Scientific Investigations Report 2006–5106
U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2006–5106
The severity and the extent of episodes of low concentrations of DO are greater in the landward regions of Hood Canal (http://www.hoodcanal.washington.edu/documents/document.jsp?id=1334, accessed January 26, 2006). The analysis of long-term annual loads of DIN (discussed in section “Annual Freshwater and Saline Loads of Dissolved Inorganic Nitrogen of Hood Canal, 1971–2002”) suggested that freshwater loads of DIN to the upper layer of Lynch Cove would contribute a greater percentage of the total DIN loads than the percentage in Hood Canal as a whole. However, determination of DIN loads for a more landward reach of Hood Canal required additional information, especially a better understanding of DIN loads from surface water and the strength of net currents that transport DIN in the lower layer. Available resources did not allow for a year long investigation of DIN loads to Lynch Cove, but as part of this study, DIN loads to Lynch Cove from the five previously discussed pathways were calculated for September and October 2004, the critical period in which DO concentrations are lowest there. Understanding DIN loads from (1) precipitation falling directly on marine waters, (2) surface streams flowing into Hood Canal, (3) ground water flowing from adjacent hillsides directly in Hood Canal, (4) shallow subsurface flow from septic systems within 150 m of the shoreline, and (5) transport by the net estuarine circulation in the lower layer of Hood Canal during the critical season contributes to the understanding of the DO problem, although the loads probably do not reflect annual means.
The seaward boundary for the Lynch Cove study area was set as a transect that extended from Sisters Point to the south shore of Lynch Cove at an angle of 45° to the west (fig. 7). The boundary was located there partly because the 300-kH Acoustic Doppler Current Profilers (ADCP) available for the study cannot be used in water deeper than 50 m. About 27.4 km2 of Lynch Cove is included in the area for which loads were calculated. The 158.4 km2 of subbasins include the Union River subbasin, the Mission Creek subbasin, shoreline subbasin 8, 94 percent of shoreline subbasin 7 (7A), and 67 percent of shoreline subbasin 9 (9A).
On August 25, 2004, two 300-kHz ADCPs (RDI Instruments) were deployed within rigid-bottom tripods in the upward-looking mode at 1.2 m above the bottom of the cove to measure current velocity at two ADCP sites. In addition, a CTD probe (conductivity, temperature, and depth) was mounted on each tripod. One tripod was deployed at mid-channel in the constriction off Sisters Point in 54 m of water (ADCP site A in fig. 7) and the second tripod was deployed outside the study area in 45 m of water at mid-channel in an area with a wider cross section off the town of Union (ADCP site B). The ACDP measured current velocity 12 times an hour with a vertical resolution of 1 m from 7 m above the bottom to about 4 m below the surface. The tripod at site A was recovered on October 27, 2004, and the tripod at site B on October 28, 2004. A complete description of the current measurements is presented in Noble and others (2006).
Two water-quality sampling sites, sites L13 and L14 that were numbered following the labeling convention of the HCDOP, were established in the Lynch Cove study area to collect water samples between July and October (fig. 7). A profile of the water column properties was obtained from a small boat with a Seabird 19+ CTD sensor. Water-column properties for all CTD water column profiles and the quality-assurance data for the CTD measurements are shown in Appendix A (at back of report). During each sampling event, one or more water samples from the lower layer of the water column were collected in a 5-L Niskin bottle suspended on a wire.
Samples for analysis of DO were collected immediately after flushing the tube attached to the 5-L Niskin bottle and were preserved onboard before shipping at ambient temperature to the Hood Canal Salmon Enhancement Group in Belfair, Wash. Salinity samples then were collected in tightly sealed glass bottles and shipped at ambient temperature to the University of Washington Marine Chemistry Laboratory. The saline samples for nitrogen nutrient analyses were filtered onboard through a 25-mm, 0.45-µm pore-size inline syringe filter and shipped to the University of Washington Marine Chemistry Laboratory.
DO was measured by the Hood Canal Salmon Enhancement Group using the Winkler titration method. At the University of Washington Marine Chemistry Laboratory, saline samples were analyzed for salinity by electrical conductivity and concentrations of nitrogen nutrients by colorimetry. Nitrate and nitrite were analyzed following the method of Armstrong and others (1967), and ammonia was analyzed following the method of Slawyk and MacIsaac (1972). The University of Washington Marine Chemistry Laboratory received acceptable ratings for nitrate, nitrite, and ammonia in three recent intercomparison exercises (Katherine Krogslund, University of Washington, written commun., 2005). Quality-control samples for the sampling program included field blanks for equipment contamination and replicates and are presented in table 14 (at back of report). The analyses of two filtered distilled water field blanks using calibrations of nutrients in a seawater matrix for colorimetric methods were conducted to detect gross contamination during seawater sample processing. Nitrate, nitrite, and ammonia concentrations in the field blanks were less than the reporting level. The results of replicate samples with nitrate concentrations between <0.002 and 0.18 mg/L were within one reporting unit of each other, although the relative percent difference of one sample with an average nitrate concentration of 0.38 mg/L was 21 percent. The variance of replicate analyses had little effect on the uncertainty of the DIN loading. For the computation of DIN, no contributions for ammonia or nitrite are added if concentrations are less than the reporting level.
Atmospheric wet deposition to the surface of Lynch Cove is calculated using data on the wet deposition of nitrate and ammonia taken from the National Atmospheric Deposition Program site (Hoh Ranger Station; see fig. 1) in the Olympic Mountains (National Atmospheric Deposition Program, 2005). During September and October 2004, the mean wet deposition for the 2 months was 0.05 kg/ha of DIN per month, which is 55 percent of the long-term autumn mean. When this wet-deposition value of DIN is applied to the surface area of Lynch Cove (27.2 km2), and mean monthly DIN load of 0.14 ± 0.05 MT to Lynch Cove is estimated for September and October 2004.
Estimates of DIN loads delivered by streamflow to Lynch Cove are based on streamflow data for Union River and Mission Creek and on estimates of streamflow for three ungaged shoreline subbasins (shoreline subbasins 7A, 8, and 9A) as described in the section “Annual Freshwater and Saline Loads of Dissolved Inorganic Nitrogen to Hood Canal, 1971–2002.” Streamflow data for September and October 2004 for the Union River and October for Mission Creek (table 6) were provided by the Hood Canal Salmon Enhancement Group (Matthew Korb, written commun., 2005). Streamflow in September for Mission Creek is estimated by establishing a regression relation between the mean daily streamflow values in October for the two subbasins. Additionally, because the Union River gaging station was not located at the river mouth, the streamflow values are scaled up using a comparison of an instantaneous discharge measurement made near the mouth to the discharge recorded at the gaging station.
Nutrient concentrations in the Union River were measured in June 2004 (Frans and others, 2006) and were low throughout the part of the river sampled. Ammonia concentrations generally were less than the detection level and nitrate concentrations ranged from 0.12 to 0.28 mg/L of N. The mean calculated monthly DIN load for the Union River for September and October is 0.64 MT of DIN using the mean monthly streamflows (table 6) and a nitrate concentration of 0.20 mg/L of N measured at the mouth of the river.
Nutrient concentrations also were low in storm-water samples collected in the subbasins of Lynch Cove during March 2004 (Frans and others, 2006). The total DIN loads for Mission Creek and the three shoreline subbasins for September and October were calculated using the mean monthly streamflows in table 6 and the DIN concentration of 0.16 mg/L for Mission Creek and the DIN concentration of 0.13 mg/L for the shoreline subbasins. The estimated monthly DIN load for Mission Creek is 0.10 MT of DIN and for the sum of the three shoreline subbasins is 0.18 MT of DIN as N. The mean monthly DIN load from all surface water for September and October 2004 was estimated to be 0.9 ± 0.3 MT.
The monthly loads of DIN from all ground water for September and October 2004 from the five subbasins used to calculate the DIN load to Lynch Cove sum to a DIN load of 1.7 ± 0.7 MT (table 6). This value is estimated by dividing the annual loads (table 2) by 12 after adjusting for the percentages of shoreline subbasins 7 and 9 discharging to Lynch Cove.
Calculations of DIN load from shallow subsurface flow were based on seasonal estimates of the number of people in housing units along the shore of Lynch Cove and per capita nitrogen load (2.95 kg of DIN per person per year). The census tracts in the five Lynch Cove subbasins yield a shoreline population of 2,600 for October. To account for increased occupancy of residences during September, the number of housing units (1,934) is multiplied by an estimated average occupancy rate of 2.2 people per residence for a population of 4,250. The mean monthly inorganic nitrogen loads from septic effluent in shallow subsurface flow to Lynch Cove during September and October were estimated to be 0.84 ± 0.35 MT as N.
Estimating the load of DIN transported by estuarine circulation in Lynch Cove requires estimating water transport in the lower layer of the water column from the net velocity of currents in the area that averages out tidal influences over the cross-sectional area at ADCP site A, and determining the DIN concentrations of saline water moving landward into the cove at water-quality sites L13 and L14.
The estimate of water transport to Lynch Cove in the lower layer requires that the boundary between the upper and lower layers be defined; and the boundary generally is located at a depth around the pyncocline. The pycnocline is the vertical position of large change in density, which is calculated from salinity and temperature measurements obtained by the CTD sensor. Water column properties were determined each month between July and October 2004 (Appendix A), not only to define the boundary between the upper and lower layers, but also to provide the physical context of the DIN data. The bottom of the steepest gradient in temperature and salinity ranged between 5 and 9 m below the water surface at water-quality site L14 (fig. 8). The near-bottom temperature and salinity off Sisters Point were stable until the third week of September, at which time a front of warm, salty, near-bottom water reached the ADCP site A on September 22 (fig. 9). Temperature and salinity of near-bottom water increased steadily from September 22 to the end of the record on October 27, but exhibited short-term oscillations because of the tides. Although the strength and depth of the density gradient changed at water-quality site L14 between August and October 2004 (fig. 8), the boundary between the lower layer and upper water was set at 7 m for the calculation of net advective transport into Lynch Cove.
The directions of the tidal currents at ADCP sites A and B, obtained from ADCP measurements, were aligned with the regional topography, and the frequencies of the tidal components were normal for the coastal regions of Puget Sound (Noble and others, 2006). The relative strengths of the tidal components were consistent with the relative cross-sectional area perpendicular to the primary direction of the tides. Filtering out the tidal currents using a 66-hour running average reveals the weaker sub-tidal currents that describe the short-term average transport of water within the water column. Unlike most estuarine systems, the sub-tidal currents in both the upper and lower layer in Lynch Cove showed flow reversals with periods between 2 and 7 days (fig. 10). The fluctuations in each layer were in phase between sites, with no lag period. In contrast, the fluctuations in the upper layer were 180 degrees out of phase with the lower layer (that is, in opposite directions). The reversals in the sub-tidal currents correlated with the wind pattern measured at Shelton, Wash. (Noble and others, 2006). The fluctuations in sub-tidal currents in Lynch Cove must be better understood in order to properly model nutrient biogeochemistry and its effect on DO.
The advective current in the lower layer can be derived by simply averaging the instantaneous currents over the period of record because the period of record was greater than two lunar cycles. The error for the mean along-shore current depends on the number of independent observations, which in turn depends on the period over which the current was not correlated with itself. Current measurements in Lynch Cove become independent of previous current measurements at the same depth only when the lag period between the two measurements becomes greater than 2.5 days, which yields a degree of freedom of 19 for the record. At ADCP site A off Sisters Point, mean advective currents are strongest at the bottom and decrease with decreasing depth (fig. 11). Only the currents between 33 and 47 m, the depth of recorded current measurements, were significantly different from zero (note that the first current measurements were made 7 m above the bottom). The mean velocity in this depth interval between 33 and 47 m for the period of record was -1.9 ± 1.4 cm/s (in the landward direction). Assuming that the net velocity is uniform across the width of the cross section, the along-shore transport of water across the Sisters Point transect (ADCP site A) between 33 and 47 m was 119 ± 83 m3/s, based on a cross-sectional area of 5,944 m2 relative to the vertical datum NGVD 29.
The mean along-shore velocities at ADCP site B, obtained from the ADCP measurements, were lower than those at ADCP site A because of the much wider cross-sectional area at site B. Below the thermocline, the mean along-shore velocities were not significantly different than zero at any depths. The mean along-shore velocity within 20 m of the bottom was –0.2 ± 0.7 cm/s (in other words, landward flowing).
The passage of the front of the warm salty water also provides evidence for a significant net advection in the region of Sisters Point. The lag time of 7 days between the passage of the front by ADCP site B and passage by ADCP site A (4.3 km) indicates a net speed of 1.0 cm/s. This speed is well within the range measured for the near-bottom waters at site A by the ADCP and between the velocities measured at sites A and B.
The concentration of nitrate in the lower layer off Sisters Point was fairly constant during the deployment of the ADCP (fig. 12). The mean concentration of DIN in lower-layer water from Sisters Point and seaward was 0.42 ± 0.04 mg/L (n = 7), with nitrate constituting at least 90 percent of DIN (table 7). The mean estuarine transport of DIN into Lynch Cove between 33 and 47 m during September and October 2004, calculated from the mean DIN concentration and the net advective transport in the bottom section of the lower layer, was estimated to be 132 ± 93 MT per month.
Not all of the 0.42 mg/L of DIN that was transported into Lynch Cove in the lower layer originated from seawater flowing into Hood Canal from Admiralty Inlet. Paulson and others (1993) found that the DIN concentration at the entrance to Hood Canal was 0.29 mg/L. The mean DIN below the pycnocline in the northern part of Hood Canal between 2000 and 2002, measured by the Washington State Department of Ecology (2004), was 0.27 mg/L, for samples at their site HC006 that were deemed to have contained little remineralized DIN. Thus, about 0.14 mg/L of DIN in the lower layer (0.42 mg/L in water entering Lynch Cove minus 0.28 mg/L in water flowing over the sill) originated from recycling of DIN through biogeochemical processes within Hood Canal. This transport of recycling nitrogen corresponds to 44 ± 31 MT of the 132 ± 93 MT per month of DIN that was estimated to be transported into Lynch Cove.
During late autumn 2004 (September and October), the surface- and ground-water flows from subbasins contributed an average of about 2.6 MT of N per month, or 72 percent of the DIN entering the upper layer of Lynch Cove (table 8). These loads reflected atmospheric sources to the terrestrial landscape (0.82 MT per month), natural biological processes that produce DIN, and the agricultural and residential loads of people living greater than 150 m from the shoreline in the subbasins. The surface-water load of DIN is based on a combination of DIN concentrations measured during the 2004 surface-water component of the study (Frans and others, 2006) and streamflow data collected in autumn 2004 by the Hood Canal Salmon Enhancement Group. The ground-water discharge of nitrate has not been measured in Lynch Cove, so the estimate of ground-water discharge was based on a generalized water balance and median nitrate concentrations from the Hood Canal drainage basin obtained from Statewide databases. Nitrate load from shallow shoreline septic systems was estimated to be about 23 percent of the total nitrogen load to the upper layer of Lynch Cove. The estimates are based on per capita nitrogen discharges from residential septic systems obtained from the literature and population data obtained from the Census Bureau. Inorganic nitrogen in rain falling directly on the surface of Lynch Cove was estimated to contribute about 4 percent of the load to the upper layer, based on autumn 2004 data obtained from the Hoh Ranger Station. The total amount of DIN estimated to have entered the upper layer of Lynch Cove from atmospheric and terrestrial sources during September and October 2004 was 3.6 MT per month.
The estuarine circulation of Hood Canal transported a mean of 132 MT of DIN per month (± 93 MT) into Lynch Cove in the landward-flowing lower layer between 33 and 47 m, of which about two-thirds originated from the saline waters in Admiralty Inlet and about one-third was biological recycled within Hood Canal. However, the DIN in the bottom must mix upward into the euphotic zone in order to be taken up by phytoplankton that settle and consume DO in the lower layer.
For more information about USGS activities in Washington, visit the USGS Washington Water Science Center home page.
U.S.
Department of the Interior | U.S. Geological
Survey
Persistent URL: https://pubs.water.usgs.gov/sir20065106
Page Contact Information: Publications Team
Page Last Modified: Thursday, 01-Dec-2016 19:03:34 EST
