Estimating Groundwater Discharge and Nutrient Loading to Lynch Cove, Hood Canal, Washington Rich Sheibley1,, Don Rosenberry2, James Foreman1, Ed Josberger1, Chris Chickadel3, Bill Simonds4, Steve Cox1, Peter Swarzenski5

Abstract DISTANCE ALONG TRANSECT, IN METERS 0 12 24 36 48 60 72 84 96 108 4 14 24 34 44 54 64 74 84 94 104 A. 0 H. 0 Low dissolved-oxygen concentrations in the waters of Hood Canal, WA (fig. 1) 5.2 5.0 Thermal Infrared Imaging Impact on Lynch Cove 10.4 10.1 threaten marine life in late summer and early autumn. Eutrophication and subsequent 15.5 15.1 20.7 20.2 To determine the spatial distribution of groundwater seeps along the shores of Lynch Cove, we Although the estimation methods vary, groundwater input can be an important component in the nitrogen oxygen depletion in the landward reaches of the Canal (Lynch Cove) has been linked B. 0 12 24 36 48 60 72 84 96 108 I. 4 14 24 34 44 54 64 74 84 94 104 imaged the shoreline of Lynch Cove with an airborne thermal infrared camera (7-12 micrometers budget in Lynch Cove. This and other ungaged sources become increasingly important through the summer, to phytoplankton growth, which is controlled by nutrients (primarily nitrogen) that 0 0 5.2 5.0 wavelength) and a true-color camera at the beginning of a flood tide in September 2008. combining with solar heating to produce a highly stratified surface layer. To evaluate the importance of enter the Canal from various sources. Previous work has shown that seawater entering 10.4 10.1 15.5 SR 15.1 Approximately 3,500 thermal infrared (TIR) and true-color images were collected, although these sources on the oceanography of Lynch Cove, we used a numerical hydrodynamic model of Hood the Canal is the largest source of nitrogen; however, groundwater discharge also may 20.7 ET 20.2 EN M EN only one-half of the images covered the beach. Figures 6A and 6B are examples of the imagery Canal that simulated groundwater input by a series of small point sources distributed along the shoreline contribute significant quantities, particularly during summer months (fig. 2), when 4 14 24 34 44 54 64 74 84 94 104 4 14 24 34 44 54 64 74 84 94 104 C. 0 J. 0

D, I and the corresponding visual image. The thermal-infrared camera resolved 0.1°C temperature of Lynch Cove. The amount of flow was set by the results of our field studies. The sensitivity of Lynch increased nutrient availability in the Canal directly effects eutrophication. 5.2 E 5.0 B

10.4 A 10.1 differences and had a spatial resolution of approximately 2 meters, which easily allowed for the Cove to this previously ignored freshwater source was evaluated by running the model with different S E

R 15.5 15.1 ETE The amount of nitrogen entering Hood Canal from groundwater was estimated using 20.7 W S 20.2 detection of cooler groundwater. The individual images were referenced to an open-water scene levels of freshwater input (table 1). For all cases, the model included the climatological average flow from O LE

N M toward the west end of Lynch Cove. In the images, cooler groundwater appears dark as it flows the Tahuya and Union Rivers (fig. 8). Preliminary results show that groundwater and small streamflows direct and indirect measurements of groundwater discharge, analysis of nutrient 4 14 24 34 44 54 64 74 84 94 104 4 14 24 34 44 54 64 74 84 94 104 H B K. 0 D, I D.

0 T E P 5.0

B 5.2 concentrations, and estimates of denitrification in near-shore sediment. In areas E over the exposed beach and into the warmer surface water. The side-by-side comparison of the increase the stratification in the surface layers and restrict the vertical mixing in Hood Canal. When these A D 10.1

E 10.4 with confirmed groundwater discharge, shore-perpendicular electrical resistivity 15.5 15.1 two types of images clearly shows both the sporadic location and relative strengths of cooler freshwater sources are included in the model, the salinity of the surface water in Hood Canal is reduced, W S 20.2

O 20.7 profiles, manual and automatic seepage-meter measurements, fiber optic distributed B LEH water flowing into Lynch Cove. Visible in the imagery are many small ungaged streams flowing when compared to simulations without the freshwater sources.

4 14 24 34 44 54 64 74 84 94 104 4 14 24 34 44 54 64 74 84 94 104 temperature sensors, and continuous radon measurements were used to characterize T E. 0 L. 0 across the exposed beach, springs and groundwater discharge from alluvial fans, and cooler PE 5.2 5.0 D temporal variations in groundwater discharge over several tidal cycles. Although 10.4 10.1 water flowing from beneath bulkheads in front of homes. The complete set of images available 15.5 15.1 nitrogen concentrations in groundwater are generally low (average 0.5 mg N/L), the 20.7 20.2 online at: http://wa.water.usgs.gov/projects/hoodcanal/seepage.htm. Nutrient Loading to Lynch Cove flux of groundwater discharge associated nitrogen may be large in some areas of the 4 14 24 34 44 54 64 74 84 94 104 1.65 F. 0 Groundwater discharge measurements across all methods were scaled up in order to estimate an annual A B Hood Canal coastline. 5.2 1.60 Full Color Image Thermal-infrared Image 10.4 C nitrogen (N) load from groundwater (table 1). Discharge rates from each method were applied across an 1.55 Figure 3A. Diagram of electromagnetic seepage meter. This 15.5 D intertidal area representing the 5 m tidal range at the site (shaded areas in fig. 8) and multiplied by total Groundwater discharge was estimated using all methods and ranged from 2.5 to 20.7 1.50 meter allows continuous flow measurements during several E 1.45 40 cm/d in the study area. Groundwater discharge carries nitrogen loads to the F dissolved nitrogen concentrations from numerous wells, seeps, and piezometers sampled during the study. tidal cycles. 4 14 24 34 44 54 64 74 84 94 104

AGE, IN METERS 1.40 G. 0 L G 5.2 The N loading estimates were highly variable across methods ranging from 14 to 749 MT/yr. Using a marine environment ranging from 14 to 749 MT/yr. However, some of this nitrogen 1.35 10.4 H K TIDAL ST 1.30 zonal approach (fig. 8), estimates of N loading from the terminus of Lynch Cove are much greater than is attenuated by sediment denitrification. A recent pilot study of denitrification 15.5 I J potential in the nearshore sediments indicated that sediments can remove much of 20.7 1.25 Midpoint time of resistivity profile those from the north and south shores. However, denitrification rates of the nearshore sediment can EXPLANATION 1.20

:00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 :00 00 00 :00 00 00 00 00 process on average 1,480 MT/yr (range 0 to 4800 MT/yr) of nitrate, implying that groundwater nitrate 2 4 6 8 6 8 2 4 6 8 2 4 2 4 6 8 2 4 2 4 6: 8: 2: 4: 6: 8: this groundwater-derived nitrogen as it flows into the surface water system under 50.0 ELECTRICAL 10 1 1 1 1 20 2 2 10 1 1 1 1 20 2 2 10 1 1 1 1 13.3 RESISTIVITY, June 06, 2006 June 07, 2006 June 08, 2006 IN OHM-METERS may not be as important as originally suspected. ideal conditions. These data demonstrate that groundwater discharge to Hood Canal 3.5 WATER DEPTH AT THE SEAWARD END OF THE RESISTIVITY CABLE—MERRIMONT INTENSIVE STUDY SITE is highly variable in space and time because of the heterogeneous local geology, the 0.94 0.25 Table 1. Summary of nitrogen loading into Lynch Cove from groundwater variable hydraulic gradients in the groundwater system adjacent to the shoreline, Figure 4. Electrical-resistivity profiles perpendicular to the shoreline at the Merrimont intensive study site and a large tidal range of 3 to 5 meters. Therefore, a more refined understanding near Sisters Point, Lynch Cove area of Hood Canal, Washington. Each profile represents a snapshot of electrical resistivity along a cross-section of the intertidal area. Colors represent electrical resistance in ohm- Method Groundwater Nitrogen Reference of the nutrient loads entering Hood Canal from submarine groundwater discharge, meters. Red colors indicate more electrically resistant freshwater, and blue colors indicate more electrically flux (cm/day) load* particularly in summer, is an important component of the overall nutrient budget in conductive saline water. Approximate position of the waterline on the beach is indicated by the arrow above Figure 6A. A small stream flowing across an alluvial fan and multiple groundwater sources. The thermal-infrared image (MT/yr) the profile. The plot shows the midpoint time of the data acquisition interval relative to the tidal cycle. has been colorized to show the coldest temperatures in cyan. this system. )(°C Regional water 2.5 14-47 Paulson and balance others, 2006 Lee-type seepage 5.0 28-93 Simonds and others, meters 2008 and unpublished data Electromagnetic 21.2 118-393 Simonds and others, seepage meters 2008 and unpublished data Radon-radium mass 40.4 231-749 Simonds and balance others, 2008 * Range in N loads based on range in total dissolved nitrogen (TDN) concentrations from field surveys. TDN ranged from 0.33 to 1.1 mg N/L. E 123°05' 123° 122°55' 122°50'

EXPLANATION r

e v Intertidal area i R

47° Zone 1, South shore n 27' io Zone 2, Lynch Cove terminus Un 0.03 cm/d Figure 3B. Continuous flow measurements collected in 2-minute intervals Zone 3, North shore Figure 6B. The small bay where the fiber optic Distributed Temperature Sensing (DTS) was deployed (below), showing 7 cm/d 51.0 MT/yr Belfair during several tidal cycles. Average seepage rates, in centimeters per day 0.09 cm/d freshwater inputs along the waterfront. B 8.6 cm/d Thermal-infrared images r ve 51.0 MT/yr Nitrogen loading for zone, Ri 4.3 cm/d in metric tons per year in B total dissolved nitrogen Pilot Study on the use of Fiber-Optic Distributed Temperature Sensing Landon Road

Methods for Measuring Groundwater Discharge 0.9 MT/yr 3.3 cm/d To determine a reliable and efficient method for identifying areas of substantial groundwater 6.5 cm/d

and nutrient discharge to Lynch Cove, we tested fiber-optic (DTS) technology along a stretch 47° Sunset Beach a A 24' y Lynch Cove Direct Groundwater Flux Measurements – Seepage Meters hu of the Lynch Cove shoreline that also had TIR imagery available (fig. 6B). In September 2009, a Several Lee-type seepage meters were installed at each intensive site. A T we deployed about 300 meters of fiber-optic cable on the seabed parallel to the shoreline at a capture bag was used to determine the change in volume over time and 10.5 MT/yr depth just below low-tide elevation. Using DTS technology, we measured the marine-water 3.9 cm/d to estimate groundwater flux (fig. 3). Electromagnetic (EM) seepage temperature at about 1 meter increments along the cable every 10 minutes for 3 days. The Twanoh meters were placed in the intertidal zone at each intensive site and State Park results (fig. 7) show where relatively cold (about 11◦ C) groundwater was discharging to Lynch Sisters collected continuous data at 2-minute intervals over several tidal cycles Point N Cove. These discharge locations were consistent with areas of cold water discharge indicated 13.4 cm/d (fig. 3A and 3B). 0 1 2 3 4 5 KILOMETERS in the TIR imagery as well as conductivity, depth, and temperature (CTD) profiles immediately 47° 21' offshore of the study site (CTD data not shown here). The DTS unit is capable of monitoring 0 1 2 3 4 5 MILES Indirect Groundwater Flux Measurements – Electrical Resistivity temperatures along 4 kilometers of shoreline at a given time, so the technology is appropriate Base from U.S. Geological Survey digital data, 1983, 1:100,000 A stationary electrical resistivity (ER) cable, 112 meters long containing FigureUniversal 8. Transverse Estimates Mercator projection, of nutrient Zone 10 loading to Lynch Cove using a Zone approach and average lee-type Figure 5A. Radon data from a boat survey along Lynch Cove indicating for identifying groundwater discharge zones over large areas. North American Datum of 1983 (NAD 83) seepage fluxes. Intertidal areas of Lynch Cove, based on a 5-meter tidal range, are subdivided into 56 electrodes spaced 2 meters apart, was placed perpendicular to the hot-spots of ground-water discharge (red). shore to see how fresh water and salt water interact in the intertidal zones representing the south shore, the terminus of Lynch Cove, and the north shore areas of Hood Canal, Washington. zone of Lynch Cove. A complete resistivity profile is collected about once per hour during the tidal cycle (fig. 4). References Figure 1. Location of Hood Canal on the west side of the Puget Sound lowland , adjacent to the Olympic Mountains in western Washington. Geochemical Methods (m) Paulson, A.J., Konrad, C.P., Frans, L.M., Noble, M., Kendall, C., Josberger, E.G., Huffman, , Near-continuous radon (222Rn) data were collected during both R.L., and Olsen, T.D., 2006, Freshwater and saline loads of dissolved inorganic nitrogen to , Hood Canal and Lynch Cove, western Washington: U.S. Geological Survey Scientific streaming (fig. 5A) and stationary (fig. 5B) deployments of a , Investigations Report 2006-5106, 92 p. commercial radon in air detector. The amount of 222Rn in a water Atmospheric input , Simonds, F.W., Swarzenski, P.W., Rosenberry, D.O., Reich, C.D., and Paulson, A.J., 2008, Shallow subsurface flow sample is proportional to the quantity of groundwater discharge. Regional groundwater , Estimates of nutrient loading by ground-water discharge into the Lynch Cove area of Hood Surface water Canal, Washington: U.S. Geological Survey Scientific Investigations Report 2008-5078, 54 p. Denitrification Rates A series of laboratory flask studies were conducted to assess the

potential for nearshore sediments to process nitrate. Sediment (m) was collected from approximately 30 sites along Lynch Cove and Contacts denitrification rates were estimated using the acetylene block technique. 1U.S. Geological Survey Washington Water Science Center, 934 Broadway Tacoma, WA 98402 Results from these studies are pending. 2U.S. Geological Survey, Denver Federal Center, MS 413, Bldg. 53, Box 25046 Lakewood, CO 80225 3 University of Washington, Applied Physics Laboratory, 1013 NE 40th St., Box 355640 Seattle, WA 98105-6699 4U.S. Geological Survey, Oregon Water Science Center, 2130 SW 5th Ave., Portland, OR 97201 Air Temperature 5U.S. Geological Survey Pacific Science Center, 400 Natural Bridges Dr., Santa Cruz, CA 95060 Figure 7. Temperatures measured using DTS near the Port of Allyn boat launch on Lynch Cove, Washington. For more information contact Rich Sheibley, [email protected] 253-552-1611; Figure 2. Breakdown by source of monthly loads to Hood Canal estimated from historical data and a regional The temperatures on the left (length 670 – 680 meters) are air temperature at the dock where day and night project website http://wa.water.usgs.gov/projects/hoodcanal/ water balance (Paulson and others, 2006). Figure 5B. Continuous radon (222Rn) data and the calculated advection rates with changing cycles predominate. The thick horizontal blue bands indicate periods of cooler water temperature that tidal levels at a stationary site in Lynch Cove. coincide with incoming tides. The blue vertical “streaks” such as those near the label 750-meter label, indicate areas where a consistent cool water source is present, likely from groundwater discharge that is keeping marine water relatively cool.