The evidence:

Hypoxia in Hood Canal, Washington: controlling factors and current statusJ. Newton1, M. Alford1, A. Devol2, M. Kawase2, M. Warner2, D. Hannafious3

See the HCDOP IAM website: www.hoodcanal.washington.edu for program information and data access

Hood Canal, WA

The factors:

The strategy:

32

28

13

10

159

4

1115

% increase in integrated

production

Newton et al., 2000

Nitrogen sensitivity:

20 Aug 1998

0

5

10

15

20

25

30

0 5000 10000

P.production (mgC m-3 d-1)

ambient

spike

<5 >5-15 >15-25 >25-35

Is it a change in nutrient input ?

Is it a change in river input ?

Sisters - 50 m

0

1

2

3

4

5

6

7

0 30 60 90 120 150 180 210 240 270 300 330 360

Day of Year

Dis

solv

ed

Oxyg

en

(m

g/l)

1950-59

1960-66

2003-4

2005

Is it a change in ocean input ?

Is it climate-driven, e.g., by winds ?

The HCDOP Integrated Assessment and Modeling (IAM) study, in its first year of three, is utilizing automated technology, observing networks, and numerical modeling to:

• Quantify marine processes • Quantify loadings to Hood Canal• Assess biota-oxygen interactions• Model key processes to quantify driver(s) and• Evaluate potential corrective actions

}

These figures show that in both the “main-stem” of the Hood Canal (see red outlines) and the “around the bend” hook (see green outlines), the seasonal oxygen amplitude is flattened in recent years, with the largest historical change in the early and late part of a year. This change could have a significant impact on biota that cannot withstand sustained hypoxia. Fish kills, although recorded sporadically throughout history, have occurred repeatedly during 2002, 2003, and 2004.

Seattle

Pacific Ocean

0

20

40

60

80

100

120

140

160

180

200

0 2 4 6 8 10 12

Calendar month

Mo

nth

ly m

ean

flo

w (

ft3

s-1)

1940's1950's1960's1970's1980's1990's2000's

See 600% higher flows during summer in 2000's than 1950-80’s

Skokomish River mean monthly flow at Potlatch station

USGS data

Kawase (UW) model output

Higherthan normalsalinity

Lower than normal salinity

It could be.Climate variation could be influencing oxygen concentrations from several mechanisms, including sunnier summers (2003 and 2004), changes in precipitation (drought in 2000-2001, very wet in 2005), temperature (though this alone cannot account for the variation in oxygen), and wind.

It could be.The flushing of Hood Canal is driven by a “push” from the incoming high-density Pacific Ocean water. We see that the density of the incoming ocean water was relatively light during 2003, whereas the water within Puget Sound and Hood Canal was relatively dense, primarily driven by a strong ‘densification’ during the 2000-2001 drought.

It could be.Phytoplankton production in Hood Canal is limited by nitrogen (N). Surface production can be enhanced as much as 300%. Regionally, Hood Canal is the area most sensitive to N.

It could be.The main river to Hood Canal, the Skokomish, is impounded for hydroelectric power generation.

-40

-30

-20

-10

0

10

20

30

40

-45 -35 -25 -15 -5 5 15 25 35 45

Upwelling

Riv

er F

low

97

01

94

99

03

96

95

93

98

02

00

strongest density gradient strong flushing inhibited v. mixing

high density waterlow oxygen content

low density waterhigh oxygen content

low

riv

er in

put

high

riv

er in

put

weakest density gradient weak flushing easier v. mixing

red = low DO anomalygreen = high DO

Low DO years when coastal upwelling weak

WEAK STRONGNewton, 2005 PS-GB RC

23.0

23.5

24.0

24.5

25.0

25.5

26.0

26.5

27.0

D-98 D-99 D-00 D-01 D-02 D-03 D-04

den

sit

y, kg

/m^

3

JEMS

22

22.2

22.4

22.6

22.8

23

23.2

23.4

23.6

23.8

Mar-97 Mar-98 Mar-99 Mar-00 Mar-01 Mar-02 Mar-03 Mar-04

50 m

Den

sity

at

50 m

2000 20041998 2002

Lower Admiralty Inlet Ecology-PSAMP

140

m

80 m

Den

sity

at

Strait Juan de FucaIncoming ocean water

Deep PS basin water

Stream-Riparian ForestComposition Change

0%

20%

40%

60%

80%

100%

1 1 1 3 3 4 5 7 10 15 20 26 40 40 41 46 50 54 55

Distance from Stream Edge (m)

Bas

al A

rea

Douglas-fir Cedar Hemlock Other Conifer Alder Bigleaf Maple Other Hardwood

0%

20%

40%

60%

80%

100%

c. 1870

2003

Ted Labbe

Port Gamble S’Klallam Tribe

Natural Resources Department

GLO survey notes

Re-occupation of GLO

As you can appreciate from the evidence below, the complexity of the factors driving hypoxia in Hood Canal is high. A study to model and assess controlling factors is needed if mitigation or corrective actions are to be evaluated and recommended.

The goal of the Hood Canal Dissolved Oxygen Program (HCDOP) is to:

• determine the sources of low dissolved oxygen in Hood Canal and the effect on marine life, and then

• work with local, state, federal, and tribal government policy makers to evaluate potential corrective actions that will restore and maintain a level of dissolved oxygen that will reduce stress on marine life. 

HCDOP is a partnership of 38 organizations that conducts monitoring and analysis and develops potential corrective actions to address the low dissolved oxygen problem in Hood Canal.

1 Applied Physics Lab, University of Washington; 2 School of Oceanography, University of Washington; 3 Hood Canal Salmon Enhancement Groupnewton@apl.washington.edu

Further evidence for a strong oceanic role in driving hypoxia is shown by this plot. The red diamonds show years of lower than average DO. River input does not appear to correlate with low DO, but weak coastal upwelling does. Weak upwelling means that the incoming ocean water is less dense, although it has more oxygen.

But what is the source of the N? Septic fields and agricultural runoff are easy target suspects, but other factors could be important. For example, a change in forests to alders that fix N may be a large factor. Quantification of the loads of N as well as C is needed.

While the dam dates back to 1926, the release of freshwater to Hood Canal has changed substantially over the decades, as population and power selling have both increased. Higher summertime flows began in the 1990s and 2000s, similar in timing to the advent of increased hypoxia. The added freshwater flow in summertime, good for fish passage, would lead to enhanced stratification and minimized vertical mixing and could enhance the occurrence of hypoxia.

Shown here is POM model output by Mitsuhiro Kawase (UW) under northerly and southerly winds. In color is the surface salinity anomaly. The “Great Bend” area of Hood Canal is where fish kills are often most severe. These model data show that with southerly winds, surface salinity is high, indicating that deep waters (with low oxygen) have upwelled and could potentially eliminate any refugia that biota could be using at the surface.

Responding to blooms and fish kills.

Measuring fluxes

Using real-time observing tools and posting data

Involving citizens in monitoring

Assessing biota effects

ABSTRACT: Hood Canal, a fjord-like sub-basin of Puget Sound, Washington State, USA, is a long (110 km), deep (100-200 m), narrow (1-2 km), productive (4000 mg C mg-2 d-1) estuary with strong seawater density stratification (Δ σt >2 all year) and slow circulation (months to year). These conditions are conducive to seasonally low oxygen concentrations, which have been observed in records dating back to the 1930's. However, in recent years, especially since the mid-1990's, the frequency, duration, and spatial extent of the hypoxia has increased. Although biota (fish) kills have been reported as far back as the 1920's, fish kill events occurred during 2002, 2003, and 2004. Of concern is that the dissolved oxygen inventories of the deep waters in the southern portion of the canal (where the hypoxia is strongest) measured during the 2000's are among the lowest, with 2004 the absolute lowest, on record compared to the data from the 1950's, 60's, and 90's. Also of concern is that the hypoxia is sustained through a longer portion of the year or, in some locations, all year long.

Causes for this severe and seemingly deteriorating condition could be many and potentially include: 1. changes in oceanic water properties that could affect flushing; 2. human-mediated loading of nitrogen or organics that could affect oxygen demand; 3. changes in river flow delivery that could affect stratification; and 4. changes in local weather forcing, that could have wide-ranging effects. We present evidence for some of these contributing factors.

A focused observational-modeling study, the Hood Canal Dissolved Oxygen Program’s Integrated Assessment and Modeling Study (HCDOP-IAM), will address the quantitative balance of these factors in driving the observed hypoxia using a suite of observational data and models, both watershed and coupled physical-biological marine models. This approach is necessary because different causes of the problem have different solutions. The HCDOP IAM strategy is possible through federal (U.S. Navy) funding by Congressman Norm Dicks, and significant leverage from NOAA IOOS/NANOOS, state, tribal, and volunteer commitments.