HomeMy WebLinkAbout2004 Lower Dosewallips River Reach AnalysisIntroduction
The SRFB application for the Dosewallips Estuary Restoration Project established three
fundamental problems in the ecological function of the lower Dosewallips River:
1. Loss /isolation of distributary channel habitat
2. Loss /modification of tidal estuarine circulation
3. Riparian forest degradation
4. Loss /impairment of natural channel processes in the mainstem river (This
problem has been identified post - application).
This reach analysis seeks to investigate current conditions and outline actions to remedy
these problems. An exception may be problem 3 which, in some part, is a biological not
physical problem.
Reach analysis or assessment is, in general, "An ... `objective procedure to characterize
the present [or historic] state of a natural resource in a watershed and to diagnose
resource impairment that can be remedied". In the context of stream habitat restoration,
assessment is typically used to document current and historic conditions, identify and
determine the extent of problems and habitat deficiencies, identify the causes of those
problems and deficiencies, identify restoration opportunities and constraints, and to
determine the dominant processes that create and maintain habitat within the stream
corridor2."
Physical habitat assessment can be approached in a variety of ways. A purely
quantitative analysis is a large undertaking associated with significant infrastructure and
12/22/2004 2 Dosewallips Reach Analysis
2
Introduction.......................................................................................... ...............................
3
Physical characteristics ........................................................................ ...............................
.........................
...... 3
Historical perspective ...................................................................
...............................
5
Channel characteristics ................................................................
5
Water, sediment and wood discharge .......................................... ...............................
6
Stream bed, bank and general soil characteristics ....................... ...............................
7
Channelprofile ............................................................................. ...............................
8
Channel stability and equilibrium ............................................... ...............................
12
CrossSections ............................................................................ ...............................
Crosssection 5 ............................................................................. .............................12
13
Crosssection 4 ........................................................................... ...............................
13
Crosssection 3 ........................................................................... ...............................
.............................13
Crosssection 2 .............................................................................
14
Crosssection 1 ........................................................................... ...............................
...............................
14
Distributaries..............................................................................
15
Habitat................................................................................................ ...............................
17
Conclusions........................................................................................ ...............................
18
Restoration Action Alternatives ......................................................... ...................
...... ......
19
Recommended Restoration Activities ................................................ ...............................
Introduction
The SRFB application for the Dosewallips Estuary Restoration Project established three
fundamental problems in the ecological function of the lower Dosewallips River:
1. Loss /isolation of distributary channel habitat
2. Loss /modification of tidal estuarine circulation
3. Riparian forest degradation
4. Loss /impairment of natural channel processes in the mainstem river (This
problem has been identified post - application).
This reach analysis seeks to investigate current conditions and outline actions to remedy
these problems. An exception may be problem 3 which, in some part, is a biological not
physical problem.
Reach analysis or assessment is, in general, "An ... `objective procedure to characterize
the present [or historic] state of a natural resource in a watershed and to diagnose
resource impairment that can be remedied". In the context of stream habitat restoration,
assessment is typically used to document current and historic conditions, identify and
determine the extent of problems and habitat deficiencies, identify the causes of those
problems and deficiencies, identify restoration opportunities and constraints, and to
determine the dominant processes that create and maintain habitat within the stream
corridor2."
Physical habitat assessment can be approached in a variety of ways. A purely
quantitative analysis is a large undertaking associated with significant infrastructure and
12/22/2004 2 Dosewallips Reach Analysis
high risk to human and habitat values. Accurate physical measurements and numerical
modeling are commonly the basis for such an analysis.
A qualitative approach uses expert opinion and readily obtained physical measurements
associated with rural, low risk projects. A qualitative approach is employed in this study.
The study area is the reach of the Dosewallips River mainstem downstream of the US
101 bridge as well as the adjacent flood plain and estuary.
Physical characteristics
Historical perspective
The Dosewallips estuary is delta -type estuary, as opposed to a drowned river valley -type
estuary also common in Puget Sound. Deltas form when the amount of sediment
produced by the watershed, and transported by the river, exceeds the volume of the valley
and sea level rise 3. The formation of the Dosewallips delta probably has its roots in the
distant, more geologically active past when glacial processes led to significant releases of
sediment to fill the glacially carved valley bottom.
Even though the delta is a depositional structure, the regional trend is one of subsidence.
A recent study of the Duckabush and Hamma Hamma River estuaries (3 and 12 miles
south of the Dosewallips, respectively) indicates that a combination of tectonic
subsidence and sea level rise has lead to a lowering of the estuary 2.7 to 4 mm/yr in the
last 100 years4. This is roughly twice the eustatic sea level rise in the last 1500 years.
Current morphology is dominated by episodic river floods, coarse sediment and a
macrotidal range, much lower energy processes than those that formed the delta. The
general shape of the delta, channel and distributaries has not changed substantially since
the 1883 (U.S. Coast & Geodetic Survey T- sheet), Figure 1. The remnant channel
running along the north edge of the delta clearly shows that the Dosewallips River once
ran there, but it is not clear when this occurred, certainly longer than 120 years.
Hydraulically, the remnant channel to the north is very attractive to the Dosewallips: it is
a roughly 30% shorter route to the base elevation at Hood Canal and, as a result, much
steeper with greater sediment transport capacity. The resistance of the delta to modern
change may have one of several roots. We did not seek an answer in this study because,
with the left bank dike, the development of Brinnon, and the US 101 road fill, it is
unlikely that this northern route will be occupied soon.
Development of the estuary had already begun when the U.S. Coast & Geodetic Survey
Topographic Sheet was drawn in 1883. This is true particularly in the area south of the
main channel where there was an orchard in that year and which, in all the subsequent
aerial photographs, is consistently cleared. The former orchard is now the site of State
Park's day use area and a number of camp sites. This same area shows the greatest
topographic simplification in our study area.
The currently wooded section, defined by the distributary we call N3 and the mainstem
left bank, was probably the location of the mainstem sometime before 1883. A large
12/22/2004
Dosewallips Reach Analysis
sediment pulse could have been deposited here, forcing the river to the right. This
section is rich with distributaries and flood swales. Analysis of aerial photographs shows
that in 1957 it was the only portion of the estuary east of US 101 that supported a dense
concentration of taller vegetation, and by 1965 is almost entirely treed.
Generally the area on the right bank, the side with the developed campgrounds, has been
simplified over the years. Nineteen -thirty-nine aerial photography shows nearly
featureless pasture or fields to the southwest of the main channel. Washington State Parks
archives have records of dredging and diking from the earliest days of Parks ownership.
A 1964 plan sheet shows dredging in the study area, primarily bar scalping. It is not
indicated where the dredge spoils were placed, although it is easy to imagine them placed
in the right bank area to increase its elevation.
In 1968 bank protection occurred along the right bank and further dredging in the main
channel with the note that "material to be used as fill behind bank protection." By 1978
the right bank was largely rip rapped, although it is difficult to tell exactly which
proposed construction activities were actually done and which weren't. The distributary
mouths, N1 and N2 (Figure 4), were cleared of debris at least once in 1978. It is
assumed that this was done to decrease flood elevations in the mainstem.
Focusing on the current alignment downstream of US 101, the main Dosewallips channel
runs down the center of a prograding delta cone, Figure 3. This morphology is a common
condition when the river alignment is constrained (in this case by the US 101 road fill and
bridge) and not allowed to move laterally across the face of the delta. The typical delta is
fan- shaped and the delta front moves out as sediment is added episodically across the
conical surfaces. In an unconstrained condition, the lower Dosewallips river would
abandon the current alignment for one that flowed to the north or south, around the
aggraded delta cone, seeking a more efficient route to salt water. Those familiar with
JimmyComeLately Creek will see that pattern.
JimmyComeLately Creek; avulsion
of a prograded delta cone
The Dosewallips Delta topography has been determined in a rough way using Lidar data
from the Port Gamble S'Klallam tribe and a limited amount of ground survey. We
12/22/2004 4 Dosewallips Reach Analysis
understand the general shape of the area below US 101, but not exactly (see Figures 2
and 3, and the cross sections).
Channel characteristics
The channel is confined at the bridge and unconfined at the estuary. The US 101 bridge
and approach fill limit the down -valley movement of flow over the flood plain. The
effective width of the bridge (compensating for the skew of 40° and eliminating the State
Parks access roadway) is 345 feet, the flood prone width. The average bankfull width
upstream of the bridge is approximately 230 feet, which makes the entrenchment ratio in
the bridge section 1.5, or moderately entrenched. Entrenchment indicates a channel with
high shear stress, high efficiency and a lack of flood plain storage. Downstream of the
bridge the flood prone width extends for an indeterminate distance since the flood plain is
a conical surface, falling away laterally. The meaning of this term is stretched in this
context, but we could say that it is unconfined.
Sinuosity in the study reach is essentially 1.0. At low channel slopes we expect high
sinuosity ( >1.4), although diked channels and those on alluvial fans tend to be straight.
This channel could move laterally at any point downstream of the bridge so that the entire
area should be considered the channel migration zone. Bank armoring on the right bank
hinders movement, as does a developed riparian on the left bank.
Water, sediment and wood discharge
While the USGS gage was in operation, 37 years of flow data were accumulated with the
following peak flow statistics (Table 1).
Table 1
USGS station 12053400 (1931 -1968)
Recurrence
Dosewallips R.
Interval, yr.
peak discharge, cfs
1.01
1,362
1.25
2,852
2
4,335
5
6,586
10
8,194
50
12,021
100
13,762
The bankfull channel in this reach can contain a 4000 cfs 2 year flood. Flows in excess
of this cover the flood plain to a greater or lesser extent. Distributary flow from the
bridge to roughly 2100 feet downstream may account for up to 2000 cfs. The 5 major
distributaries carry a maximum of about 400 cfs each. This may be overoptimistic since 3
of the 5 combine into one channel. These channels flow only within a few feet of
bankfull elevation, depending on the elevation of the sediment at the inlet. When flow
goes overbank, distributary flow likely increases as the width and depth of these channels
increases with distance from the mainstem.
12/22/2004 5 Dosewallips Reach Analysis
The elevation of the area north of the wooded riparian, the Bloomfield property, is
substantially lower than the bankfull elevation at the channel. The flood path between
the bank top at the bridge and bank top of Sylopash Slough drops 7 feet, and the lack of
heavy riparian vegetation along this path makes it very prone to avulsion.
The wood that enters the study reach is either pressed against the riparian vegetation by
overland flow and distributary flow, or it ends up the end of the channel, about MHHW,
at the head of the new distributaries. Very little of the wood is retained in the mainstem
channel and cannot be considered a significant factor in channel morphology or habitat
formation. Key piece size for a channel this size is minimum of one quarter of the
bankfull width long and one half to three quarters of the bankfull depth in diameterb. This
is equivalent to a log about 40 feet long and 2.5 to 3.75 feet average diameter, with root
wad. In his survey of in- channel wood in the lower half of the river, Labbe, et. al., found
the majority (99 %) was smaller than key piece size.
Stream bed, bank and general soil characteristics
Bank soils are generally sand and finer materials. The sort of soil to be expected on a
flood plain. In themselves, these soils are sensitive to lateral channel movement. Along
the main channel on both banks large cottonwoods, alders and conifers have become
established, as well as various species of native and non - native brush, to create a root
mass that is relatively resistant to erosion. In approximately 1968, Washington State
Parks installed rip rap bank protection all along the right bank from the US 101 bridge to
1800 ft downstream. This was done to stop the lateral movement of the lower
Dosewallips River into park property. This rock has now become embedded in roots and
is quite resistant to erosion. Visual inspection does not reveal the original extent of the
installation, although historical photographs show a well - armored bank.
The stream bed material consists of gravel and cobble, with the occasional boulder -sized
particle. Pebble counts were done at three locations in and upstream of the study reach.
One at a point 2644 feet downstream of the US 101 bridge. This location is the "end" of
the channel at about mean higher high water, where sediment transport stops as the
channel is backwatered at high tides. It was chosen to represent the sort of material
transported through the reach creating the emerging delta cone. Below is table of the
sediment distribution.
Table 2 Particle size distribution,
Dosewallips River mainstem, 2644 ft
downstream of US 101.
Percent finer
Particle
Particle
size, ft
size, mm
16
0.03
8
35
0.09
26
50
0.16
48
84
0.30
92
90
0.37
112
The second and third pebble counts were
100
0.60
183
done upstream of the US 101 bridge. These
12/22/2004
6
Dosewallips Reach Analysis
samples were taken on the face of bars representing the coarsest bedload transported by
the Dosewallips river in this reach.
Table 3 Particle size distribution,
Dosewallips River mainstem, 360 ft
upstream of US 101.
Percent finer Particle Particle
Table 4 Particle size distribution,
Dosewallips River mainstem, 1200 ft
upstream of US 101.
Percent finer
size, ft
size, mm
16
0.02
7
35
0.08
24
50
0.15
47
84
0.36
110
90
0.39
120
100
0.70
213
Table 4 Particle size distribution,
Dosewallips River mainstem, 1200 ft
upstream of US 101.
Percent finer
Particle
size, ft
Particle
size, mm
16
0.02
6
35
0.10
30
50
0.21
64
84
0.40
122
90
0.47
143
100
0.70
213
Channel profile
The Table 5 below shows the hydraulic properties of the cross sections along the main
channel profile downstream of US 101.
Table 5 Dosewallips River mainstem hydraulic conditions
Xsec
Distance
D/s US 101
ft.
Low flow
WS slope
ftfft
BFelev
Avg.
ft.
Depth
Avg.
ft.
Discharge
Bankfull
cfs
Width
Bankfull
ft.
Shear Stress
Bankfull
Ibstft "2
6
0
18.0
5.8
7174
207
1.0
5
278
0.38%
17.1
5.9
6157
185
0.9
4
817
0.14%
14.6
4.2
4248
175
0.8
3
1316
0.35%
13.1
4.4
3399
130
0.9
2
1753
0.29%
11.9
4.5
4159
150
0.9
1
2151
0.18%
8.8
2.1
112
0
2697
0.04%
8.8
2.3
The lower river is at an overall water surface slope of 0.19 %, over the entire reach from
the US 101 bridge to the terminus of the channel where it breaks off into multiple
distributaries. Low flow water surface slope varies along this profile, with a maximum of
0.38% downstream of the bridge, to a minimum of 0.04% in the last 500 feet.
At bankfull flow, water surface slope is approximately 0.3 %, the value used in bankfull
hydraulic computations for this study.
As one moves downstream, the hydraulic capacity of the main Dosewallips channel
decreases, indicated by the decrease in the average depth of the channel (in this instance
the difference in the elevation between the average bankfull elevation and the low flow
water surface), the bankfull width, and consequently the bankfull discharge. This
decrease in main channel capacity is compensated by an increase in the number of
distributaries, overflow channels, and general overbank flow.
12/22/2004 7 Dosewallips Reach Analysis
It is somewhat difficult to hydraulically model a channel which is "leaking" all along its
length. A one dimensional backwater model would require multiple water diversion
points along the profile to prevent the water surface elevation from "building up"
progressively downstream. For this study each cross section was treated individually to
determine the theoretical flow at bankfull stage. Flows in excess of the cross section's
capacity is assumed to flow overbank or down distributary channels. Hydraulic
computations were done with the aid of WinXsPro9, a normal flow cross section analysis
program.
The main channel type is pool riffle, although this is drowned out by sediment deposition,
with the channel tending to a plane bed through most of the reach. Low flow water depths
in the cross sections did not exceed 2.5 feet, except in the backwatered lower section.
This rather low pool depth for this size of channel indicates a lack of hydraulic diversity.
The lower 600 feet is backwatered at low flow by the terminus of the channel. At tides
greater than MHHW, the backwatered area extends further upstream.
Channel stability and equilibrium
As the main channel lengthens with delta progradation (6 ft/yr) and the average elevation
of Hood Canal increases with eustatic sea level rise (1 to 2 mm/yr.), channel slope
decreases (providing the upstream channel is generally in equilibrium). The consequence
of this decrease in slope is, among other things, reduced sediment transport capacity and
subsequent aggredation, leading to channel avulsion (abandonment of the current
mainstem). All else remaining equal, if the channel continues to prograde, a threshold
will be reached where an avulsion becomes imminent.
Since the 1960s the slope of the channel in the reach downstream of the bridge has
decreased. The 1968 USCOE Dosewallips River flood model plotted an "observed water
surface" in the reach roughly 1200 feet downstream of the US 101 bridge. Slope of this
observed surface was 0.43 %. In 2004 the low flow water surface slope was 0.28 %, a
decrease of 34 %. From this data it is assumed that this portion of the channel is aggrading
at a rate greater than the reach immediately upstream. The rate of this increase is greater
than that predicted by delta progradation and sea level rise and likely results from more
local aggredation in this reach.
12/22/2004 8 Dosewallips Reach Analysis
140-
120
100
E
E
ai 80
N
A
A2 60
M
o.
40
20
0
Particle size and shear stress along the Dosew alliips profile
0.98
0.96
0.94
0.92
0.9
0.88
0.86
0.84
0.82
0.8
0.78
500 1000 1500 2000 2500 3000 3500 4000 4500
River station, d/s of uppermost station, ft
The chart above pictures the change in particle sizes and shear stress along the
Dosewallips River profile starting from a point 1200 feet upstream of the US 101 bridge.
The particle size data is from Tables 2, 3 and 4, representing 2 pebble counts upstream
of the bridge and one at the terminus of the channel. Shear stress data is from Table 5.
The size of D50 is virtually unchanged through the study reach, indicating that this size
particle is readily transported under the shear stress present at all cross sections. The size
of D84, on the other hand, decreases, indicating a deposition of the coarser fraction within
the reach. The dip in shear stress at cross section. 4 means that this is a likely location for
deposition. As we have seen, there has been a reduction in slope since the 1960s in this
reach. The reach containing cross section 4 will tend to aggrade forcing high backwater
elevations which lead to more overbank flow and an increase in the likelihood of channel
avulsion.
While we can never know exactly when or where an avulsion will occur, the threat needs
to be evaluated. Figure 6 shows a number of possible avulsion routes. They begin at
points in the main channel of known elevation and proceed out perpendicular to contour
lines and roughly along existing distributary alignments. They are not all likely routes,
but offer a basis for comparison. Lateral channel changes are often stopped by wood
racking against the inlet when it is in a forested riparian, blocked by a deposit of sediment
in the inlet, or by trees falling into the newly enlarged channel as it passes through a
forested riparian. Routes A2 -5 are not likely avulsion paths for these reasons. Although
12/22/2004 9 Dosewallips Reach Analysis
6
N
N
a
N
N
N
N
m
a�
x
US 101 �a
+M
1 '0
y r
r . D50
D84
• Shear stress
Xsec. 4
0.98
0.96
0.94
0.92
0.9
0.88
0.86
0.84
0.82
0.8
0.78
500 1000 1500 2000 2500 3000 3500 4000 4500
River station, d/s of uppermost station, ft
The chart above pictures the change in particle sizes and shear stress along the
Dosewallips River profile starting from a point 1200 feet upstream of the US 101 bridge.
The particle size data is from Tables 2, 3 and 4, representing 2 pebble counts upstream
of the bridge and one at the terminus of the channel. Shear stress data is from Table 5.
The size of D50 is virtually unchanged through the study reach, indicating that this size
particle is readily transported under the shear stress present at all cross sections. The size
of D84, on the other hand, decreases, indicating a deposition of the coarser fraction within
the reach. The dip in shear stress at cross section. 4 means that this is a likely location for
deposition. As we have seen, there has been a reduction in slope since the 1960s in this
reach. The reach containing cross section 4 will tend to aggrade forcing high backwater
elevations which lead to more overbank flow and an increase in the likelihood of channel
avulsion.
While we can never know exactly when or where an avulsion will occur, the threat needs
to be evaluated. Figure 6 shows a number of possible avulsion routes. They begin at
points in the main channel of known elevation and proceed out perpendicular to contour
lines and roughly along existing distributary alignments. They are not all likely routes,
but offer a basis for comparison. Lateral channel changes are often stopped by wood
racking against the inlet when it is in a forested riparian, blocked by a deposit of sediment
in the inlet, or by trees falling into the newly enlarged channel as it passes through a
forested riparian. Routes A2 -5 are not likely avulsion paths for these reasons. Although
12/22/2004 9 Dosewallips Reach Analysis
6
N
N
a
N
N
N
N
m
a�
an avulsion could be forced by removing trees, if other factors are favorable. Routes Al
and A6 are not forested and do not surfer these controls.
Another factor in the occurrence of avulsions is the relative stream energy after the event
occurs. If the stream energy is in a lower state after the change occurs, it is more likely to
happen. This can be thought of a lowering of the water surface at the upstream end of the
avulsion route. What follows is an analysis to determine the change in water surface
elevation after an avulsion.
The slope of the avulsion routes was taken from the existing channel, taking into
consideration the concave -up shape to the profile of a channel approaching a level pool.
The point was made earlier that the slope of the existing channel is measurably lower
than it was just a few decades ago and it would seem prudent to assume a greater slope.
The main difference between then and now is that the current channel is much longer.
This increased length accommodates a low gradient reach directly above tidal influence,
in addition to a transition to the higher gradient and coarser channel just above it.
Hood Canal establishes a base elevation for any stream entering it. This simple
observation is complicated by the fact that it is macrotidal and that the base elevation will
be variable depending on the stream energy entering it and the size and abundance of
sediment transported into the interface. Rather than trying to solve this rather complicated
problem, we can use existing conditions at the mouth and make the assumption that base
elevation of any high order, sediment - carrying, channel will be similar to what we find at
the end of the existing channel. The bed elevation at the mouth was measured at 6.5 feet.
This is somewhat higher than the elevation of the tidal channels and distributaries
(approximately 5 feet, MSL is 4.1 feet) at the same relative location on the delta front.
These channels do not carry a coarse sediment load that would be deposited at a much
20
18
16
= 14
C
12
w 10
Avulsion Route Al b Profile
Approximate ground surface
Avulsion route Al
Sylopash slough
500 1000 1500 2000
Distance, ft.
higher elevation than the generally fine material present in tidal channels.
The profile for Al b is shown above. It assumes a base elevation of 6.5 feet and a slope
of 0.22 %. It does not follow the elevation of the Sylopash slough bottom since the river
flow that would establish this hypothetical channel would be carrying sediment that
12/22/2004 10 Dosewallips Reach Analysis
would be deposited under the influence of backwater from Hood Canal. The upstream
end has a final elevation approximately 1.1 feet below the current channel elevation.
This is the only profile, of the ones shown in Figure 6, that offers any decrease in relative
stream energy and this only occurs if the shore bar is breached and the main channel
enters the Canal toward the north east, rather than following Sylopash out to the east.
The Table 6, below, shows the average slope of the existing mainstem river beginning at
the delta front. The assumption is that given the sediment gradation of the current
bedload, similar flood events and the effect of tidal backwater, these are the slopes the
bed would normally assume. The slope increases as one moves upstream. These are
average slope for the lengths given, which are applied to similar length channels in the
hypothetical avulsion routes.
Table 6
Cross Distance from Average
section
Canal
slope
0
0
Avulsion
1
546
0.04%
2
944
0.11%
3
1381
0.17%
4
1880
0.21%
5
2419
0.22%
6
2697
0.23%
The Table 7 below shows the results of the analysis that was described above. For each
avulsion route with given length and slope (from Table 6), the final elevation at the
existing main channel is determined. As can be seen, in only one case will river energy
be decreased, Al b. In all other cases the routes are just too long to result in any
substantial decrease in stream energy, and are therefore less like to occur.
Table 7
Avulsion
Length
Slope'
ExistingZ
3
Avulsion
Route
Chan. Elev
Chan. Elev
Al
3000
0.23%
12.1
13.4
Al a
2510
0.22%
12.1
12.1
Al b
2080
0.22%
12.1
11.0
A2
2300
0.22%
10.8
11.5
A2 a
1900
0.21%
10.8
10.6
A3
1800
0.21%
10.1
10.2
A3 a
1240
0.15%
10.1
8.3
A4
1200
0.14%
8.8
8.2
A5
880
0.10%
7.4
7.4
A6
2700
0.23%
12.1
12.7
Average slope of channel with similar length. See text.
2 Elevation of
the existing channel at the beginning of the avulsion route.
3 Elevation of the beginning
of the avulsion channel when projected up
from elev. 6.5 feet at the slope given slope'.
In the early stages of the reach analysis, it appeared that an avulsion was a likely event in
the lower Dosewallips. It now appears that conditions have not reached the threshold
12/22/2004 11 Dosewallips Reach Analysis
past which an avulsion is imminent. Conditions can change rapidly and this threshold
exceeded rapidly. Such a condition would be a significant flood accompanied by a large
release of sediment and wood. Such an event could increase the elevation of the bed
downstream of US 101, deposit one or more significant log jams, or extend the length of
the current channel. If this were to happen, an avulsion would likely occur.
Cross Sections
Six cross sections were established in the study area, Figure 2. The cross sections are a
combination of ground surveys and Lidar aerial surveys. In areas where vegetation does
not obscure the ground surface, Lidar data was used without correction. In high marsh
areas, Lidar data was corrected with spot ground elevations. In areas of higher
vegetation, brush and trees, a ground survey was done. These two methods were merged
together to give the estuary-wide cross sections. The distance measurement begins at the
intersection with US 101, on the southern end of the cross sections. This results in plots
of these cross sections looking upstream, opposite the more conventional downstream-
looking cross section. This means that channel right is on the left hand side of the plots.
Unfortunately this was built into the analysis from the beginning and has not been
changed.
In a number of cases where the cross sections intersect the main channel and
distributaries obliquely, they have been modified to show true width to accurately
estimate hydraulic conditions.
Cross section 5
The plot of elevation as a function of distance from the southern edge of the study area
shows a marked drop in elevation as one moves north from the main channel. The slope
of this line is somewhat greater than the slope of the channel, creating conditions
conducive to lateral flow. Without the robust left bank riparian vegetation, northerly
channel movement would have already occurred.
The gradient between the left bank at the US 101 bridge to the bottom of Sylopash
slough, a distance of approximately 1014 ft, is 0.9 %. This is three times the bankfull
slope of the mainstem and offers a very attractive route for overbank flow. A significant
flood, greater than the 10 year event, may begin to cut a permanent distributary channel
here and may be the site of a future avulsion.
The N3 distributary is the most upstream channel to leave the mainstem in this reach.
Bankfull flows at the bridge are 7000 cfs N3 only has the capacity to contain flows of
200 to 300 cfswhich amounts to roughly 4% of the main channel discharge at this stage.
During higher recurrence interval storms, discharge may be significantly higher. Shear
stress in this channel during a 5 year flood in the mainstem is sufficient to move the
general bedload found in the main river (a pebble count was not done in any of the
distributaries). This means that relatively frequent floods will keep this distributary
scoured, providing that wood does not plug the mouth or a bar does not develop just
inside the mouth. Water also enters the distributary from overland flow and, in part,
accounts for the general downstream increase in width.
12/22/2004 12 Dosewallips Reach Analysis
Cross section 4
Cross section 4 clearly shows the aggredation of material in the vicinity of the channel.
The difference in elevation between the Dosewallips River left top -of -bank and the flood
plain surface 200 hundred feet north from the main channel is about 2.5 feet lower. That
amounts to 57% of the bankfull depth.
Cross section 3
Cross section 3 shows that the main channel is significantly narrower at this point than
where the previous cross sections intersected the channel. This channel cross section has
the least hydraulic capacity of any of the sampled channel cross sections, and as a result
of the limited channel capacity here, the adjacent flood plain is scarred with overflow
channels. This cross section shows that the area on the right bank has been built up,
possibly with dredge spoils.
In the table below are tidal statistics at Quilcene bay, converted to the study datum,
NAVD 88.
Table 8 Quilcene Bay tidal statistics.
All of the channel thalwegs, mainstem and distributary, have a very similar elevation at
this point in the estuary, with the exception of Sylopash Slough and the channel that runs
along the base of US 101. Cross section 3 is the first point measured where the influence
of tide begins to be felt, backwatering up the channel and forcing sediment deposition.
All the channel thalwegs are below MSL, indicating that they are dewatered half the time
and experience normal flow capable of scour from overbank flow. But they are
backwatered by tide 10 to 20 percent of the time, forcing deposition.
Surface texture patterns on the left bank flood plain along this cross section are complex.
Some patches with no signs of recent deposition or scour are at the same elevation as
similar, nearby patches with significant signs. It is the nature of cross sections to miss
features that occur between them. This complexity, and the inability of the methods to
discern it, is not considered important for this study.
Cross section 2
Most of cross section 2 is under the influence of tide. All but the wooded area right along
the banks of the main channel is salt marsh with an elevation of approximately 9 ft. Lida'
shows a slightly higher and more variable surface than actual ground elevation.
12/22/2004 13 Dosewallips Reach Analysis
NAVD
MEAN HIGHER HIGH WATER (MHHW)
8.7
MEAN HIGH WATER (MHW)
7.9
MEAN SEA LEVEL (MSL)
4.1
MEAN LOW WATER (MLW)
0.3
MEAN LOWER LOW WATER (MLLW)
-2.6
All of the channel thalwegs, mainstem and distributary, have a very similar elevation at
this point in the estuary, with the exception of Sylopash Slough and the channel that runs
along the base of US 101. Cross section 3 is the first point measured where the influence
of tide begins to be felt, backwatering up the channel and forcing sediment deposition.
All the channel thalwegs are below MSL, indicating that they are dewatered half the time
and experience normal flow capable of scour from overbank flow. But they are
backwatered by tide 10 to 20 percent of the time, forcing deposition.
Surface texture patterns on the left bank flood plain along this cross section are complex.
Some patches with no signs of recent deposition or scour are at the same elevation as
similar, nearby patches with significant signs. It is the nature of cross sections to miss
features that occur between them. This complexity, and the inability of the methods to
discern it, is not considered important for this study.
Cross section 2
Most of cross section 2 is under the influence of tide. All but the wooded area right along
the banks of the main channel is salt marsh with an elevation of approximately 9 ft. Lida'
shows a slightly higher and more variable surface than actual ground elevation.
12/22/2004 13 Dosewallips Reach Analysis
delta progradation rate of 6 feet per year is at all representative of the past, the delta front
was back at the entrance to N3 roughly 440 years ago.
N1, N2, and N3 are all blind channels. They do receive scouring flow during flood
events. The lowermost entrance to N3, the site of the cabled cottonwood log, has only
recently been reestablished and is not continuous. Considering how rapidly these old
distributary mouths become blocked with wood and sediment, the removal of debris and
sediment in order to increase the frequency of active flow should not be part of a long
term habitat restoration plan.
Habitat
There are a number of habitat types in the study area. The primarily types are
o mud flats
o shallow water habitat below MHW with a coarse texture
o tidal channels
o salt marsh
o freshwater marsh
o distributary rearing
o distributary spawning
o main channel rearing
o main channel holding
o main channel spawning
o main channel edge
A detailed evaluation of the area and condition of these habitats will not be a part of this
report. While we might benefit from such an analysis, it is likely that it will not change
the restoration activities recommended.
Land use, land cover, and infrastructure influence habitat in a number of ways in the
study area, but primarily it affects channel evolution. Sinuosity has been reduced to a
minimum in the lower Dosewallips. This was accomplished by diking and dredging over
the last 50 years, and possibly for some time before that. As far as we can tell, the
channel sinuosity was never very high. Sinuosities greater than 1.4 can be found in broad,
flat, and fine - grained coastal estuaries. This coarse - grained high gradient delta would
naturally have a relatively straight main channel, diversity primarily formed by bars,
distributaries, and islands of deposited sediment pulses. The 1883 T sheet shows a
channel that is relatively straight for 2500 feet and then bifurcates around an island. The
current channel travels straight from the US 101 bridge to about MHHW (2644 feet),
where it splits into a number of distributaries, basically falling off the deposited sediment
wedge.
Edge habitat on the right bank main channel has been severely impacted. This bank was
formed by a rip rap dike in 1978, significantly simplified compared to the more natural
left bank. Vegetation is minimal, large wood nearly absent, and the bankline straight and
12/22/2004 15 Dosewallips Reach Analysis
uniformly steep. There is some overhanging vegetation and a few large trees, but this
would be considered minimal.
The 1978 dike has also eliminated the possibility of distributary development on the right
bank. At the point where dike ends, the distributary S 1 begins. Land use on the right
bank, Dosewallips State Park, generally precludes channel development, either as
increased sinuosity, distributary development or avulsion of the main channel to the
south. Depending on Park's attitude toward natural systems, there may still be
opportunities to foster channel complexity. The history of Park ownership has been one
of manipulating the river for specific human needs, so a change in attitude would be
necessary. There are portions of this reach that have been built up over the elevation of
the adjacent natural flood plain. Cross section 3 shows where fill has elevated the original
flood plain between stations 700 and 900. Excavation of this fill is not likely. A first step
is to remove rip rap in selected areas and establish a vegetated buffer with a low natural
slope.
Table 9 Main Dosewallips River distributaries and blind channels.
The restoration goals specifically state that the loss or isolation of distributary channel
habitat is a problem. Table 9 gives the area of the various channel habitats on the
Dosewallips delta shown in Figure 4. Of the 18.6 acres of tidal and distributary habitat,
only 24% is connected distributary habitat and of this only S 1 conveys out - migrating
juveniles to adjacent salt marsh habitat. Juveniles must go out to the delta front to move
north to the adjacent salt marsh habitat, exposing themselves to deeper water and
increased predation at an early age. We have discussed the reactivation of the
distributaries and how they are quickly occluded by natural mechanisms.
There is a substantial amount of tidal channel habitat in the Dosewallips estuary, although
much of it is not connected to the river, and not likely to be through maintenance -free
restoration techniques.
12/22/2004 16 Dosewallips Reach Analysis
Approximate area in acres
Distributary
Total area
Non -tidal Tidal
Connected
Blind
N1
0.8
0.8
0.8
N2
3.4
0.7 2.7
3.4
N3
3.6
1.0 2.6
3.6
N3 trib
0.3
0.3
0.3
N4
1.9
1.9
1.9
S1
1.2
1.2
1.2
Sylopash slough*
4.2
4.2
4.2
New distributaries
1.6
1.6
1.6
0.9
0.9
0.9
0.7
0.7
0.7
Totals
18.6
1.7 16.9
4.5
14.1
*Sylopash slough has not been a distributary for at least 120 years.
The restoration goals specifically state that the loss or isolation of distributary channel
habitat is a problem. Table 9 gives the area of the various channel habitats on the
Dosewallips delta shown in Figure 4. Of the 18.6 acres of tidal and distributary habitat,
only 24% is connected distributary habitat and of this only S 1 conveys out - migrating
juveniles to adjacent salt marsh habitat. Juveniles must go out to the delta front to move
north to the adjacent salt marsh habitat, exposing themselves to deeper water and
increased predation at an early age. We have discussed the reactivation of the
distributaries and how they are quickly occluded by natural mechanisms.
There is a substantial amount of tidal channel habitat in the Dosewallips estuary, although
much of it is not connected to the river, and not likely to be through maintenance -free
restoration techniques.
12/22/2004 16 Dosewallips Reach Analysis
The role and risks of large wood in the main channel is a complex problem in this reach.
The reach is simplified and one of the primary benefits of large wood is complexity,
although that complexity is accompanied by some risks. Significant accumulations of
wood create backwater, increase flooding, and produce turbulence and local scour. At
this conceptual stage in the design process, and without doing sophisticated modeling, an
estimate of acceptable wood loading is given here. Gippell1 found that the measurable
backwater threshold for wood is 0.0056 ft3 /ft2, which for this reach is about 142 ft3/W�h,
or about one key piece per 2.7 channel widths. This is a relatively small amount,
considering that for a jam to be stable it must be composed of a number of key pieces and
a still greater number of functional ones. Another threshold is that wood structures should
not exceed a blocking ratio of greater than 0.112. The blocking ratio is the proportion of
the channel cross section blocked by the wood structure. Clearly, if wood structures are
proposed, a better analysis of the hydraulic consequences should be done.
Depending on which bank the wood is located and its reach scale effects, scour may
affect Park activities and/or infrastructure, and could increase the risk of avulsion.
Key piece size for this reach of the Dosewallips has a volume of roughly 380 ft3 (10.75
m), from Martin Fox's recommendation cited in the Stream Habitat Restoration
Guidelines 13. Abbe 14 suggests that, in large alluvial channels, the ratio of log length to
bankfull width should be greater than 0.25, and the ratio of average log diameter to
bankfull depth should be from 0.5 to 0.8, with root wad. This leads to a log roughly 40
feet long and 2.5 to 3.75 feet in diameter, which has a volume of 195 to 443 ft3 without a
root wad.
Conclusions
Physical habitat deficiencies that critically limit fish, wildlife, and plant productivity
within the lower Dosewallips stream corridor in the study reach are few, but they are
large scale. Generally, water quality is good, sediment is abundant, much of the
immediate upland and estuary are left to develop naturally.
• As the east Olympic peninsula subsides and sea level rises, the north half of the
original Dosewallips estuary will sink without sediment input. The location of the
US 101 bridge constrains the lateral movement of the river and limits the range of
sediment deposition to the southern estuary.
• Channel avulsion is an inevitable consequence of delta progradation. At the rate
of 6 feet per year, providing the upstream channel elevation remains the same and
including sea level rise of about 300 mm in that time, progradation will lower
slope by about one third in the next 100 years. This is more than enough to halt
sediment transport in the main channel and force an avulsion at some point
upstream.
• In addition to the gradual increase in channel length mentioned in the previous
point, there is also the possibility of a catastrophic change in the channel
downstream of US 101 due to a major flood accompanied by a release of
sediment and wood. Rather than a gradual build -up to an avulsion threshold,
12/22/2004 17 Dosewallips Reach Analysis