Search USGS⌕ Search

SEARCH · Search USGS

Results for “Optical Engineering”

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.

41 records · Page 3Linked to original sources
USGS2008

Validation of the ASTER instrument level 1A scene geometry

An independent assessment of the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) instrument geometry was undertaken by the U.S. ASTER Team, to confirm the geometric correction parameters developed and applied to Level 1A (radiometrically and geometrically raw with correction parameters appended) ASTER data. The goal was to evaluate the geometric quality of the ASTER system and the stability of the Terra spacecraft. ASTER is a 15-band system containing optical instruments with resolutions from 15- to 90-meters; all geometrically registered products are ultimately tied to the 15-meter Visible and Near Infrared (VNIR) sub-system. Our evaluation process first involved establishing a large database of Ground Control Points (GCP) in the mid-western United States; an area with features of an appropriate size for spacecraft instrument resolutions. We used standard U.S. Geological Survey (USGS) Digital Orthophoto Quads (DOQS) of areas in the mid-west to locate accurate GCPs by systematically identifying road intersections and recording their coordinates. Elevations for these points were derived from USGS Digital Elevation Models (DEMS). Road intersections in a swath of nine contiguous ASTER scenes were then matched to the GCPs, including terrain correction. We found no significant distortion in the images; after a simple image offset to absolute position, the RMS residual of about 200 points per scene was less than one-half a VNIR pixel. Absolute locations were within 80 meters, with a slow drift of about 10 meters over the entire 530-kilometer swath. Using strictly simultaneous observations of scenes 370 kilometers apart, we determined a stereo angle correction of 0.00134 degree with an accuracy of one microradian. The mid-west GCP field and the techniques used here should be widely applicable in assessing other spacecraft instruments having resolutions from 5 to 50-meters. ?? 2008 American Society for Photogrammetry and Remote Sensing.

Photogrammetric Engineering and Remote Sensing↗
USGS2011

Borehole geophysical investigation of a formerly used defense site, Machiasport, Maine, 2003-2006

The U.S. Geological Survey, in cooperation with the U.S. Army Corps of Engineers, collected borehole geophysical logs in 18 boreholes and interpreted the data along with logs from 19 additional boreholes as part of an ongoing, collaborative investigation at three environmental restoration sites in Machiasport, Maine. These sites, located on hilltops overlooking the seacoast, formerly were used for military defense. At each of the sites, chlorinated solvents, used as part of defense-site operations, have contaminated the fractured-rock aquifer. Borehole geophysical techniques and hydraulic methods were used to characterize bedrock lithology, fractures, and hydraulic properties. In addition, each geophysical method was evaluated for effectiveness for site characterization and for potential application for further aquifer characterization and (or) evaluation of remediation efforts. Results of borehole geophysical logging indicate the subsurface is highly fractured, metavolcanic, intrusive, metasedimentary bedrock. Selected geophysical logs were cross-plotted to assess correlations between rock properties. These plots included combinations of gamma, acoustic reflectivity, electromagnetic induction conductivity, normal resistivity, and single-point resistance. The combined use of acoustic televiewer (ATV) imaging and natural gamma logs proved to be effective for delineating rock types. Each of the rock units in the study area could be mapped in the boreholes, on the basis of the gamma and ATV reflectivity signatures. The gamma and mean ATV reflectivity data were used along with the other geophysical logs for an integrated interpretation, yielding a determination of quartz monzonite, rhyolite, metasedimentary units, or diabase/gabbro rock types. The interpretation of rock types on the basis of the geophysical logs compared well to drilling logs and geologic mapping. These results may be helpful for refining the geologic framework at depth. A stereoplot of all fractures intersecting the boreholes indicates numerous fractures, a high proportion of steeply dipping fractures, and considerable variation in fracture orientation. Low-dip-angle fractures associated with unloading and exfoliation are also present, especially at a depth of less than 100 feet below the top of casing. These sub-horizontal fractures help to connect the steeply dipping fractures, making this a highly connected fracture network. The high variability in the fracture orientations also increases the connectivity of the fracture network. A preliminary comparison of all fracture data from all the boreholes suggests fracturing decreases with depth. Because all the boreholes were not drilled to the same depth, however, there is a clear sampling bias. Hence, the deepest boreholes are analyzed separately for fracture density. For the deepest boreholes in the study, the intensity of fracturing does not decline significantly with depth. It is possible the fractures observed in these boreholes become progressively tighter or closed with depth, but this is difficult to verify with the borehole methods used in this investigation. The fact that there are more sealed fractures at depth (observed in optical televiewer logs in some of the boreholes) may indicate less opening of the sealed fractures, less water moving through the rock, and less weathering of the fracture infilling minerals. Although the fracture orientation remained fairly constant with depth, differences in the fracture patterns for the three restoration sites indicate the orientation of fractures varies across the study area. The fractures in boreholes on Miller Mountain predominantly strike northwest-southeast, and to a lesser degree they strike northeast. The fractures on or near the summit of Howard Mountain strike predominantly east-west and dip north and south, and the fractures near the Transmitter Site strike northeast-southwest and dip northwest and southeast. The fracture populations for the boreholes on or near the summit of Howard Mountain show more variation than at the other two sites. This variation may be related to the proximity of the fault, which is northeast of the summit of Howard Mountain. In a side-by-side comparison of stereoplots from selected boreholes, there was no clear correspondence between fracture orientation and proximity to the fault. There is, however, a difference in the total populations of fractures for the boreholes on or near the summit of Howard Mountain and the boreholes near the Transmitter Site. Further to the southwest and further away from the fault, the fractures at the Transmitter Site predominantly strike northeast-southwest and northwest-southeast.Heat-pulse flowmeter (HPFM) logging was used to identify transmissive fractures and to estimate the hydraulic properties along the boreholes. Ambient downflow was measured in 13 boreholes and ambient upflow was measured in 9 boreholes. In nine other bedrock boreholes, the HPFM did not detect measurable vertical flow. The observed direction of vertical flow in the boreholes generally was consistent with the conceptual flow model of downward movement in recharge locations and upward flow in discharge locations or at breaks in the slope of land surface. Under low-rate pumping or injection rates [0.25 to 1 gallon per minute (gal/min)], one to three inflow zones were identified in each borehole. Two limitations of HPFM methods are (1) the HPFM can only identify zones within 1.5 to 2 orders of magnitude of the most transmissive zone in each borehole, and (2) the HPFM cannot detect flow rates less than 0.010 + or - 0.005 gal/min, which corresponds to a transmissivity of about 1 foot squared per day (ft2/d). Consequently, the HPFM is considered an effective tool for identifying the most transmissive fractures in a borehole, down to its detection level. Transmissivities below that cut-off must be measured with another method, such as packer testing or fluid-replacement logging. Where sufficient water-level and flowmeter data were available, HPFM results were numerically modeled. For each borehole model, the fracture location and measured flow rates were specified, and the head and transmissivity of each fracture zone were adjusted until a model fit was achieved with the interpreted ambient and stressed flow profiles. The transmissivities calculated by this method are similar to the results of an open-hole slug test; with the added information from the flowmeter, however, the head and transmissivity of discrete zones also can be determined. The discrete-interval transmissivities ranged from 0.16 to 330 ft2/d. The flowmeter-derived open-hole transmissivity, which is the combined total of each of the transmissive zones, ranged from 1 to 511 ft2/d. The whole-well open-hole transmissivity values determined with HPFM methods were compared to the results of open-hole hydraulic tests. Despite the fact that the flowmeter-derived transmissivities consistently were lower than the estimates derived from open-hole hydraulic tests alone, the correlation was very strong (with a coefficient of determination, R2, of 0.9866), indicating the HPFM method provides a reasonable estimate of transmissivities for the most transmissive fractures in the borehole. Geologic framework, fracture characterization, and estimates of hydraulic properties were interpreted together to characterize the fracture network. The data and interpretation presented in this report should provide information useful for site investigators as the conceptual site groundwater flow model is refined. Collectively, the results and the conceptual site model are important for evaluating remediation options and planning or implementing the design of a well field and borehole completions that will be adequate for monitoring flow, remediation efforts, groundwater levels, and (or) water quality. Similar kinds of borehole geophysical logging (specifically the borehole imaging, gamma, fluid logs, and HPFM) should be conducted in any newly installed boreholes and integrated with interpretations of any nearby boreholes. If boreholes are installed close to existing or other new boreholes, cross-hole flowmeter surveys may be appropriate and may help characterize the aquifer properties and connections between the boreholes.

Maine↗
USGS2019

Optical wave gauging using deep neural networks

We develop a remote wave gauging technique to estimate wave height and period from imagery of waves in the surf zone. In this proof-of-concept study, we apply the same framework to three datasets: the first, a set of close-range monochrome infrared (IR) images of individual nearshore waves at Duck, NC, USA; the second, a set of visible (i.e. RGB) band orthomosaics of a larger nearshore area near Santa Cruz, CA, USA; and the third, a set of oblique (unrectified) images from the same site. The network is trained using coincident images and in situ wave measurements. The optical wave gauge (OWG) consists of a deep convolutional neural network (CNN) to extract features from imagery — called a ‘base model’, with additional layers to distill the feature information into lower dimensional spaces, and a final layer of dense neurons to predict continuously varying quantities. Four base models are compared. The OWG is trained for both individual wave height and period, and statistical quantities like significant wave height and peak wave period. The best performing OWG on the IR dataset achieved RMS errors of 0.14 m and 0.41 s for height and period, respectively, capturing up to 98% of the variance in these quantities. The best performing OWG on the visible band rectified dataset achieved RMS errors of 0.08 m and 0.79 s, respectively, for height and period. The same values for the oblique RGB imagery were 0.11 m and 0.81 s for height and period, respectively. Overall, wave height and period accuracy is sensitive to choice of base model; OWGs built upon MobilenetV2 tend to perform worst and those built on Inception-ResnetV2 have the smallest RMS error. The presence or otherwise of residual layers in the model makes little systematic difference to the final OWG accuracy. Smaller batch sizes used in model training tend to result in more accurate OWGs. An out-of-calibration validation, using images associated with wave heights or periods outside the range of values represented in the training data, showed that the ability for OWGs to predict the bottom 5% of low wave heights and the top 5% of high wave heights was reasonably good, but the same was not generally true of wave period. The same framework, not optimized for either dataset, predicts both quantities with high accuracy when trained on imagery, despite the differences in electromagnetic band, perspective, and scale. The OWG estimates wave properties from an image in less than 100 ms on a modestly sized CPU, allowing for the possibility of continuous real-time wave estimates.

California, North Carolina↗
USGS2013

Determination of flow losses in the Cape Fear River between B. Everett Jordan Lake and Lillington, North Carolina, 2008-2010

During 2008-2010, the U.S. Geological Survey conducted a hydrologic investigation in cooperation with the Triangle J Council of Governments Cape Fear River Flow Study Committee and the North Carolina Division of Water Resources to collect hydrologic data in the Cape Fear River between B. Everett Jordan Lake and Lillington in central North Carolina to help determine if suspected flow losses occur in the reach. Flow loss analyses were completed by summing the daily flow releases at Jordan Lake Dam with the daily discharges at Deep River at Moncure and Buckhorn Creek near Corinth, then subtracting these values from the daily discharges at Cape Fear River at Lillington. Examination of long-term records revealed that during 10,227 days of the 1983-2010 water years, 408 days (4.0 percent) had flow loss when conditions were relatively steady with respect to the previous day's records. The flow loss that occurred on these 40 days ranged from 0.49 to 2,150 cubic feet per second with a median flow loss of 37.2 cubic feet per second. The months with the highest number of days with flow losses were June (16. percent), September (16.9 percent), and October (19.4 percent). A series of synoptic discharge measurements made on six separate days in 2009 provided "snapshots" of overall flow conditions along the study reach. The largest water diversion is just downstream from the confluence of the Haw and Deep Rivers, and discharges substantially decrease in the main stem downstream from the intake point. Downstream from Buckhorn Dam, minimal gain or loss between the dam and Raven Rock State Park was noted. Analyses of discharge measurements and ratings for two streamgages-one at Deep River at Moncure and the other at Cape Fear River at Lillington-were completed to address the accuracy of the relation between stage and discharge at these sites. The ratings analyses did not indicate a particular time during the 1982-2011 water years in which a consistent bias occurred in the computations of discharge records that would indicate false flow losses. A total of 34 measured discharges at a streamgage on the Haw River below B. Everett Jordan Lake near Moncure were compared with the reported hourly flow releases from Jordan Lake Dam. Because 28 of 34 measurements were within plus or minus 10 percent of the hourly flow releases reported by the U.S Army Corps of Engineers, use of the current discharge computation tables for reporting Jordan Lake Dam flow releases is generally supported. A stage gage was operated on the Cape Fear River at Buckhorn Dam near Corinth to collect continuous stage-only records. Throughout the study period, flow over the dam was observed along its length, and flow loss within the study reach is not attributed to river-level fluctuations at the dam. Water-use information and (or) data were obtained for five industrial facilities, a regional power utility, two municipalities, one small hydropower facility on the Deep River, and one quarry operation also adjacent to the Deep River. The largest water users are the regional power producer, a small hydropower operation, and the two municipalities. The total water-use diversions for these facilities range from almost 25.5 to 38.5 cubic feet per second (39.5 to 59.5 million gallons per day) during the winter and summer periods, respectively. This range is equivalent to 69 to 104 percent of the 37 cubic feet per second median flow loss. The Lockville hydropower station is on the Deep River about 1 mile downstream from the streamgage near Moncure. Run-of-river operations at the facility do not appear to affect flow losses in the study reach. The largest water user in the study area is a regional power producer at a coal-fired power-generation plant located immediately adjacent to the Cape Fear River just downstream from the confluence of the Haw an Deep Rivers. Comparisons of daily water withdrawals, sup-plied by the regional power producer, and discharge records at a streamgage on the diversion canal indicated many days when consumption exceeded the producer's estimates for the cooling towers. Uncertainty surrounding reasonable estimates of consumption remained in effect at the end of the study. Data concerning evaporative losses were compiled using two approaches-an analysis of available pan-evaporation data from a National Weather Service cooperative observer station in Chapel Hill, North Carolina; and a compilation of reference open-water evaporation computed by the State Climate Office of North Carolina. The potential flow loss by evaporation from the main stem and the Deep River was estimated to be in the range of 4 to 14 cubic feet per second during May through October, equivalent to 10 to 38 percent of the 37 cubic feet per second median flow loss. Daily water-use diversions and evaporation losses were compared to flow-loss occurrences during the period April 2008 through September 2010. In comparing the surface-water, water-use, and evaporation data compiled for 2008-2010, it is evident that documented water diversions combined with flow losses by open-water evaporation can exceed the net flow gain in the study area and result in flow losses from the reach. Analysis of data from a streamgage downstream from the regional power plant on the diversion canal adjacent to the Cape Fear River provided insight into the occurrence of an apparent flow loss at the streamgage at Lillington. Assessment of the daily discharges and subsequent hydrographs for the canal streamgage indicated at least 24 instances during the study when the flows suddenly changed by magnitudes of 100 to more that 200 cubic feet per second, resulting in a noted time-lag effect on the downstream discharges at the Lillington streamgage, beginning 8 to 16 hours after the sudden flow change. A fiber-optic distributed temperature-sensing survey was conducted on the Cape Fear River at the Raven Rock State Park reach August 12-14, 2009, to determine if the presence of diabase dikes were preferentially directing groundwater discharge. No temperature anomalies of colder water were measured during the survey, which indicated that at the time of the survey that particular reach of the Cape Fear River was a "no-flow" or losing stream. An aerial thermal-infrared survey was conducted on the Haw and Cape Fear Rivers on February 27, 2010, from Jordan Lake Dam to Lillington to qualitatively delineate areas of groundwater discharge on the basis of the contrast between warm groundwater discharge and cold surface-water temperatures. Dis-charge generally was noted as diffuse seepage, but in a few cases springs were detected as inflow at a discrete point of discharge. Two reaches of the Cape Fear River (regional power plant and Bradley Road reaches) were selected for groundwater monitoring with a transect of piezometers installed within the flood plain. Groundwater-level altitudes at these reaches were analyzed for 1 water year (October 1, 2009, to September 30, 2010). Data collected as part of this study represent only a brief period of time and may not represent all conditions and all years; however, the data indicate that, during the dry summer months, the Cape Fear River within the study area is losing an undetermined quantity of water through seepage. Analyses completed during this investigation indicate a study reach with complex flow patterns affected by numerous concurrent factors resulting in flow losses. The causes of flow loss could not be solely attributed to any one factor. Among the factors considered, the occurrences of water diversions and evaporative losses were determined to be sufficient on some days (particularly during the base-flow period) to exceed the net gain in flows between the upstream and downstream ends of the study area. Losses by diversions and evaporation can exceed the median flow loss of 3 cubic feet per second, which indicates that flow loss from the study reach is real. Groundwater data collected during 2009-2010 indicate the possibility of localized flow loss during the summer, particularly in the impounded reach above Buckhorn Dam. However, no indication of unusual patterns was noted that would cause substantial flow loss by groundwater and surface-water interaction at the river bottom.

North Carolina↗
USGS2013

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

This study characterized the amount and quality of organic matter in the Clackamas River, Oregon, to gain an understanding of sources that contribute to the formation of chlorinated and brominated disinfection by-products (DBPs), focusing on regulated DBPs in treated drinking water from two direct-filtration treatment plants that together serve approximately 100,000 customers. The central hypothesis guiding this study was that natural organic matter leaching out of the forested watershed, in-stream growth of benthic algae, and phytoplankton blooms in the reservoirs contribute different and varying proportions of organic carbon to the river. Differences in the amount and composition of carbon derived from each source affects the types and concentrations of DBP precursors entering the treatment plants and, as a result, yield varying DBP concentrations and species in finished water. The two classes of DBPs analyzed in this study-trihalomethanes (THMs) and haloacetic acids (HAAs)-form from precursors within the dissolved and particulate pools of organic matter present in source water. The five principal objectives of the study were to (1) describe the seasonal quantity and character of organic matter in the Clackamas River; (2) relate the amount and composition of organic matter to the formation of DBPs; (3) evaluate sources of DBP precursors in the watershed; (4) assess the use of optical measurements, including in-situ fluorescence, for estimating dissolved organic carbon (DOC) concentrations and DBP formation; and (5) assess the removal of DBP precursors during treatment by conducting treatability "jar-test" experiments at one of the treatment plants. Data collection consisted of (1) monthly sampling of source and finished water at two drinking-water treatment plants; (2) event-based sampling in the mainstem, tributaries, and North Fork Reservoir; and (3) in-situ continuous monitoring of fluorescent dissolved organic matter (FDOM), turbidity, chlorophyll- a , and other constituents to continuously track source-water conditions in near real-time. Treatability tests were conducted during the four event-based surveys to determine the effectiveness of coagulant and powdered activated carbon (PAC) on the removal of DBP precursors. Sample analyses included DOC, total particulate carbon (TPC), total and dissolved nutrients, absorbance and fluorescence spectroscopy, and, for regulated DBPs, concentrations of THMs and HAAs in finished water and laboratory-based THM and HAA formation potentials (THMFP and HAAFP, respectively) for source water and selected locations throughout the watershed. The results of this study may not be typical given the record and near record amounts of precipitation that occurred during spring that produced streamflow much higher than average in 2010-11. Although there were algal blooms, lower concentrations of chlorophyll- a were observed in the water column during the study period compared to historical data. Concentrations of DBPs in finished (treated) water averaged 0.024 milligrams per liter (mg/L) for THMs and 0.022 mg/L for HAAs; maximum values were about 0.040 mg/L for both classes of DBPs. Although DBP concentrations were somewhat higher within the distribution system, none of the samples collected for this study or for the quarterly compliance monitoring by the water utilities exceeded levels permissible under existing U.S. Environmental Protection Agency (USEPA) regulations: 0.080 mg/L for THMs and 0.060 mg/L for HAAs. DOC concentrations were generally low in the Clackamas River, typically about 1.0-1.5 mg/L. Concentrations in the mainstem occasionally increased to nearly 2.5 mg/L during storms; DOC concentrations in tributaries were sometimes much higher (up to 7.8 mg/L). The continuous in-situ FDOM measurements indicated sharp rises in DOC concentrations in the mainstem following rainfall events; concentrations were relatively stable during summer base flow. Even though the first autumn storm mobilized appreciable quantities of carbon, higher concentrations of DBPs in finished water were observed 3-weeks later, after the ground was saturated from additional rainfall. The majority of the DOC in the lower Clackamas River appears to originate from the upper basin, suggesting terrestrial carbon was commonly the dominant source. Lower-basin tributaries typically contained the highest concentrations of DOC and DBP precursors and contributed substantially to the overall loads in the mainstem during storms. During low-flow periods, tributaries were not major sources of DOC or DBP precursors to the Clackamas River. Although the dissolved fraction of organic carbon contributed the majority of DBP precursors, at times the particulate fraction (inorganic sediment and organic particles including detritus and algal material) contributed a substantial fraction of DBP precursors. Considering just the main-stem sites, on average, 10 percent of THMFP and 32 percent of HAAFP were attributed to particulate carbon. This finding suggests water-treatment methods that remove particles prior to chlorination would reduce finished-water DBP concentrations to some degree. Overall, concentrations of THM and HAA precursors were closely linked to DOC concentrations; laboratory DBP formation potentials (DBPFPs) clearly showed that THMFP and HAAFP were greatest in the downstream tributaries that contained elevated carbon concentrations. However, carbon-normalized "specific" formation potentials for THMs and HAAs (STHMFP and SHAAFP, respectively) revealed changes in carbon character over time that affected the two types of DBP classes differently. HAA precursors were elevated in waters containing aromatic-rich soil-derived material arising from forested areas. In contrast, THM precursors were associated with carbon having a lower aromatic content; highest STHMFP occurred in autumn 2011 in the mainstem from North Fork Reservoir downstream to LO DWTP. This pattern suggests the potential for a link between THM precursors and algal-derived carbon. The highest STHMFP value was measured within North Fork Reservoir, indicating reservoir derived carbon may be important for this class of DBPs. Weak correlations between STHMFP and SHAAFP emphasize that precursor sources for these types of DBPs may be different. This highlights not only that different locations within the watershed produce carbon with different reactivity (specific DBPFP), but also that different management approaches for each class of DBP precursors could be required for control. Treatability tests conducted on source water during four basin-wide surveys demonstrated that an average of about 40 percent of DOC can be removed by coagulation. While the decrease in THMFP following coagulation was similar to DOC, the decrease in HAAFP was much greater (approximately 70 percent), indicating coagulation is particularly effective at removing HAA precursors'likely because of the aromatic nature of the carbon associated with HAA precursors. Several findings from this study have direct implications for managing drinking-water resources and for providing useful information that may help improve treatment-plant operations. For example, the use of in-situ fluorometers that measure FDOM provided an excellent proxy for DOC concentration in this system and revealed short-term, rapid changes in DOC concentration during storm events. In addition, the strong correlation between FDOM values measured in-situ and HAA5 concentrations in finished water may permit estimation of continuous HAA concentrations, as was done here. As part of this study, multiple in-situ FDOM sensors were deployed continuously and in real-time to characterize the composition of dissolved organic matter. Although the initial results were promising, additional research and engineering developments will be needed to demonstrate the full utility of these sensors for this purpose. In conclusion, although DBPFPs were strongly correlated to DOC concentration, some DBPs formed from particulate carbon, including terrestrial leaf material and algal material such as planktonic species of blue-green algae and sloughed filaments, stalks, and cells of benthic algae. Different precursor sources in the watershed were evident from the data, suggesting specific actions may be available to address some of these sources. In-situ measurements of FDOM proved to be an excellent proxy for DOC concentration as well as HAA formation during treatment, which suggests further development and refinement of these sensors have the potential to provide real-time information about complex watershed processes to operators at the drinking-water treatment plants. Follow-up studies could examine the relative roles that terrestrial and algal sources have on the DBP precursor pool to better understand how watershed-management activities may be affecting the transport of these compounds to Clackamas River drinking-water intakes. Given the low concentrations of algae in the water column during this study, additional surveys during more typical river conditions could provide a more complete understanding of how algae contribute DBP precursors. Further development of FDOM-sensor technology can improve our understanding of carbon dynamics in the river and how concentrations may be trending over time. This study was conducted in collaboration with Clackamas River Water and the City of Lake Oswego water utilities. Other research partners included Oregon Health and Science University in Hillsboro, Oregon, Alexin Laboratory in Tigard, Oregon, U.S. Geological Survey National Research Program Laboratory in Denver, Colorado, and the U.S. Geological Survey Water Science Centers in Portland, Oregon, and Sacramento, California. This project was supported with funding from Clackamas River Water, City of Lake Oswego, the U.S. Geological Survey, and the Water Research Foundation.

Oregon↗