1
Hydrographic and dissolved oxygen variability in a seasonal Pacific Northwest estuary
2
3
4
5
6
7
8
9
10
Molly A. O’Neill1 and David A. Sutherland1*
11
Abstract
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Hypoxia is an issue of growing concern for coastal communities. In the California
Current System, a prototypical eastern boundary current, attention has been focused on
explaining the trend in increasing shelf hypoxia. Despite the regional focus on hypoxia in
eastern boundary regions, relatively few studies have examined smaller estuarine
systems. Here, we present results from an observational study in Coos Bay, a small
estuary on the southern Oregon coast, subject to seasonal upwelling/downwelling winds
and wide fluctuations in freshwater input. Coos Bay exhibits characteristics of a saltwedge type estuary under high river discharge conditions (>150 m3 s-1), a well-mixed
estuary under low discharge conditions (0-30 m3 s-1), and partially-mixed estuary during
times of moderate discharge (30-150 m3 s-1). The observed vertical stratification and
along-estuary salinity gradients correlate significantly with river discharge, although the
tidally-averaged estuarine circulation is also sensitive to local wind forcing. Despite a
strong coupling with coastal waters where hypoxia has been present, we did not find
evidence for pervasive hypoxia in Coos Bay. The primary physical driver of seasonal
variability in dissolved oxygen levels is the estuarine exchange flow that controls
estuarine residence times. We find that upwelling on the shelf advects low dissolved
oxygen water into the estuary on synoptic timescales, but that the overall strength of the
upwelling season is not a good predictor of low dissolved oxygen levels in the estuary.
1
Department of Geological Sciences, 1272 University of Oregon, Eugene, OR 974031272, USA, 541-346-8753, [email protected],
*corresponding author
In review.
Keywords dissolved oxygen; hypoxia; estuarine dynamics; upwelling; Pacific
Northwest; Coos Bay estuary
1
35
1. Introduction
36
In the past decade, there has been growing concern about the increase in hypoxia
37
on the mid- and inner continental shelf of the California Current System (CCS) (Chan et
38
al. 2008; Grantham et al. 2004; Bograd et al. 2008; Connolly et al. 2010). Hypoxia occurs
39
when waters become undersaturated in dissolved oxygen (DO), causing organisms to
40
suffer adverse and potentially lethal effects (Rabalais et al. 2010). Depending on the
41
effects and the organisms being assessed, thresholds for hypoxia vary widely in the
42
literature (Vaquer-Sunyer and Duarte 2008; Rabalais et al. 2010). Frequently, a threshold
43
of 2 mg O2 L-1 is cited, which is what we will use here in defining hypoxic waters.
44
In the CCS during the summer, equatorward winds drive upwelling of deep,
45
nutrient-rich and oxygen-poor waters onto the outer shelf (Huyer 1983). Coupled
46
physical and biological processes regulate the DO concentrations of these waters (Bograd
47
et al. 2008; Adams et al. 2013, Monteiro et al. 2006). While outer shelf hypoxia is
48
natural, the recent development of mid- and inner shelf hypoxia is linked to changes in
49
basin-scale atmospheric and oceanic processes that have led to decreases in the oxygen
50
content of upwelled water (Chan et al., 2008; Peterson et al. 2013; Bograd et al. 2008;
51
Pierce et al. 2012), an increase in upwelling-favorable wind stress (Bakun 1990; Snyder
52
et al. 2003), and productivity-driven increases in respiration (Grantham et al. 2004;
53
Thomas et al. 2003). Understanding changes in the conditions of CCS waters has wide
54
implications for other eastern boundary current systems around the world that experience
55
similar dynamics (Epifanio et al. 1983; Chavez and Messié 2009).
56
Despite the overall increase in hypoxic area in the CCS, there is significant along-
57
coast variability in the observed incidence of hypoxia (Peterson et al. 2013). This spatial
58
variability is attributed to wider shelf regions facilitating longer residence times and more
59
organic matter input, thus elevating the potential for the development of hypoxia
60
(Peterson et al. 2013; Barth et al. 2005).
61
Considerable attention has been directed towards understanding the drivers of
62
spatiotemporal variability in hypoxia on the shelf (Adams et al. 2013; Peterson et al.
63
2013; Pierce et al. 2012; Connolly et al. 2010) and in larger estuarine environments like
64
the Columbia River (Roegner et al. 2011) and Hood Canal (Newton et al. 2007) on the
65
US West Coast or the Gulf of Mexico (Rabalais et al. 2002) and Chesapeake Bay (Hagy
2
66
et al. 2004; Scully 2013) on the US East Coast. Much less attention has been given to the
67
vulnerability of smaller coastal estuarine environments in the Pacific Northwest (NOAA
68
1998; Brown and Power 2011). Given the observed along coast variability in DO levels,
69
one might ask which estuaries are most susceptible to intrusions of low-DO water from
70
the shelf, and whether this has occurred simultaneously with the increase in hypoxic area
71
on the inner shelf. These estuaries provide critical habitat for many species of
72
commercially valuable fish. Fishery landing data indicate estuarine species comprised
73
68% of commercial landings and 80% of recreational landings from 2000-2004 (Lellis-
74
Dibble et al. 2008).
75
Here, we focus on Coos Bay, a small estuary on the southern Oregon coast, which
76
is subject to the highly seasonal conditions common throughout the coastal Pacific
77
Northwest. Since no seasonal description of the water properties along the Coos Bay
78
estuary exists, we first focus on identifying the dominant dynamics through a new
79
monthly along-estuary hydrographic surveying dataset coupled with several longer-term
80
time series of water properties. Then we investigate variations in DO levels and compare
81
them with the observed hydrography and circulation. Finally, we put our new data in
82
context with a look at a historic DO dataset that extends back to the late 1950s in Coos
83
Bay.
84
85
2. Study Location
86
The Coos Bay estuary is mesotidal with mixed semidiurnal tides ranging from 2.3
87
m at the mouth to 2.2 m at the city of Coos Bay (Rumrill, 2006). It is located south of
88
Heceta Bank, adjacent to a relatively narrow continental shelf (Fig. 1). High
89
sedimentation rates and tidal fluctuations result in large intertidal areas that make up
90
approximately half of the estuary’s 54 km2 surface area (Hickey and Banas 2003;
91
Rumrill, 2006). These extensive flats, in conjunction with a deep, dredged navigation
92
channel, produce an ebb-dominant system where flood tides are dampened by friction
93
with the flats and ebb tides rush out the channel (Hyde 2007). The tidal currents average
94
1 m s-1, with maximum-recorded currents at 1.7 m s-1 (Baptista 1989).
95
The estuary has one opening to the Pacific Ocean at its southern end, near the
96
town of Charleston (Fig. 1). The main channel extends north, almost parallel to the coast,
3
97
before turning sharply to the southeast near the town of North Bend. The main estuary
98
ends in the town of Coos Bay, 21 km from the mouth. The Coos River is the primary
99
source of freshwater input to the system, although the majority of previous work on the
100
estuary (e.g., Rumrill 2006) has focused on the smaller southern arm, called the South
101
Slough. This southern arm is located just south of the mouth of Coos Bay and is the site
102
of the South Slough National Estuarine Research Reserve (SSNERR).
103
104
105
106
107
108
109
110
Figure 1. Map of Coos Bay with physical features and bathymetry contoured. Distance alongestuary (x) is numbered (km), with select sensor positions indicated. Inset shows Coos Bay’s
location along with other PNW estuaries, as well as Heceta Bank (HB) offshore and the location
the Newport C-MAN meteorological station and NOAA buoy 46015, in 420 m water depth.
3. Data and Methods
3.1. Monthly CTD Transects
111
To describe the seasonal changes in hydrography and DO levels along the estuary,
112
we conducted monthly sampling over a roughly two-year period. During each sampling
113
cruise we obtained along-channel hydrographic sections of salinity and temperature using
4
114
a conductivity/temperature/depth (CTD) sensor. CTD profiles were collected from a 20-
115
foot aluminum boat, the R/V Pugettia, of the Oregon Institute of Marine Biology
116
(OIMB). Data used in this study span sampling cruises starting in Nov. 2012 and
117
continuing though Jul. 2014 (supplementary Table S1 lists information about each of the
118
sampling cruises).
119
For the majority of the fieldwork we used a RBR Titanium XR-620 profiling
120
CTD with three cable-mounted sensors (Rinko DO, Seapoint Turbidity and Seapoint
121
Fluorometer). The sample rate was set at 6 Hz. The instrument was calibrated each year
122
of the fieldwork at the factory. Downcast data were pressure averaged into 1 dbar bins.
123
On a few occasions, a different sensor set up was used (Table S1), including a SeaBird
124
19plus CTD, and a RBR Concerto CTD with DO (bulkhead-mounted Oxyguard). The
125
SeaBird lacked a DO sensor, so for these data no DO data were collected along estuary.
126
Sampling began near the mouth and proceeded up-estuary, increasing in along-
127
estuary distance (x) from the mouth along the main channel (Fig. 1). From Nov. 2012 to
128
Jan. 2014, sampling followed the channel from the mouth past downtown Coos Bay,
129
towards Isthmus Slough (Fig. 1). Beginning in Feb. 2014, the transect was modified to go
130
up the Coos River instead of heading towards Isthmus Slough. The intention was to
131
capture more of the freshwater signal and inputs to Coos Bay.
132
133
3.2. Water Quality Loggers
134
The monthly sampling cruises are adequate to resolve large seasonal differences
135
in water properties, but do not give sufficient temporal resolution to resolve tidally-driven
136
variability or synoptic, weather-driven variability. To put our monthly cruises in context,
137
we obtained records from three YSI data loggers (model 6600) that measure temperature,
138
salinity, DO, turbidity, and pH. These loggers were deployed 0.5 m off the bottom at 3.1,
139
6.9, and 8.4 km from the mouth and provide time-series of water quality along the estuary
140
(Fig. 1). Two of the loggers are maintained by the Confederated Tribes of the Coos,
141
Lower Umpqua, and Siuslaw (CTCLUSI) water quality-monitoring program, which has
142
been sampling continuously since Oct. 2011. One is located at the Empire Docks at
143
43.3942ºN, 124.2804ºW (EMP; x = 6.9 km) in water depth of 6 m, while the other is at
144
the Bureau of Land Management boat ramp at 43.4139ºN, 124.2789ºW (BLM; x = 8.1
5
145
km) on the North Spit of Coos Bay in 5 m water depth. The loggers take measurements
146
every 15 minutes and are quality controlled and processed by the water quality managers
147
for each tribe. Here we use data spanning the 2012 and 2013 water years that cover Oct-
148
2012 to Sept-2013 and Oct-2013 to Sept-2014, respectively. The third water quality
149
logger (YSI model 6600) is by the Charleston Bridge at 43.3380ºN, 124.3210ºW (Fig. 1)
150
and has been maintained by the SSNERR since Apr. 2002 in a similar setup to the other
151
YSI loggers. The SSNERR logger is in 2.5 m water depth and at x = 3.1 km.
152
153
3.3. Acoustic Doppler Current Profiler (ADCP)
154
Although much can be inferred about estuarine dynamics from salinity and
155
temperature gradients along the estuary, direct observations of velocity are invaluable.
156
We gathered a limited dataset of velocity starting in late November 2013, when the
157
SSNERR expanded their sensor network into the greater Coos Bay area. Data from one
158
of these instruments, a 1500 kHz Sontek Argonaut acoustic Doppler current profiler
159
(ADCP), are used here to characterize observed circulation patterns in the main channel
160
of the estuary. The ADCP is located x = 8 km up-estuary from the mouth and slightly out
161
of the channel (Fig. 1). The ADCP is moored to the bottom at 9 m depth looking upward
162
and sampling with 1-meter bins. Data from 25-Nov-2013 to 27-May-2014 are presented
163
here. Note that because the ADCP is not moored in the deepest part of the channel, it
164
potentially misses part of the deepest up-estuary flow.
165
166
3.4. Department of Environmental Quality Data
167
To put our recent data into historical context, we obtained a fifty-year record
168
(1957-2007) of temperature, salinity, and oxygen in Coos Bay from the Oregon
169
Department of Environmental Quality (DEQ) LASAR database (Brown and Power
170
2011). From 1957-1999, data were collected as part of a monitoring program for fecal
171
bacteria in shellfish growing waters. After 1999, the data were collected in partnership
172
with the EPA’s National Coastal Assessment western pilot project. Latitude, longitude,
173
station, date, time, and sampling matrix (surface water or bay/estuary/ocean) were
174
recorded along with DO, salinity, temperature, and percent saturation DO. Unfortunately,
175
however, the depths of each observation were not recorded. DO measurements preceding
6
176
1989 were made through Winkler titrations, while post-1989 rapid-pulse polarographic
177
oxygen sensors were used (pers. comm., Larry Caton, OR DEQ).
178
179
3.5. Environmental Conditions
180
Two main forcing mechanisms, river discharge and along-shelf winds, control
181
estuarine water properties in small PNW estuaries like Coos Bay (Hickey and Banas
182
2003). The most significant shift in wind conditions along the Oregon coast occurs during
183
the spring transition, when along-shelf winds turn predominantly southward and drive
184
upwelling (e.g., Barth et al. 2007). Although variations in upwelling wind strength do
185
exist across the entire West Coast, observations suggest that conditions over the Oregon
186
coastline are coherent. Thus, we use wind conditions taken from a Coastal-Marine
187
Automated Network (C-MAN) station at Newport, OR (Fig. 1), as it represents the most
188
complete time series available. Comparing winds local to Charleston, either measured
189
offshore at NOAA buoy 46015 (located 15 nautical miles west of Port Orford and
190
approximately 75 km southwest of Coos Bay) or from the SSNERR meteorological
191
station (Campbell CR-10X; Rumrill 2006), located in Charleston, would be ideal but
192
these time series have significant gaps during 2012-2013. The Newport C-MAN station
193
data is low-pass filtered to remove diurnal variability and rotated into a coordinate system
194
aligned with the local coastline (Barth et al. 2007). These data show the onset of
195
upwelling-favorable winds and storm events during wintertime that characterize the PNW
196
atmospheric conditions, discussed below in conjunction with the hydrographic and DO
197
results.
198
Discharge data were extracted from the Coos Watershed Association’s (CWA;
199
http://www.cooswatershed.org) river gauge network spanning the water years 2002-2013.
200
From this data, total freshwater input to the estuary was estimated by extrapolating
201
discharge data and drainage areas from the major instrumented tributaries (the South Fork
202
of the Coos River, SFCR; the East Fork of the Millicoma River, EFM; the West Fork of
203
the Millicoma River, WFM; and Marlow Creek, MC) to the total drainage area (~1542
204
km2) of the Coos watershed.
205
206
7
207
208
209
210
211
212
213
214
Figure 2. (a) Along-estuary salinity section (color) in the channel during the fall months, starting
at the mouth (x = 0 km) and ending near the Coos River (x = 22 km). Black lines are dissolved
oxygen contours every 0.5 mg L-1. Black triangles show the locations of CTD casts. (b) Same as
in a, but for a representative springtime transect. (c) Same as in a, but for a representative
wintertime transect. (d) Same as in a, but for a representative late summer transect.
215
4. Results
216
4.1. Hydrography and estuarine classification
217
4.1.1. Seasonal variability
218
Salinity profiles from the CTD transects show the remarkable seasonality in Coos
219
Bay (Fig. 2), and are representative of conditions during each season. The variability in
220
the along-estuary salinity gradient and stratification reveal a system shifting seasonally
221
through different estuarine classifications. In the fall (Fig. 2a), the estuary is partially
222
mixed—isohalines are slightly tilted, and considerable freshwater is present, mixing
223
along the length of the estuary. Wintertime conditions in Coos Bay cause the estuary to
224
become a salt wedge. Fig. 2c shows the much fresher, and much more stratified salinity
225
section. Isohalines are nearly horizontal as large inputs of freshwater drive the salt
226
intrusion down-estuary. Springtime conditions are largely variable (Fig. 2b), dependent
8
227
on the freshwater input and wind conditions. The Apr. 2013 section shown was taken
228
following an upwelling-favorable wind event on the shelf. The depth-averaged along-
229
estuary salinity gradient is large, although the isohalines are nearly vertical and a plug of
230
saltier water was observed in the lower layer at the mouth. Despite the strong along-
231
estuary gradient, stratification was weak, and the estuary was well-mixed. Summertime
232
conditions in Coos Bay are well-mixed, especially by late summer in September (Fig 2d).
233
The salt intrusion extended to the limit of our observations, with the freshest water
234
having S > 30 and vertical isohalines.
235
The monthly CTD transect profiles did not resolve differences due to tidal
236
changes, as each transect took ~2-4 hours to complete. Nonetheless, the along-estuary
237
sections serve as useful tools in understanding hydrographic variation on seasonal
238
timescales, as the seasonality was observed to be repeatable over 2012-2013.
239
The seasonal variability in hydrography can be linked to the interconnected
240
oceanic, atmospheric, and terrestrial forcing on the estuary. Like other systems within the
241
CCS, Coos Bay’s conditions are dominated by the strong seasonal shift in atmospheric
242
pressure that brings poleward winds, large rain events, and downwelling in the winter,
243
and equatorward winds, dry conditions, and upwelling in the summer.
244
245
246
247
248
Figure 3. (a) Daily alongshore wind stress, τalong, measured at Newport, OR, for 2012 (blue) and
2013 (red) water years. The gray line shows the average daily alongshore wind stress from 20022013. (b) Daily river discharge, Qr, for Coos Bay, OR, in 2012 (blue) and 2013 (red). The mean
daily discharge over 2002-2013 is in gray.
9
249
The wind data (Fig. 3a) show the predominantly poleward winds in the wet
250
season and the switch during the spring transition to equatorward winds in the dry season.
251
The seasonal winter winds bring strong precipitation events, which drive high discharge
252
events in Coos Bay from November until late April (Fig. 3). Discharge diminishes to ~2-
253
3 m3 s-1 through the dry season. The observational CTD data showed the manifestation of
254
these conditions in Coos Bay’s water properties (Fig. 2).
255
To summarize the seasonal change in estuarine conditions, Fig. 4 illustrates the
256
along-estuary variation in stratification and salinity as a function of time using all the
257
monthly CTD data. The depth-averaged, along-estuary salinities supports the general
258
picture obtained by examining the full S transects with stronger gradients in the winter
259
months coinciding with the freshest waters observed (Fig. 4a). Previous theory predicts
260
functional dependencies of both the vertical stratification and the along-estuary salinity
261
gradient on the river discharge (MacCready and Geyer, 2010; Monismith et al. 2002;
262
Ralston et al. 2008). There is a slight dependence of the along estuary salinity gradient
263
(∂S/∂x) on Qr, the river discharge (Fig. 4b), where we have taken the section mean for
264
each monthly transect to get a section averaged value of ∂S/∂x. Qr is calculated as the
265
mean discharge over the 4 days preceding the CTD transect. We find that ∂S/∂x ~ Qr0.19
266
(R2 = 0.90, p < 0.01), which is slightly higher than the power of 1/7 (0.14) estimated
267
previously (Monismith et al. 2002; Ralston et al. 2008), but much lower than that
268
predicted by classic estuarine theory (Hansen and Rattray 1965; MacCready and Geyer
269
2010). The correlation does not change significantly if Qr1/7 is used. However, at the
270
highest Qr, there is a hint of saturation in ∂S/∂x, leveling off near 1 psu km-1.
271
The along-estuary vertical stratification, ∂S/∂z, also changes seasonally: during
272
summer and fall months, the stratification is low throughout the estuary, increasing into
273
the winter and spring. There is no persistent spatial trend in ∂S/∂z, although it does
274
increase slightly up estuary in certain winter and spring months. However, this bias might
275
reflect the calculation method of ∂S/∂z as the total depth (Δz) decreases as one approaches
276
the Coos River and exits the dredged channel (Fig. 1). Here, we define the stratification
277
as ΔS/Δz, where ΔS is the difference between S over the upper 2 m and the lower 2 m of
278
the water column (Fig. 4c). For vertical stratification, ∂S/∂z, theory predicts it will
279
increase as ~ Qr2/3 (MacCready and Geyer 2010). We find that the change in stratification
10
280
is well explained by changes in river discharge, ∂S/∂x ~ Qr0.67 (R2 = 0.84, p < 0.01) and
281
compares well with the Qr2/3dependence.
282
283
284
285
286
287
288
289
290
Figure 4. (a) Depth-averaged salinity (S) normalized by the observed salinity at the mouth
(Smouth) as a function of along-estuary distance for the monthly transects (colored by season). The
mouth is at x = 0 km. (b) The mean value of the along-estuary salinity gradient for each monthly
section as a function of river discharge, Qr. The black line shows a power law fit. (c) Vertical
stratification as a function of along-estuary distance for all of the monthly transects (color). The
mouth is at x = 0 km. (d) The mean vertical stratification for each monthly section as a function
of river discharge, Qr. The black line shows a power law fit.
291
4.1.2. 2012/2013 Comparison
292
The strong seasonality in the hydrography of Coos Bay was also apparent in the
293
data from the water quality loggers, which allows us to examine two seasonal cycles for
294
comparison. Fig. 5 shows temperature and salinity (T-S) data from the BLM logger for
295
the 2012 and 2013 water years. The water properties varied dramatically over the course
296
of a year (Fig. 5) in a similar fashion to the picture that emerged from the monthly along-
297
estuary transects, supporting the notion that the monthly transects resolved the seasonal
298
signal. In addition to the seasonal variability, the T-S data reveal clear variation between
299
the two years (Fig. 5).
11
300
301
302
303
304
305
306
Figure 5. Temperature and salinity data from the BLM boat dock logger (x = 9 km, up estuary
from mouth). Each month of data is represented with a different color. (a) 2012 water year data.
(b) 2012 water year data in grey with 2013 water year data overlaid in color.
In 2012, salinity ranged from 2-34 and temperature ranged from 5.7-16.6°C. The
307
influence of freshwater did not start to dominate until February, corresponding to the later
308
onset of large discharge events to the estuary that year (Fig. 3). In 2013, conditions were
309
warmer and saltier overall at the same logger location, with the minimum S = 13.8, the
310
maximum S = 34.2, and temperatures from 6.8-20.4°C. January 2013 experienced a
311
larger range in salinity than in 2012 due to some large, early season discharge events.
312
However, for all other months, the salinity range in 2013 was much reduced from that of
313
2012. Diminished discharge in the mid to late wet season of 2013 might account for the
314
disparity in the observed range of S (Figs. 3-5).
315
The weaker freshwater inflow to the estuary in 2013 likely caused a reduction in
316
the estuarine residual flow and an increase in the extent of the salt intrusion, affecting the
317
local T-S properties at the BLM location. Coupled with the overall weaker summer
318
upwelling season in 2013 (Fig. 3a), the water temperatures were higher in Coos Bay that
319
year. The 2013 upwelling season started and ended earlier (1-Apr Spring transition to a
320
end of August Fall transition) than 2012 (1-May Spring transition to 1-Oct Fall transition)
12
321
based on estimates of upwelling strength on the OR coast (available from
322
http://damp.coas.oregonstate.edu/windstress/). No matter the upwelling strength, a
323
smaller residual flow would prevent the coldest upwelled summer waters from reaching
324
the logger, and also would facilitate longer residence times in the estuary. We present a
325
closer analysis of the residual flow and residence times for these two years below to
326
support this hypothesis.
327
328
329
330
331
332
333
334
Figure 6. Data from 25-Nov-2013 to 27-May-2014 showing the seasonal and synoptic forcing on
flows in Coos Bay. (a) Discharge extrapolated from the Siuslaw to represent the inflows to Coos
Bay. (b) Alongshore wind velocities available from NOAA buoy 46015 (shaded red) and the
SSNERR meteorological station (black). Positive velocities indicate northward wind. (c) Rotated,
along-channel, depth-averaged current velocities, Ualong. Ualong > 0 are out-estuary. (d) Rotated,
along-channel, tidally-averaged current velocities, Ue. The solid black line indicates 0 m s-1.
335
4.2 Estuarine circulation and residence times
336
4.2.1. Seasonal variability
337
The ADCP data provide the most direct insight into the current structure and
338
time-varying flows in Coos Bay (Fig. 6). While the ADCP dataset failed to encompass
13
339
the full range of seasons, it did capture currents from late Nov-2013 to late May 2014—
340
fall, winter, and spring of the 2014 water year. The data show tidal currents on the order
341
of ~1 m s-1 and residual currents on the order of ~0.1 m s-1 (Fig. 6). The dominant
342
variability in the along-estuary flow was due to the tides. A strong spring/neap cycle is
343
apparent in the data (Fig. 6). The residual circulation is seasonally variable. In the late fall
344
and winter months, residual flows are weaker (Fig. 6), corresponding to the low discharge
345
in the estuary from November- February. In spring months, the residual flow strengthens
346
with the arrival of large discharge events in mid February-May.
347
Although the along-estuary residual flow is the dominant component in many
348
estuarine processes, lateral flows can be important to estuarine dynamics in certain
349
systems (MacCready and Geyer 2010). However, in Coos Bay, the velocity data indicate
350
tidally-averaged lateral flow about an order of magnitude weaker than the along-estuary
351
flows. Lateral velocities were typically order 10-2 m s-1, while along-estuary velocities
352
were order 10-1 m s-1.
353
354
4.2.1. 2012/2013 Comparison
355
To investigate the interannual variability in residual flows implied from the T-S
356
data, we approximated residual flows from the long-term time series of salinity between
357
the water quality loggers. We scaled the magnitude of the residual flow, Ue (also called
358
the estuarine exchange flow), using observational data of along-estuary salinity gradients,
359
and the theoretical approximation for partially-mixed estuaries Ue = (gβh03∂S/∂x) / (48
360
Km), where β is the coefficient of expansivity for salinity (7.7 × 10−4 psu−1), g is the
361
gravitational constant 9.8 m s−2, and h0 is water depth (MacCready and Geyer, 2010).
362
Here, the vertical eddy viscosity, Km = a0CDuth0, where ut is a representative tidal
363
velocity, CD = 2 × 10−3 is a bottom drag coefficient, and a0 is a numerical constant, 0.028.
364
We use salinities from the BLM and SSNERR logger locations to estimate a time-
365
series of along-estuary salinity gradients. Because the loggers are positioned at a fixed
366
depth, the salinity gradient was not depth-averaged, and represented only the bottom
367
salinity gradient. However, our monthly CTD transects showed that bottom salinity
368
gradients were similar in magnitude to surface and depth-averaged salinity gradients in
369
all but a few cases.
14
370
A low-pass filter of successive 24 hr, 24 hr, and 25 hr boxcar filters (Godin 1991)
371
was applied to remove tidal variability from the salinity data over the time period 1-Oct-
372
2011 to 31-Sept-2013. To estimate ∂S/∂x, we took the difference in the low-passed,
373
tidally-averaged salinities (∂S) and divided by ∂x = 9 km.
374
Just as the T-S data indicated, the magnitude of residual flows was greater in 2012
375
than in 2013, with the mean magnitudes equal to 0.05 m s-1 in 2013 and 0.08 m s-1 in
376
2012. Although data gaps prevented a full picture of the continuous two year time series
377
of Ue, for the record available, it is clear there were several instances where the exchange
378
flow exceeds 0.15 m s-1 in 2012, whereas, in 2013, the exchange flow is almost always
379
smaller in magnitude (supplemental figure, S1). The theoretical scales for Ue compare
380
favorably with the magnitudes observed by the ADCP, although we cannot compare the
381
methods directly as the 2014 water year logger data are unavailable at present.
382
From the rotated along-estuary, low-passed ADCP velocities, we estimated transit
383
and filling times, two measures that often are used as proxies for residence time. The
384
transit time quantifies the time a particle takes to exit the estuary from a particular point
385
of input (Vallino and Hopkinson, 1998; Sheldon and Alber, 2002). The filling time is the
386
time it would take to fill the volume of the estuary, given a specified volume transport
387
(Lemagie and Lerczak 2014; Sutherland et al., 2011).
388
For transit times, at each time step, we calculated the mean velocities for the
389
upper and lower layers. The coordinate system was aligned such that Ue < 0 indicates
390
movement up-estuary, Ue > 0, down-estuary. A transit time from the mouth to the head of
391
the estuary was calculated by dividing the distance (21 km) by the lower layer velocity. A
392
total transit time was then found by adding the up-estuary transit time to the down-
393
estuary transit time.
394
For the ADCP time series from late November to late May, transit times were on
395
the order of two weeks. The median transit time was 14.0 days, the mode was 5.2 days,
396
and the mean and standard deviation were 25.5±103.2 days. The breakdown in this
397
approximation occurs when the whole water column was moving either up or down
398
estuary, which is caused by a breakdown in two-layer estuarine flow due to strong local
399
winds ramping up over the long north-south fetch (Fig. 6).
15
400
For filling times, we used a volume determined from an updated bathymetric
401
dataset of Coos Bay, which includes a NOAA digital elevation model and channel
402
bathymetry from the USACE. Zero crossings in the ADCP along-estuary current profiles
403
were identified for every time step. Using those zero crossings, cross sectional areas for
404
the bottom and top layers were calculated. Transports were then found by multiplying the
405
cross sectional areas by the mean velocities in the upper and lower layers. The filling
406
time was calculated using the lower layer transport. The median filling time was 22.7
407
days, with a mode of 6.1 days, and mean and standard deviation of 63.9±522.9 days. Like
408
the transit times, this approximation broke down during periods where the along-estuary
409
components of the velocities were going in the same direction at all depths. The filling
410
times were longer than the transit times, but followed the same pattern.
411
Although several assumptions underlie these approximations, we believe they
412
provide an order of magnitude estimate of residence times that are useful to estuarine
413
managers and the discussion of dissolved oxygen variability. However, to accurately
414
portray the spatial and temporal variability in estuarine residence times across Coos Bay,
415
we would have to show: (i) the assumption that cross-estuary advection is negligible, (ii)
416
that the influence of tidal motions on residence times are small, and (iii) the along-estuary
417
ADCP velocities were spatially constant (Lemagie and Lerczak 2014).
418
419
4.3. Dissolved Oxygen Variability
420
4.3.1. Seasonal Variability in DO
421
Dissolved oxygen levels were lower overall in the dry season than in the wet
422
season (Fig. 7), a result that indicates either 1) water properties in Coos Bay were
423
strongly coupled to shelf processes where a similar temporal signal is observed, or 2)
424
biological processes drew down ambient DO levels during summer. The coupling
425
between shelf waters and the estuarine waters has been previously documented for Coos
426
Bay (Roegner and Shanks 2001; Czielsa 1999), and other, similar small Pacific
427
Northwest estuaries (Roegner et al. 2002; Brown and Ozretich 2009; Brown and Power
428
2011). However, the higher incidence of lower DO levels late in the dry season coincides
429
with longer residence times as the dry season progresses, potentially allowing biological
430
processes to decrease DO. Quite likely, both processes are important; though quantifying
16
431
the biological component of this DO drawdown is beyond the scope of this study. Fig. 7
432
shows that in general, the lowest DO waters observed in the CTD transects along the
433
main channel coincide with the warmest waters, for example in September (Fig. 7b). The
434
relatively low DO waters generally are found at higher salinities, although the minimum
435
DO levels are not found at maximum S (Fig. 7c). Instead, there appears to be a DO
436
minima at intermediate salinities, both in the CTD data and the DEQ data.
437
438
439
440
441
442
443
Figure 7. (a) Minimum dissolved oxygen observed during each monthly transect versus distance
along the estuary (x = 0 km is at the mouth), colored by month. (b) Dissolved oxygen versus
temperature observations from all monthly transect data (colors same as in a) and the OR DEQ
historic dataset (pink dots). (c) Same as in b but for dissolved oxygen versus salinity.
The exception to the DO-T relationship occurs in April of both years (Fig. 7),
444
where low DO waters are observed near the mouth, with a strong along-estuary gradient.
445
The minimum in DO in these April transects coincides with colder, saltier water, in
446
contrast to the other transects. This observation, which we hypothesize as due to the
17
447
presence of upwelled water at the mouth of Coos Bay, is discussed further below in
448
combination with time series data from the other sensors.
449
To give historical context to the observed occurrence of low DO in Coos Bay, we
450
examined the DEQ dataset that spans 1957-2007 and covers a large portion of the estuary
451
(Fig. S2, inset). Limiting the DEQ data to only the dry season months, we find only one
452
measurement of DO <2 mg L-1 (Fig. 7), with the majority of all observations >5 mg L-1.
453
The low DO measurement appears cold and salty relative to all the other data, suggesting
454
it is derived from upwelled waters, much like the April CTD data (Fig. 7). The general
455
trend in DEQ DO levels shows that the lowest DO levels are found at the warmest
456
temperatures and at S ~24-28. These T-S properties only occur in the riverine end of the
457
estuary and in the summer months, when surface heat fluxes and relatively warm river
458
water contribute to the high temperatures and relatively lower salinities (compared to
459
oceanic values where S>30). Unfortunately, while the DEQ data are temporally robust,
460
they lack any information about sample depths, making it difficult to draw conclusions
461
about the historic occurrence of hypoxia in the estuary with these data alone. In addition,
462
sampling was widely inconsistent between years (Fig. S2); some years had less than a
463
dozen samples, while others had hundreds. Nonetheless, combined with the more recent
464
water quality data discussed above, these data suggest that pervasive hypoxia has not
465
occurred in the main channel of Coos Bay over the last 50 years.
466
467
468
469
Table 1. Percentage of total number of DO measurements <6.5 mg L-1 for each logger
location. Dates for BLM and EMP span Oct-2011 to Sept-2013, while the SSNERR (SS)
Charleston logger spans Apr-2002 to Dec-2013.
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
SS
0
0
0
2
2
4
6
10
10
9
1
1
EMP
1
0
0
0
<1
3
7
4
38
0
0
0
BLM
0
0
0
2
<1
2
1
3
6
0
0
0
470
471
The two-year data record from the CTCLUSI and the eleven-year record from the
472
SSNERR show that although near bottom DO levels diminished as the dry season
473
progressed, the majority of the time (~90% in the dry season), the waters are not hypoxic
474
(Table 1). September had the maximum occurrence of low DO levels at all locations,
475
although the Empire logger was significantly higher than the other two. It is likely lower
18
476
than the BLM logger because of its closer proximity to the mouth. Thus it might receive a
477
greater proportion of low DO upwelled water (Fig. 7).
478
479
4.3.2. 2012/2013 Comparison
480
The seasonal occurrence of lower DO waters in Coos Bay is due to a combination
481
of physical (advection and residence time) and biologic (in situ respiration; Rumrill,
482
2006) processes. However, similar to the observed hydrographic variability shown here
483
for 2012 and 2013, and the interannual variability observed previously in shelf DO (Chan
484
et al., 2008; Grantham et al., 2004), we find interannual changes in estuarine DO levels in
485
Coos Bay.
486
487
488
489
490
491
492
493
494
Figure 8. Dissolved oxygen (a) and salinity (b) measurements from Oct-2011 to Sept-2013 from
the BLM, EMP, and SSNERR sensor locations. The yellow, orange, and red horizontal dashed
lines various cited thresholds (see text). The black line seperates the 2012 and 2013 water years,
while the green box indicates the Apr-2013 time period zoomed in on in panels c-f. (c)
Alongshore wind stress, τalong, from buoy 46015. τalong < 0 is upwelling favorable. (d) Low-pass
filtered salinity from the BLM, EMP, and SS locations. Light grey lines in the back show
unfiltered data with tidal variability. (e) Same as is d, but for T. (f) Same as is d, but for DO.
495
threshold found to maintain 90% of all species (Vaquer and Duarte 2008) and none below the 2
496
mg L-1 threshold (Fig. 8). Late in the dry season of 2013, there were more measurements
In 2012, waters were well oxygenated, with few values below the 4.6 mg L-1
19
497
of waters below 4.6 mg L-1, but still none below the 2 mg L-1 threshold. Despite a
498
stronger, longer, and more consistent upwelling season in 2012 (Fig. 3) that incorporated
499
cold, salty shelf waters into the estuary (Fig. 7), there were fewer instances of lower DO.
500
Thus, it is likely that the DO values <4.6 mg L-1 can be attributed to the diminished 2013
501
residual flow facilitating longer residence times and allowing in situ biologic DO
502
drawdown. Additionally, the estuarine water temperatures in the summer of 2013 were
503
higher than in 2012 (Fig. 3), facilitating increased biologic oxygen demand and decreased
504
solubility of oxygen across the air-water interface.
505
506
4.3.3. Synoptic Variability in DO
507
The advection of shelf waters into Coos Bay can cause DO variability on synoptic
508
timescales. Fig. 8 shows an example from April 2013 of how upwelling favorable winds
509
on the shelf allowed an inflow of relatively cold and salty water into the estuary. The
510
onset of the wind event on April 22nd, and the peak wind stress on the 23rd resulted in
511
minimum DO and temperatures, and maximum salinity on the 27th. During this period,
512
DO dropped by 2 mg L-1 at both the South Slough and BLM stations, with a slight lag at
513
the BLM station (Fig. 8). Unfortunately, the EMP DO sensor was malfunctioning during
514
this event, but given the gradients observed above (Fig. 7; Table 1) we would expect even
515
lower DO levels at EMP. These time series capture the sequence of the movement of
516
upwelled water into Coos Bay, likely explaining the strong along-estuary gradient in DO
517
observed in the CTD transect in Apr-2013 (Fig 7). With these large spatial DO gradients,
518
tidal currents can induce swings of 3-4 mg L-1 in DO levels at a single location (Fig. 8f).
519
520
5. Discussion
521
5.1. Hydrographic variability and estuarine dynamics
522
Physical processes set the ambient DO conditions in the estuary, while biological
523
processes may modulate DO within the estuary. Rapid decreases of DO in Coos Bay can
524
be linked to upwelling-favorable wind events on the shelf during early summer (Fig. 8).
525
And the inundation of the estuary with predominantly salty, oceanic water through the
526
dry season (Figs 2, 5) confirmed that shelf waters are key in setting the conditions for the
527
estuary, especially when freshwater flow is virtually nonexistent. Coos Bay is a seasonal
20
528
estuary, transitioning from a strongly stratified, salt-wedge type regime in winter during
529
storms, to a well-mixed, oceanic-influenced type in the summer.
530
We further corroborate the observational classification of Coos Bay from the
531
CTD transects by quantifying various estuarine non-dimensional parameters (Table 2).
532
Simpson numbers, Si, also referred to as horizontal Richardson numbers, show the ratio
533
of tidal mixing strength to the estuarine circulation strength (Stacey et al., 2001). Here, Si
534
= (gβh02∂S/∂x) / (CDut2), where ∂S/∂x is the depth-averaged, along-estuary salinity
535
gradient, and other variables are defined above. The CTD data provided ∂S/∂x, and tidal
536
velocities and water depth were approximated at 1 m s-1 and 10 m, respectively. For
537
months where more than one CTD transect was collected, Si represents the mean Si of
538
those months.
539
540
541
542
543
Table 2. Seasonal estimates of various estuarine parameters for Coos Bay. Si is the Simpson, or
horizontal Richardson number. Rie is the estuarine Richardson number, Frf is the freshwater
Froude number, and M is the mixing parameter (Geyer and MacCready 2014). Dashes indicate
data were not available to estimate Si for those months. See text for definitions.
Jan
Feb Mar Apr May Jun
Jul
Aug Sep
Oct Nov Dec
Si
0.36 0.43 0.53 0.46 0.61 0.36 0.30
-0.09 0.25 0.36
-Rie
1.03 0.63 0.98 0.57 0.23 0.16 0.04 0.02 0.04 0.06 0.52 1.05
Frf
0.40 0.24 0.38 0.22 0.09 0.06 0.02 0.01 0.02 0.02 0.20 0.40
M
0.5—1.6
544
545
Higher Si indicates stronger stratification that inhibits mixing, and the resultant
546
increased current shear facilitates stronger baroclinic flow (Stacey et al., 2001). Low
547
values of Si, which occurred more often in the summer in Coos Bay (Table 2), indicate
548
that tidal mixing dominates. This supports the well-mixed conditions observed in
549
summer, when the baroclinic forcing is weak and tidal mixing occurs more easily. The Si
550
value for September was the lowest, falling under the threshold where well-mixed
551
conditions dominate (Si < 0.1; (MacCready & Geyer, 2010; Simpson et al. 1990). Higher
552
values of Si in the winter and spring months indicated the potential for stronger baroclinic
553
flow (Table 2).
554
The estuarine Richardson number, Rie = gβh0S0uf / (ud2), where uf = Qr / A is
555
freshwater discharge rate per unit cross-sectional area of the estuary, and the densimetric
556
velocity ud = (gβh0Smouth)1/2 is a ratio of the energy input by river flow to the work done
21
557
by bottom stress (Geyer & Ralston, 2011). Large Richardson numbers, Rie > 0.8, imply a
558
stratified water column, while smaller values, Rie < 0.8, imply a well-mixed system
559
(Fischer, 1976). For Coos Bay, Rie numbers followed our observations. In months of high
560
river flow, Rie > 0.8, and the system was stratified (Fig. 2). As the dry season progressed,
561
Rie decreased, and the system became more well-mixed, with Rie < 0.8 (Fig. 2).
562
A more recent approach to estuarine classification uses the familiar freshwater
563
Froude number, Frf = uf / (gβh0Smouth)1/2, in addition to a mixing parameter that provides
564
insight to variability on tidal and seasonal timescales (Geyer and MacCready, 2014). The
565
mixing parameter, M = (CDut2 / ωN0h02)1/2, is a ratio of tidal stirring to stratification,
566
where N0 = (gβSmouth / h0)1/2 is a buoyancy frequency, and ω is the tidal frequency.
567
For Frf, we used scaled discharge data from the CWA to calculate freshwater flow
568
velocities. A mean discharge for each month from the thirteen-year dataset was used to
569
find a representative uf. Frf for Coos Bay was on the order of ~10-3 to ~10-2 for months
570
that experience low to moderate discharge. Winter and spring months had Frf on the
571
order ~10-1. The mixing parameter, M, does not vary significantly seasonally, instead
572
most of the variability in M is over a tidal cycle and between spring/neap cycles. For
573
Coos Bay, M ~ 0.5-1.6.
574
Put into context of the Frf—M parameter space (see Fig. 6 in Geyer and
575
MacCready, 2014) Coos Bay spans several classification regimes. During low to medium
576
discharge, when Frf vales are smaller, and the tidal velocities are strong (causing higher
577
M), Coos Bay is in the strain-induced periodic stratification regime. Otherwise, during
578
weaker tidal flows, Coos Bay falls in the partially mixed regime. If discharge is high,
579
then Frf values are high, and Coos Bay is in the salt wedge classification. Using this
580
classification scheme, Coos Bay notably never was in the well-mixed regime—a result
581
that disagreed with our dry season CTD observations and calculated Rie values.
582
In addition to seasonal and interannual variability in the hydrography and
583
estuarine circulation in Coos Bay, the ADCP data revealed unexpected disruptions to the
584
two-layer estuarine exchange flow on synoptic timescales. At times, the ADCP showed
585
along-estuary currents flowing out of the estuary at all depths (2-Dec-2013, 2-Jan-2014,
586
11-Jan-2014). These events corresponded to strong local southward winds occurring
587
during spring tides (Fig. 6). The reverse happened when strong local northward winds
22
588
blow over the estuary during neap tides (11-Mar-2014). During these events, the along-
589
estuary currents flowed up-estuary at all depths.
590
The effect of wind forcing on estuaries has been studied extensively, and its
591
influence can be predicted using the Wedderburn number, W = N0 = τxL / βΔSgH12,
592
where τx is the along-estuary wind stress (N m-2), L is the length of the estuary, and ΔS is
593
the horizontal salinity difference averaged over the upper layer depth H1. W compares the
594
energy input directly by winds to an estuary to the potential energy available for driving
595
the baroclinic exchange flow (Chen and Sanford 2009; Geyer 1997). If W = 1, then the
596
wind input and baroclinic forcing are comparable and the estuarine circulation will be
597
significantly wind-influenced.
598
For Coos Bay, the along-estuary salinity gradient, estimated in W by ΔS/L, ranges
599
from 0.1-1.0 psu km-1 (Fig. 4), H1 is order 5 m (Fig. 2), and along-estuary winds can be
600
0.05 N m-2 in both directions (Fig. 3). Thus, minimum W occurs during light winds or
601
large along-estuary salinity gradients. Large W occurs during well-mixed conditions or
602
strong winds. The reversals observed in the ADCP data suggest that wind forcing can
603
overcome even strong buoyancy forcing in Coos Bay. Indeed, for ΔS/L = 10 / 10 km, W =
604
20, when τx = 0.05 N m-2.
605
606
5.2. Dissolved oxygen variability and link to estuarine dynamics
607
The data in this analysis show that hypoxia is not observed in the main channel of
608
Coos Bay, and has not been for the past decade and possibly longer. DO conditions prior
609
to 1957 need to be assessed further, possibly through the use of paleo-oxygenation
610
proxies (Gooday et al. 2009).
611
Lower DO waters were observed more frequently as the dry season progressed
612
(Table 1; Fig. 7). The location of the minima in DO levels along the estuary also migrated
613
from the mouth in the spring to the head of the estuary in the summer (Fig. 7). This
614
spatiotemporal shift is due to upwelled shelf water spending longer in the estuary.
615
Increased residence times in Coos Bay in the dry season were facilitated by weakened
616
residual flow resulting from diminished discharge lowering the buoyancy forcing to the
617
estuary. Decreased estuarine exchange flow in late-summer inhibits the advection of low
618
DO shelf waters into the estuary; upwelled waters are observed early in the summer after
23
619
the spring transition, but not in late summer. Thus, the low DO observed in Coos Bay in
620
late summer must occur due to local processes—when waters spend more time in the
621
estuary, they are subject to increased biologic respiration that draws down DO levels.
622
Further research on the spatiotemporal variability in net ecosystem metabolism in Coos
623
Bay would serve to corroborate this finding.
624
The absence of hypoxic waters in the main channel of Coos Bay is in part due to
625
the well-mixed conditions in summer (Fig. 2). Unlike other Pacific Northwest estuaries
626
that experience intermittent hypoxia—the Columbia River (Roegner et al. 2011) which
627
has year-round moderate to high discharge causing strong stratification, and Hood Canal
628
(Newton et al. 2007), which is a deep fjord-like estuary—Coos Bay does not have a
629
strong enough summer freshwater inflow nor is it deep enough nor to cause stratification
630
that prevents mixing of aerated surface waters with deeper, DO-depleted waters. The
631
physical processes affecting hypoxia in estuaries that face eutrophication issues, such as
632
in the Gulf of Mexico (Rabalais et al. 2002) or even many of Denmark’s small estuarine
633
systems (Conley et al. 2012), are different than those found here. For example, strong
634
stratification often inhibits vertical mixing and enables DO levels to decrease further,
635
with re-oxygenation only occurring during strong wind events that mix the water column
636
(Scully 2013). The well-mixed conditions and weak exchange flow in Coos Bay in late
637
summer implies that, even if hypoxic shelf waters existed offshore, they could only enter
638
the estuary through tidal dispersion processes.
639
However, the waters observed entering Coos Bay from the shelf were not
640
hypoxic. Along-shelf variability in the occurrence of inner-shelf hypoxia could explain
641
this finding (Peterson et al. 2011; Send and Nam 2012). Regions in the CCS with narrow
642
continental shelves experience less hypoxia than areas with wider shelves (Peterson et al.
643
2011). Upwelling on a wide shelf occurs in a relatively deeper depth range than over a
644
shallow shelf, given the same wind-forcing and shelf bathymetry. The deeper depth range
645
limits the re-oxygenation, due to photosynthesis or wind and wave driven mixing, of the
646
naturally low-DO upwelled waters (Send and Nam 2012). On a narrow shelf, such as near
647
Coos Bay, upwelled water reaches shallower depths where water parcels would benefit
648
from oxygen introduced by photosynthetic organisms and wind-driven mixing processes.
24
649
While hypoxia is not yet a problem in Coos Bay, monitoring should continue as
650
the effects of climate change advance, as many eastern boundary current systems will see
651
conditions move to being more favorable for increased inner-shelf hypoxia (Chavez and
652
Messié 2009). Increased surface-ocean warming, projected under IPCC AR5 (Rhein et al.
653
2013), will enhance stratification and decrease mixing. Upwelling winds in the CCS are
654
also intensifying, causing a greater volume of water to be upwelled onto the shelf (Bakun
655
1990). The DO content of these waters is decreasing as the OMZ shoals in the CCS
656
(Bograd et al. 2008; Pierce et al. 2012).
657
In tandem with the coastal effects of climate change on hypoxia, alterations to
658
Coos Bay’s hydrography should be expected under future climate change. An analysis of
659
climate models from the Intergovernmental Panel on Climate Change (IPCC) Fourth
660
Assessment Report showed the PNW experiencing little change in mean precipitation in
661
the coming century, however, the winters are projected to be wetter and the summers,
662
drier (Mote and Salathé, 2010). Less precipitation later in the wet season and into the dry
663
season could alter the estuarine circulation by further reducing the along-estuary salinity
664
gradient, analogous to the differences discussed above between the 2012 (drier, lower
665
DO) and 2013 (wetter, higher DO) water years. This will cause longer residence times,
666
and presumably, higher susceptibility to biologically-driven hypoxia.
667
668
6. Conclusions
669
Coos Bay is a strongly seasonal system as evidenced by the T-S properties of
670
water in the main channel of the estuary. Most of the seasonality can be linked to variable
671
freshwater inflow to the estuary and coastal shelf processes. The seasonality of Coos Bay
672
results in large swings in its estuarine parameters and circulation, so much so that Coos
673
Bay can be classified into a seasonally-shifting pattern of estuarine regimes, from
674
strongly stratified during episodic wintertime storms to well-mixed during the
675
summertime dry months.
676
This seasonal variability has implications for dissolved oxygen levels and the
677
health of the estuary. While stratification is strong under high discharge conditions in the
678
winter, colder, well-oxygenated river waters, and downwelling conditions on the shelf
679
promote high DO levels. In the summer, discharge diminishes and residual flows
25
680
stagnate, causing long residence times in the estuary. Upwelled shelf waters in the
681
estuary are relatively low in DO, yet stronger upwelling over a season did not result in
682
lower DO levels overall in Coos Bay. Despite recent alarm about inner-shelf hypoxia on
683
the Oregon coast, there appears to be no hypoxia in the main channel of Coos Bay
684
currently, and there is little evidence of past hypoxia in the estuary.
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
Acknowledgments
This project was partially supported by Oregon Sea Grant (NA14OAR4170064) and the National
Science Foundation grant OCE-1259603. We thank Larry Draper at the Oregon Institute of
Marine Biology for help with vessel operations, the staff at the South Slough National Estuarine
Research Reserve for help in data collection and quality control, and Jon Souder and his crew at
the Coos Watershed Association for acquiring discharge data for the area. Margaret Corvi of the
Confederated Tribes of Coos, Lower Umpqua, and Siuslaw Indians kindly provided the logger
data. Larry Caton of the Oregon Department of Environmental Quality provided the historic data.
References
Adams, K. A., Barth, J. A., and F. Chan, F. 2013. Temporal variability of near‐bottom dissolved
oxygen during upwelling off central Oregon. J. Geophys. Res.-Oceans, 118(10), 4839-4854.
Bakun A. 1990. Global climate change and intensification of coastal ocean upwelling. Science
247:198-201.
Baptista, A.M. 1989. Salinity in Coos Bay, Oregon: review of historical data (1930-1989). Report
ESE-89- 001, U.S. Army Corps of Engineers, Portland, OR. 40 pp.
Barth, J.A., et al. 2007. Delayed upwelling alters nearshore coastal ocean ecosystems in the
northern California current. Proc. Nation. Acad. Sci., doi:10.1073/pnas.0700462104.
Bograd, S. J., Castro, C. G., Di Lorenzo, E., Palacios, D. M., Bailey, H., Gilly, W., and F.P.
Chavez. 2008. Oxygen declines and the shoaling of the hypoxic boundary in the California
Current. Geophys. Res. Lttrs., 35(12), L12607. doi:10.1029/2008GL034185.
Brown, C. A., and R.J. Ozretich. 2009. Coupling between the coastal ocean and Yaquina Bay,
Oregon: Importance of oceanic inputs relative to other nitrogen sources. Estuar. Coasts, 32(2),
219-237.
Brown, C. A, and J.H. Power. 2011. Historic and recent patterns of dissolved oxygen in the
Yaquina Estuary (Oregon, USA): Importance of anthropogenic activities and oceanic conditions.
Estuar., Coast., Shelf Sci., 92(3), 446–455. doi:10.1016/j.ecss.2011.01.018
Chan, F., Barth, J. A, Lubchenco, J., Kirincich, A, Weeks, H., Peterson, W. T., and B.A. Menge.
2008. Emergence of anoxia in the California current large marine ecosystem. Science, 319(5865),
920.
Chavez, F. P., and M. Messié, 2009. A comparison of eastern boundary upwelling ecosystems.
Progress in Oceanography, 83(1), 80-96.
26
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
Chen, S.-N. and L.P. Sanford. 2009. Axial wind effects on stratification and
longitudinal salt transport in an idealized partially mixed estuary. J. Phys. Oceanogr., 39,
doi:10.1175/2009JPO4016.1.
Conley, D.J., H. Kaas, F. Mohlenberg, B. Rasmussen, and J. Windolf, 2000. Characteristics of
Danish estuaries. Estuaries, 23(6) 820-837.
Connolly, T. P., Hickey, B. M., Geier, S. L., and W.P. Cochlan. 2010. Processes influencing
seasonal hypoxia in the northern California Current System. J. Geophys. Res.-Oceans (1978–
2012), 115(C3).
Cziesla, C. A. 1999. The transport and distribution of the toxic diatom Pseudo-nitzschia spp. in
the Coos Bay estuary and the adjacent continental shelf (Doctoral dissertation, Thesis (MS)-University of Oregon, 1998.).
Epifanio, C.E., D. Maurer, and A.I. Dittel, 1983. Seasonal changes in nutrients and dissolved
oxygen in the Gulf of Nicoya, a tropical estuary on the Pacific Coast of Central America.
Hydrobiologia, 101, 231-238.
Fischer, H. B. 1976. Mixing and dispersion in estuaries. Ann. Rev. Fluid Mech., 8 (1), 107-133.
Geyer, W.R. 1997. Influence of wind on dynamics and flushing of shallow estuaries.
Estuar. Coast. Mar. Sci., 44, 713—722.
Geyer, W. R., and D.K. Ralston. 2011. The Dynamics of Strongly Stratified Estuaries. Treatise
on Estuarine and Coastal Science (Vol. 2, pp. 37–52). Elsevier Inc. doi:10.1016/B978-0-12374711-2.00206-0.
Geyer, W. R., and MacCready, P. 2014. The estuarine circulation. Ann. Rev. Fluid Mech., 46(1),
175.
Godin, G., 1991: The analysis of tides and currents (review). Progress in Tidal Hydrodynamics.
B.B. Parker, Ed., John Wiley, 675 – 709.
Gooday, A.J. et al. 2009. Historical records of coastal eutrophication-induced hypoxia,
Biogeosci., 6, 1707-1745.
Grantham, B. A., Chan, F., Nielsen, K. J., Fox, D. S., Barth, J. A., Huyer, A., and B.A. Menge.
2004. Upwelling-driven nearshore hypoxia signals ecosystem and oceanographic changes in the
northeast Pacific. Nature, 429(6993), 749-754.
Hagy, J. D., Boynton, W. R., Keefe, C. W., & Wood, K. V. (2004). Hypoxia in Chesapeake Bay,
1950–2001: Long-term change in relation to nutrient loading and river flow. Estuaries, 27(4),
634-658.
Hansen, D.V., and M. Rattray. 1965. Gravitational circulation in straits and estuaries. J. Mar.
Res., 23, 104 – 122.
Hickey, B., and N.S. Banas. 2003. Oceanography of the U.S. Pacific Northwest Coastal Ocean
and estuaries with application to coastal ecology. Estuaries, 26(4), 1010–1031.
27
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
Huyer, A. 1983. Coastal upwelling in the California Current system. Prog. Oceanogr. 12:259–
284.
Hyde, N. 2005. Towards national estuarine modeling and characterization and classification
systems: a pilot study for Coos Bay. MSc Thesis, Oregon Health and Science University. 152 pp.
Lemagie, E. P., & Lerczak, J. A. (2014). A Comparison of Bulk Estuarine Turnover Timescales
to Particle Tracking Timescales Using a Model of the Yaquina Bay Estuary. Est. Coasts, 1-18.
Lellis-Dibble, K. A., K. E. McGlynn, and T. E. Bigford. 2008. Estuarine Fish and Shellfish
Species in U.S. Commercial and Recreational Fisheries: Economic Value as an Incentive to
Protect and Restore Estuarine Habitat. U.S. Dep. Commerce, NOAA Tech. Memo. NMFSF/SPO-90, 94 p.
MacCready, P., and W.R. Geyer. 2010. Advances in estuarine physics. Ann. Rev. Mar. Sci., 2, 3558.
Monismith, S.G., W. Kimmerer, J.R. Burau, and M.T. Stacey. 2002. Structure and flowinduced variability of the subtidal salinity field in Northern San Francisco Bay. J. Phys.
Oceanogr., 32, 3003-3019.
Monteiro, P. M. S., Van der Plas, A., Mohrholz, V., Mabille, E., Pascall, A., and W. Joubert.
2006. Variability of natural hypoxia and methane in a coastal upwelling system: Oceanic physics
or shelf biology? Geophys. Res. Lttrs., 33(16).
Mote, P. W., and E.P. Salathe. 2010. Future climate in the Pacific Northwest. Climatic
Change, 102(1-2), 29-50.
National Oceanic and Atmospheric Administration. 1998. NOAA's Estuarine Eutrophication
Survey; Volume 5: Pacific Coast Region. Office of Ocean Resources Conservation and
Assessment, Silver Springs, Maryland.
Newton, J., Bassin, C., Devol, A., Kawase, M., Ruef, W., Warner, M., and R. Rose. 2007.
Hypoxia in Hood Canal: An overview of status and contributing factors. In 2007 Georgia Basin
Puget Sound Res. Conference, Vancouver, British Columbia.
Peterson, J. O., Morgan, C. A., Peterson, W. T., and E. Di Lorenzo. 2013. Seasonal and
interannual variation in the extent of hypoxia in the northern California Current from 19982012. Limnol. Oceanogr, 58(6), 2279-2292.
Pierce, S. D., Barth, J. A., Shearman, R. K., and A.Y. Erofeev. 2012. Declining Oxygen in the
Northeast Pacific. J. Phys. Oceanogr., 42(3), 495-501.
Rabalais, N. N., Turner, R. E., & Wiseman Jr, W. J. (2002). Gulf of Mexico hypoxia, AKA" The
dead zone". Annual Review of ecology and Systematics, 235-263.
Rabalais, N. N., Díaz, R. J., Levin, L. A., Turner, R. E., Gilbert, D., and J. Zhang. 2010.
Dynamics and distribution of natural and human-caused hypoxia. Biogeosci., 7(2), 585–619.
doi:10.5194/bg-7-585-2010.
28
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
Ralston, D.K., W.R. Geyer, and J.A. Lerczak. 2008. Subtidal salinity and velocity in the Hudson
River estuary: Observations and modeling. J. Phys. Oceanogr., 38: 753-770.
Rhein, M., et al., 2013: Observations: Ocean. In: Climate Change 2013: The Physical Science
Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental
Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J.
Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA.
Roegner, G. C., and A.L. Shanks. 2001. Import of coastally-derived chlorophylla to South
Slough, Oregon. Estuaries, 24(2), 244-256.
Roegner, G. C., Hickey, B. M., Newton, J. A., Shanks, A. L., and D.A. Armstrong. 2002. Windinduced plume and bloom intrusions into Willapa Bay, Washington. Limnol. Oceanogr, 47(4),
1033-1042.
Roegner, G. C., Needoba, J. A., and A.M. Baptista. 2011. Coastal upwelling supplies oxygendepleted water to the Columbia River estuary. PloS One, 6(4), e18672.
doi:10.1371/journal.pone.0018672.
Rumrill, S. 2006. The Ecology of the South Slough Estuary  : Site Profile of the South Slough
National Estuarine Res. Reserve. NOAA / Oregon Department of State Lands, pp. 1-238.
Scully, M. E. 2013. Physical controls on hypoxia in Chesapeake Bay: A numerical modeling
study. J. Geophys. Res.-Oceans, 118(3), 1239-1256.
Send, U., and S. Nam. 2012. Relaxation from upwelling: The effect on dissolved oxygen on the
continental shelf. J. Geophys. Res.-Oceans (1978–2012), 117(C4).
Sheldon, J. E., and M. Alber. 2002. A comparison of residence time calculations using simple
compartment models of the Altamaha River Estuary, Georgia. Estuaries, 25(6), 1304-1317.
Simpson, J. H., Brown, J., Matthews, J., and G. Allen. 1990. Tidal straining, density currents, and
stirring in the control of estuarine stratification. Estuaries, 13 (2), 125-132.
Snyder, M. A., Sloan, L. C., Diffenbaugh, N. S., and J.L. Bell. 2003. Future climate change and
upwelling in the California Current. Geophys. Res. Lttrs., 30(15).
Stacey, M.T., Burau, J.R., and S.G. Monismith. 2001. Creation of residual flows in a partially
stratified estuary. J. Geophys. Res., 106, 17013–17037.
Sutherland, D. A., MacCready, P., Banas, N. S., and L.F. Smedstad. 2011. A Model Study of the
Salish Sea Estuarine Circulation. J. Phys. Oceanogr., 41 (6), 1125-1143.
Thomas, A. C., Strub, P. T., and P. Brickley. 2003. Anomalous satellite‐measured chlorophyll
concentrations in the northern California Current in 2001–2002. Geophys. Res. Lttrs., 30(15).
United States Department of Energy, Federal Energy Regulatory Commission. 2009. Final
Environmental Impact Statement: Jordan Cove Energy and Pacific Connector Gas Pipeline
Project (Volume I –Executive Summary, Sections 1, 2, 3 and 4.0 - 4.4). Washington, DC. Online
at http://www.jordancoveenergy.com/pdf/FEIS%20-%20Volume%201.pdf.
29
879
880
881
882
883
884
Vallino, J. J., and C.S. Hopkinson. 1998. Estimation of dispersion and characteristic mixing times
in Plum Island Sound estuary. Estuar., Coast. Shelf Sci., 46 (3), 333-350.
Vaquer-Sunyer, R., and C.M. Duarte. 2008. Thresholds of hypoxia for marine biodiversity.
Proceedings Nat. Acad. Sci., 105(40), 15452–7. doi:10.1073/pnas.0803833105.
30
885
886
887
888
889
890
Supplementary Information
Supplementary Table S1. Information on along-estuary transects, including date of
observations, CTD instrument used, average tidal stage during transect, number of casts, whether
dissolved oxygen was measured, and what part of the spring/neap cycle was occurring.
Number
Date
Instrument
Tidal stage
Casts
Oxygen
Tides
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
11-3-12
1-19-13
2-21-13
3-12-13
3-26-13
4-27-13
5-19-13
7-11-13
8-16-13
9-7-13
9-17-13
9-21-13
10-10-13
10-16-13
10-20-13
11-30-13
1-25-13
2-22-14
3-18-14
4-6-14
5-13-14
6-17-14
7-24-14
RBR XR-620
RBR XR-620
RBR XR-620
RBR XR-620
RBR XR-620
RBR XR-620
RBR XR-620
SBE 19plus
YSI CastAway
YSI CastAway
RBR XR-620
RBR XR-620
RBR XR-620
SBE 19plus
RBR XR-620
RBR Concerto
RBR Concerto
RBR XR-620
RBR XR-620
RBR XR-620
RBR Concerto
RBR Concerto
RBR Concerto
flood
flood
high slack
ebb
high slack
flood
ebb
flood
low slack
flood
flood
flood
flood
high slack
high slack
high slack
low slack
low slack
flood
low slack
flood
low slack
high slack
22
33
9
11
32
32
19
16
17
16
11
13
27
8
14
14
26
21
11
22
19
13
17
Y
Y
Y
Y
Y
Y
Y
N
N
N
Y
Y
Y
N
Y
Y
Y
Y
Y
Y
N
Y
Y
spring
neap
spring
spring
spring
spring
neap
spring
spring
spring
spring
spring
neap
spring
spring
spring
neap
neap
spring
neap
spring
spring
spring
891
892
893
31
894
895
896
897
898
899
Supplementary Figure S1. Magnitude of the estuarine exchange flow, Ue, calculated
using the tidally-averaged salinity gradient between the CTCLUSI BLM logger and the
SSNERR Charleston Bridge logger for 2012 (blue) and 2013 (red) water years.
900
901
902
903
904
905
Supplementary Figure S2. Dry season DO measurements within the main estuary (red
dots, inset map) of Coos Bay. Measurements from smaller sloughs and far from the main
channel (black dots, inset map) were excluded. The light gray bars show the total number
of measurements taken each year with the red portion of the bar representing DO levels
<6.5 mg L-1.
32