the study unit was 0.22 milligram per liter. About 75 percent of the total phosphorus concentrations exceeded the U.S. Environmental Protection Agency recommended total phosphorus concentration of 0.10 milligram per liter or less to minimize algal growth. Median suspended sediment and dissolved organic-carbon concentrations for the study unit were 82 and 3.5 milligrams per liter, respectively. Median concentrations of nitrogen, phosphorus, and suspended sediment varied annually and seasonally. Nitrogen, phosphorus, and suspended-sediment concentrations increased each year of the study due to increased precipitation and runoff. Median concentrations of dissolved organic carbon were constant from 1996 to 1998. Nitrogen concentrations were typically higher in the spring after fertilizer application and runoff. During winter, nitrogen concentrations typically increased when there was little in-stream processing by biota. Nitrogen and phosphorus concentrations decreased in late summer when there was less runoff and in-stream processing of nitrogen and phosphorus was high. Dissolved organic carbon was highest in February and March when decaying vegetation and manure were transported during snowmelt. Suspendedsediment concentrations were highest in early summer (May–June) during runoff and lowest in January when there was ice cover with very little overland flow contributing to rivers and streams. Based on historical and study-unit data, eastern Iowa streams and rivers are impacted by both nonpoint and point-source pollution. Indicator sites that have homogeneous land use, and geology had samples with significantly higher concentrations of total dissolved nitrogen (median, 8.2 milligrams per liter) than did samples from integrator sites (median, 6.2 milligrams per liter) that were more heterogeneous in land use and geology. Samples from integrator sites typically had significantly higher total phosphorus and suspended-sediment concentrations than did samples from indicator sites. Typically, there was very little difference in median dissolved organic-carbon concentrations in samples from indicator and integrator sites. Concentrations of nitrogen and phosphorus varied across the study unit due to land use and physiography. Basins that are located in areas with a higher percentage of row-crop agriculture typically had samples with higher nitrogen concentrations. Basins that drain the Southern Iowa Drift Plain and the Des Moines Lobe typically had samples with higher total phosphorus and suspended-sediment concentrations. Total nitrogen loads increased each year from 1996 through 1998 in conjunction with increased concentrations and runoff. Total phosphorus loads in the Skunk River Basin decreased in 1997 due to less runoff and decreased sediment transport, but increased in 1998 due to higher runoff and increased sediment transport. Total nitrogen and total phosphorus loads varied seasonally. The highest loads typically occurred in early spring and summer after fertilizer application and runoff. Loads were lowest in January and September when there was typically very little runoff to transport nitrogen and phosphorus in the soil to the rivers and streams. Total nitrogen loads contributed to the Mississippi River from the Eastern Iowa Basins during 1996, 1997, and 1998 were 97,600, 120,000, and 234,000 metric tons, respectively. Total phosphorus loads contributed to the Mississippi River from the Eastern Iowa Basins during 1996, 1997, and 1998 were 6,860, 4,550, and 8,830 metric tons, respectively. Suspendedsediment loads contributed to the Mississippi River from the Eastern Iowa Basins during 1996, 1997, and 1998 were 7,480,000, 4,450,000, and 8,690,000 metric tons, respectively. The highest total nitrogen and total phosphorus yields typically occurred in samples from indicator sites. Sampling sites located in drainage basins with higher row-crop percentage typically had higher nitrogen and phosphorus yields. Sites that were located in the Des Moines Lobe and the Southern Iowa Drift Plain typically had higher phosphorus yields, probably due to physiographic features (for example, erodible soils, steeper slopes). Synoptic samples collected during low and high base flow had nitrogen, phosphorus, and organic-carbon concentrations that varied spatially and seasonally. Comparisons of water-quality data from six basic-fixed sampling sites and 19 other synoptic sites suggest that the water-quality data from basic-fixed sampling sites were representative of the entire study unit during periods of low and high base flow when most streamflow originates from ground water.

Iowa, Minnesota↗
USGS2010

Geology, geochemistry, and geophysics of the Fry Canyon uranium/copper project site, southeastern Utah: Indications of contaminant migration

The Fry Canyon uranium/copper project site in San Juan County, southeastern Utah, was affected by the historical (1957–68) processing of uranium and copper-uranium ores. Relict uranium tailings and related ponds, and a large copper heap-leach pile at the site represent point sources of uranium and copper to local soils, surface water, and groundwater. This study was designed to establish the nature, extent, and pathways of contaminant dispersion. The methods used in this study are applicable at other sites of uranium mining, milling, or processing. The uranium tailings and associated ponds sit on a bench that is as much as 4.25 meters above the level of the adjacent modern channel of Fry Creek. The copper heap leach pile sits on bedrock just south of this bench. Contaminated groundwater from the ponds and other nearby sites moves downvalley and enters the modern alluvium of adjacent Fry Creek, its surface water, and also a broader, deeper paleochannel that underlies the modern creek channel and adjacent benches and stream terraces. The northern extent of contaminated groundwater is uncertain from geochemical data beyond an area of monitoring wells about 300 meters north of the site. Contaminated surface water extends to the State highway bridge. Some uranium-contaminated groundwater may also enter underlying bedrock of the Permian Cedar Mesa Sandstone along fracture zones. Four dc-resistivity surveys perpendicular to the valley trend were run across the channel and its adjacent stream terraces north of the heap-leach pile and ponds. Two surveys were done in a small field of monitoring wells and two in areas untested by borings to the north of the well field. Bedrock intercepts, salt distribution, and lithologic information from the wells and surface observations in the well field aided interpretation of the geophysical profiles there and allowed interpretation of the two profiles not tested by wells. The geophysical data for the two profiles to the north of the well field suggest that the paleochannel persists at least 900 m to the north of the heap leach and pond sites. Contamination of groundwater beneath the stream terraces may extend at least that far. Fry Creek surface water (six samples), seeps and springs (six samples), and wells (eight samples) were collected during a dry period of April 16–19, 2007. The most uranium-rich (18.7 milligrams per liter) well water on the site displays distinctive Ca-Mg-SO 4 -dominant chemistry indicating the legacy of heap leaching copper-uranium ores with sulfuric acid. This same water has strongly negative δ 34 S of sulfate (–13.3 per mil) compared to most local waters of –2.4 to –5.4 per mil. Dissolved uranium species in all sampled waters are dominantly U(VI)-carbonate complexes. All waters are undersaturated with respect to U(VI) minerals. The average 234 U/ 238 U activity ratio (AR) in four well waters from the site (0.939±0.011) is different from that of seven upstream waters (1.235±0.069). This isotopic contrast permits quantitative estimates of mixing of site-derived uranium with natural uranium in waters collected downstream. At the time of sampling, uranium in downstream surface water was mostly (about 67 percent) site-derived and subject to further concentration by evaporation. Three monitoring wells located approximately 0.4 kilometer downstream contained dominantly (78–87 percent) site-derived uranium. Distinctive particles of chalcopyrite (CuFeS) and variably weathered pyrite (FeS 2 ) are present in tailings at the stream edge on the site and are identified in stream sediments 1.3 kilometers downstream, based on inspection of polished grain mounts of magnetic mineral separates.

Utah↗
USGS1960

Geology and ground water in Napa and Sonoma Valleys, Napa and Sonoma Counties, California

Napa and Sonoma Valleys are adjacent alluvium-filled valleys about 40 miles northeast of San Francisco. They occupy alined and structurally controlled depressions in the northern Coast Ranges physiographic province and drain south into San Pablo Bay. The valleys are surrounded and underlain by unconsolidated marine and continental sediments and volcanic rocks of Pliocene and Pleistocene age, which are water bearing in large part and together make up relatively extensive ground-water basins. Napa Valley, the eastern valley, is the larger and has a valley-floor area of about 85 square miles. Sonoma Valley has a valley-floor area of about 35 square miles; in addition, about 10 square miles is unreclaimed tidal marsh. The rock units of Napa and Sonoma Valleys are divided into four classes on the basis of their distribution and relative capacity to yield water: (a) Consolidated virtually non-water-bearing chiefly sedimentary (some metamorphic) rocks that range in age from Jurassic ( ?) to Pliocene; (b) marine shale and sand of the Petaluma formation (Pliocene) and the Merced formation (Pliocene and Pleistocene) that do not crop out within Napa or Sonoma Valleys but perhaps are penetrated by some deep wells drilled in Sonoma Valley; (c) Sonoma volcanics of Pliocene age, parts of which are non-water-bearing and parts of which locally yield large quantities of water; and (d) unconsolidated alluvial deposits mainly of Quaternary age. The deposits of classes (c) and (d) contain the most important aquifers in the area. Most of the water used in these valleys is pumped from wells in the younger and older alluvium in the Huichica and Glen Ellen formations. and in the Sonoma volcanics. The principal aquifers are the younger and older alluvium. but appreciable quantities of water are pumped locally from the Sonoma volcanics. The Huichica and Glen Ellen formations yield water in small quantities and at most places supply water only for limited domestic uses. The younger alluvium COIM3ists of interbedded deposits of unconsolidated gravel. sand, silt. and. locally, peat. These deposits underlie the flood plains and channels of the Napa River. Sonoma Creek, and their tributaries, low alluvial fans or plains graded to these streams. and the tidal marshlands. The older alluvium of Napa and Sonoma Valleys is composed of lenticular deposits of unconsolidated and poorly sorted clay, silt. sand. and gravel. Where exposed, claypan or hardpan soil is characteristically developed at the surface. The Huichica formation is composed of deformed continental beds consisting mostly of yellow silt with some interbedded lenses of silt and gravel, and silt and boulders. At the base are beds of redeposited volcanic material, silt, clay, and lenses of coarse gravel and boulders. The Glen Ellen fcrmation consists of alluvial clay, silt, sand, and gravel in clearly stratified but generally lenticular beds. The Sonoma volcanics constitute a thick and highly variate series of continental volcanic rocks, including andesite, basalt, and minor rhyolite flows and interbedded coarse- to fine-grained pyroclastic tuff and breccia, redeposited tuff and pumice, and diatomaceous mud, silt, and sand. A prominent body of rhyolite flows and tuff with some obsidian and perlitic glass, called the St. Helena rhyolite member, occurs in the upper part. Estimated ground-water pumpage for all uses in the year 1949-50 was about 5,600 acre-feet in Napa Valley and about 2,400 acre-feet in Sonoma Valley. Of this quantity the amounts pumped for irrigation were about 2,900 and 1,900 acre-feet, respectively. Water levels in wells in the central parts of the valleys range from a few feet above the land surface to about 50 feet below, with an average of about 25 feet. The quality of the water in most wells is satisfactory for domestic use and irrigation. Locally at the southern end of the valley some contamination of the native waters is caused by the movement of salty wate

Water Supply Paper↗
USGS2001

Occurrence, distribution, and transport of nutrients in Eastern Iowa Rivers

Eleven streams in the Wapsipinicon, Cedar, Iowa, and Skunk River basins in eastern Iowa were sampled monthly from March 1996 through September 1998 to assess the occurrence, distribution, and transport of nitrogen, phosphorus, and suspended sediment as part of the U.S. Geological Survey's National Water-Quality Assessment Program. At least one form of dissolved nitrogen and phosphorus was detected in every sample collected. Nitrate accounted for 92 percent of the dissolved nitrogen. About 22 percent of the samples contained nitrate that exceeded the U.S. Environmental Protection Agency's Maximum Contaminant Level of 10 milligrams per liter (mg/L) as nitrogen in drinking water. The median concentration of total nitrogen for surface water in the study area was 7.2 mg/L. Dissolved phosphorus was predominately in the form of orthophosphate. The median total phosphorus concentration for the study area was 0.22 mg/L. About 75 percent of the total phosphorus concentrations exceeded the U.S. Environmental Protection Agency recommended total phosphorus concentration of 0.10 mg/L or less to minimize algal growth in streams not discharging directly into lakes or impoundments. Median suspended-sediment concentration for the study area was 82 mg/L. Nitrogen, phosphorus, and suspended-sediment concentrations varied annually and seasonally. Nitrogen, phosphorus, and suspended-sediment concentrations increased each year of the study due to increased precipitation and runoff. Concentrations were typically higher in the spring after fertilizer application and runoff. In winter, nitrogen concentrations typically increased when there was little instream assimilation by aquatic plants and algae. Nitrogen and phosphorus concentrations decreased in late summer when there was less runoff and instream assimilation of nitrogen and phosphorus by aquatic plants and algae was high. Suspended-sediment concentrations were highest in early summer during runoff and lowest in January when there was ice cover with very little overland flow contributing to rivers and streams. Streams draining small- to medium-sized watersheds that have uniform land use and geology had significantly (p < 0.05) higher total dissolved-nitrogen concentrations (median, 8.2 mg/L) than did samples from large rivers (median, 6.2 mg/L) that drain basins with mixed land use and geology. Samples from large rivers typically had significantly (p< 0.05) higher total phosphorus and suspended-sediment concentrations than did samples from small rivers and streams. Concentrations varied between drainage basins due to land use and physiography. Basins that are located in areas with a higher percentage of row-crop agriculture typically had samples with higher nitrogen concentrations than did basins with less row-crop agriculture. In addition, basins that drain the Southern Iowa Drift Plain and the Des Moines Lobe typically had samples with higher total phosphorus and suspended-sediment concentrations than did basins that drain other landform regions. Total nitrogen and total phosphorus loads were typically proportional to the size of the drainage basin. Total nitrogen loads increased each year from 1996 through 1998 in conjunction with increased concentrations and runoff. However, total phosphorus loads in the Skunk River Basin decreased in 1997 due to less runoff. Total phosphorus loads followed the same pattern as total nitrogen loads with increases in 1998. Total nitrogen and total phosphorus loads varied seasonally and the highest loads typically occurred in early spring and summer after fertilizer application and runoff. Loads were lowest in January and September when there was little runoff to transport nitrogen and phosphorus to the rivers and streams. Total nitrogen loads contributed to the Mississippi River from the Eastern Iowa Basins during 1996, 1997, and 1998 were 97,000, 120,000, and 230,000 metric tons respectively. Total phosphorus loads contributed to the Mississippi River from the Eastern Iowa Basins during 1996, 1997, and 1998 were 6,900, 4,600, and 8,800 metric tons, respectively. The highest nitrogen and phosphorus yields typically occurred in streams draining small watersheds that were dominated by a single land use and geology. Sampling sites located in drainage basins with higher row-crop percentage typically had higher nitrogen and phosphorus yields. Sites that were located in the Des Moines Lobe and the Southern Iowa Drift Plain typically had higher phosphorus yields probably due to more erodible soils and steeper slopes.

Iowa↗
USGS2022

Sources and characteristics of dissolved organic carbon in the McKenzie River, Oregon, related to the formation of disinfection by-products in treated drinking water

Executive Summary This study characterized the concentration and quality of dissolved organic carbon (DOC) in the McKenzie River, a relatively undeveloped watershed in western Oregon, and its link to forming disinfection by-products (DBPs) in treated drinking water. The study aimed to identify the primary source(s) of DOC in source water for the Eugene Water & Electric Board’s (EWEB) conventional treatment plant on the McKenzie River near river mile 11, upstream of Hayden Bridge. The two classes of regulated compounds examined—trihalomethanes (THMs) and haloacetic acids (HAAs)—form when organic carbon in raw source water reacts with chlorine and (or) bromine during water treatment. The objectives of the study were to: characterize the amount and quality of DOC in the McKenzie River and select tributaries during storms; identify the most common types of carbon using UV-vis spectroscopy and other methods; evaluate optical properties for predicting DBP precursors in surface water; and identify land cover classes or vegetation types that may be important sources of organic carbon and DBP precursors in EWEB’s source water. Eleven storms were sampled synoptically in upstream-to-downstream fashion to provide a “snapshot” of water quality conditions at four sites on the McKenzie River from Frissell Bridge (6 miles downstream from Trail Bridge Reservoir) to the EWEB water treatment plant at Hayden Bridge and nine contributing tributaries. Storms included late summer and early autumn “first flush” events and late autumn, winter, and spring storms spanning a range in streamflows from 3,000 to 26,000 cubic feet per second as measured in the main stem McKenzie River at the EWEB water intake. Water samples were analyzed for DOC concentrations and optical properties (fluorescence and ultraviolet absorbance [UVA]) across a range of wavelengths to characterize the quantity and quality of dissolved organic matter (DOM) in the McKenzie River at the drinking water intake and upstream locations. Paired sets of source and finished water samples were collected at the EWEB treatment plant to identify DOC quality parameters in raw source water that might predict DBP concentrations in finished drinking water. DOC concentrations were relatively low in the McKenzie River (0.4–3 milligrams per liter [mg/L]; average 1.5 mg/L) but much higher in the tributaries. The highest DOC concentrations occurred during “first flush” storms in October 2012 and September 2013; the highest value (16 mg/L) was measured at the 52nd Street stormwater outfall. The average DOC concentration in the lower basin-tributaries was 3.8 mg/L; three middle basin tributaries—Quartz, Gate, and Haagen Creeks, which drain private forestland with less coniferous forest compared with other higher elevation tributaries— had slightly lower average DOC concentrations (2.8 mg/L). These middle-basin watersheds may be important sources of DOC and DBP precursors to the McKenzie River, even more so than the lower basin tributaries, depending on their flows (and loads). This is particularly true after the September 2020 Holiday Farm fire, which burned much of this area. DOC concentrations increased 68 percent in the McKenzie River between the uppermost reference site at Frissell Bridge and Vida; this includes drainage from Quartz Creek, Blue River Lake and Cougar Reservoir, which all contributed DOC to the main stem. In contrast, the lowermost tributaries draining most of the agricultural and urban land did not have a large effect on DOC in the McKenzie River despite their higher DOC concentrations because of their presumed relatively low streamflows and, consequently, DOC loads. Apart from the continuous flow monitors in the McKenzie River and some tributaries (Blue River and South Fork McKenzie River, and streamflow at Hayden Bridge and Vida, Camp Creek and some other locations), streamflow was not assessed during sample collection for this study. This lack of streamflow data precludes a detailed analysis of loads, which is discussed in the future studies section. All DBP concentrations in finished drinking water were less than EPA maximum contaminant levels (MCLs) of 0.080 mg/L for the four trihalomethanes (THM4) and 0.060 mg/L for five haloacetic acids (HAA5). During the 11 storm sampling events the maximum summed concentrations were about 0.040 mg/L for both THM4 and HAA5. Compliance monitoring samples, collected separately by EWEB, yielded some higher concentrations—0.046 mg/L THM4 and 0.047 HAA5—during the December 2012 storm. The corresponding benchmark quotient (BQ) values, which indicate how close a measured DBP concentration is to the MCL, were 0.58 and 0.78, respectively, for THM4 and HAA5. Compared with a similar 2007–08 McKenzie River study that did not target storm events, concentrations of THM4 and HAA5 in finished water were 68 percent and 33 percent higher, respectively, during the current study. Due to the high dilution rates in the McKenzie River main stem, many of the individual fluorescence excitation-emission measurements were low (<0.1 Raman units) and approached analytical detection limits. Parallel factor analysis (PARAFAC) resulted in a five-component model (C1–C5) that represents five unique organic fluorophores. Components C1, C2, and C3 represent DOM associated with soil-derived, humic-like, more degraded organic matter. In contrast, components C4 and C5 represent “fresher” DOM, derived from terrestrial and aquatic plants, including algae and cyanobacteria that are common in the McKenzie River and its tributaries and reservoirs. The fluorescence data and PARAFAC modeling suggest that most of the DOC in the McKenzie River originated from terrestrial sources (primarily components C1 and C2). The largest increases in DOC in the main stem occurred in the reach upstream of Vida, from inflows by Quartz Creek, Blue River, South Fork McKenzie River, and other tributaries. Concentrations of DBPs in EWEB’s finished drinking water were positively correlated with DOC concentrations in raw source water (THM4, p <0.05; HAA5, p <0.01) for paired samples collected 12−24 hours apart. DOC concentrations were significantly positively correlated ( p <0.001) with laboratory-based fluorescent dissolved organic matter (fDOM) measurements, suggesting fDOM as a useful parameter for monitoring and predicting DOC concentration in surface water and DBP concentrations in finished water. Of all the PARAFAC components in surface water, C5 had the highest correlations with DBPs in finished water (rho = 0.77–0.84, p <0.01), followed by components C1 and C2 (rho = 0.75 and 0.71, respectively, p <0.01). This C5 carbon is associated with recently produced DOM, possibly from decomposed terrestrial and aquatic vegetation. Model loadings of these three components were considerably higher in the sampled tributaries relative to the main stem McKenzie River, with most of the observed increases in the main stem apparent at Vida. This points to Quartz Creek or other tributaries in the reach between Frissell Bridge and the sampling site near Vida (South Fork McKenzie and Blue Rivers) as potentially key contributors of DOM source material that leads to the production of DBPs in treated drinking water. A limited load analysis showed that the reservoirs contributed 8–37 percent of the instantaneous DOC loads observed at Vida at the time of sampling, which suggests other sources such as Quartz Creek and other streams in the reach between Frissell Bridge and Vida are more important. Random forest analyses identified PARAFAC components C1 and C5 and fluorescence peaks A, C, M, T and N as the best predictors for HAA5 concentrations in finished drinking water, explaining 62.5 percent of the variation. The best predictors for THM4 were C1, C4 + C5, and peaks T, A, and N, which explained 33 percent of the variation. Several land cover and vegetation classes were correlated with DOC concentration and other optical measurements. The percentage of evergreen forest in each of the subwatersheds sampled was negatively correlated ( p <0.001) with DOC concentration and many optical indicators of DOM quantity: UVA 254 , fDOM, and all of the fluorescence peaks. In contrast, mixed (deciduous) forest was positively correlated ( p <0.001) with DOC, fDOM, UVA 254 , and several fluorescence peaks, demonstrating the importance of deciduous leaf fall in generating DOC and DBP precursors. The high level of human activities in the middle and lower portion of the basin—including timber harvesting and road construction on private forestland, agricultural, rural, industrial, and urban development—have resulted in the greatest loss in native coniferous and mixed deciduous forests in the basin. DOC loading from these tributaries and reservoir releases, which contain DOC from terrestrial and aquatic productivity, both enrich the McKenzie River. Concentrations of DOC increased an average of 71 percent (range 30–120 percent) in the McKenzie River between Frissell Bridge, the upstream reference site, and Vida. PARAFAC components C1, C2, and C5—which were correlated with DBPs in finished water—increased, on average, 109–136 percent (range 20–250 percent) in this same Frissell-to-Vida reach. These increases occur from input of tributaries in the middle basin such as Quartz Creek and others, as noted above. Future monitoring, field, and lab studies can improve our understanding of seasonal and spatial sources of organic carbon contributing DBP precursors to the McKenzie River and allow detection of long-term trends resulting from the recent Holiday Farm Fire, which burned 173,393 acres of forestland, including riparian areas along the main stem, and numerous structures, homes, and outbuildings in September 2020. Future studies could examine DOC fluxes and flushing of carbon from the watershed, investigate the role of precipitation amount and intensity in mobilizing carbon and sediment, and evaluate impacts to aquatic communities and human health as part of a post-fire assessment. Other areas ripe for study include evaluating the impacts of potential temperature increases on carbon sequestration and decomposition in the burned and unburned forests and identifying practices that foster sequestration of carbon in forest soils. The use of fluorescence sensors such as fDOM to monitor the concentration and composition of raw water supplies may be improved for detection of specific DBP precursors, to provide continuous and real-time information to treatment plant operators. Future studies that monitor DOM amount and quality, and DBP Formation Potential (FP), particularly during storm events, paired with streamflow measurements, as suggested above, could help identify areas that contribute high DOC loads and thus help managers identify the key areas to focus restoration activities. Other studies could examine treatment options for currently regulated DBPs and potentially unregulated compounds, including advanced biological treatments for their removal. This study was a collaboration between the U.S. Geological Survey (USGS) and EWEB in Eugene, Oregon, with additional funding provided from USGS Cooperative Matching Funds Program.

Oregon↗