Geoscience Reference
In-Depth Information
40
◦
C, driving water temperatures up, especially in
simplified channels exposed to the sun. Tompkins
(2006) installed piezometers in the gravel at
the tails of pools (where surface water is driven
into the bed) and downstream within the riffles
(where hyporheic water emerges from the gravel
bed) (Figure 18.11). The results documented
hyporheic exchange from downwelling and
upwelling associated with undulations in the
gravel bed in the relatively undisturbed reach
upstream of the flood control project. Continuous
water temperature recorders in these piezometers
showed that diurnal fluctuations of surface waters
in Deer Creek commonly exceed 5
◦
Cinthe
summer, but the diurnal fluctuations of the deep
areas of pools influenced by shallow groundwater
were muted dramatically, typically only 1-2
◦
C
(Figure 18.12). Thus, hyporheic exchange can
reduce diurnal fluctuations and provide thermal
refugia for temperature-sensitive salmonids, with
potentially significant benefits to fish. Tompkins
(2006) observed that during the day, juveniles
of
and adding gravel to the channel) the geomorphic
analysis recommended addressing the underlying
problem and redesigning the flood infrastructure
to allow Deer Creek to flood much of its (still
mostly agricultural) floodplain in a controlled way,
protecting vulnerable structures by ring levees. This
approach would have the advantages of providing
better flood protection to the structures on the
floodplain, while reducing shear stresses in the
low flow channel, allowing bars to build and
riparian vegetation to establish within the floodway
that is now 'maintained' (cleared). The result
should be improved habitat in the channel of Deer
Creek. The potential for floodplain reconnection
to accommodate the design discharge through
increased cross-section area, higher hydraulic
roughness and reduced average velocities and thus
reduced shear stresses in the main channel was
demonstrated by a three-dimensional modelling
study conducted as a component of a PhD thesis
(MacWilliams, 2004).
Reduced shear stresses within the main channel
would allow riparian trees to become established
and mature along the banks, and allow complex
channel forms such as gravel bars to develop and
persist. With denser riparian vegetation along
channel banks, shading and channel complexity
would increase. Another potential benefit of
increased channel complexity is expected to
be increased exchange between surface water
and hyporheic waters (shallow groundwater
within the stream-bed gravels). The potential for
natural bed undulations to induce downwelling
of surface waters into the bed and upwelling
of hyporheic waters was demonstrated through
field observations with piezometers in the reach
upstream of the flood control project, where the
bed is still complex with large gravel bars and
pool-riffle sequences (Tompkins, 2006).
Perhaps the most significant ecological
implication of the natural surface-groundwater
interaction was its buffering effect on stream water
temperature. The alluvial reach of Deer Creek
provides rearing habitat for juvenile salmon and
trout, and summer water temperature is a major
limitation on salmonid habitat. During summer
months,
steelhead
trout
(
Oncorhynchus mykiss
)were
concentrated
almost
exclusively
in
sites
with
upwelling, cooler hyporheic water.
The proposed restoration strategy of setting
back levees would allow much of the flood waters
to flow across the floodplain, thus reducing
shear stresses on the bed of the main channel
during floods. Three-dimensional modelling
of flood flows demonstrates that significant
reductions in bed shear stress could be expected
(MacWilliams, 2004). The combination of lower
shear stresses from levee setbacks and a new
management approach that no longer involves
channel 'maintenance' (i.e. removal of in-channel
vegetation and gravel bars) could allow Deer
Creek to rebuild its channel complexity over time.
Detailed monitoring of bed mobility and channel
change over a three-year period from 2004 to 2006,
including a 5-year flood (2006), demonstrated that
the bed of Deer Creek is mobile even at flows with
return interval of less than 2 years, and that a
5-year flood results in mobilization of virtually the
entire bed, with substantial erosion, recruitment of
large wood from banks, build-up of gravel bars and
increase in complexity (Kondolf
et al
., 2008). In
daytime
air
temperatures
may
exceed
