Environmental Engineering Reference
In-Depth Information
shear velocity was calculated by Rennie et al. (2002),
Rennie & Millar (2004), and Rennie & Church
(2007) by fi tting the vertical profi le of local stream-
wise water velocity measured with the ADCP to the
log law:
2.2.1.3 Summary: active hydroacoustics as
bed-load surrogate technology
Stationary measurements of apparent bed velocity in
sand and gravel reaches have been correlated to bed-
load transport rates measured concurrently from
physical sampling, dune tracking (for sand-bed
rivers), and bed shear. Apparent bed velocity distri-
butions measured from a moving boat have been
correlated to concurrent distributions of near-bed
water velocity, depth averaged water velocity, shear
velocity, and channel depth.
Error is a signifi cant limitation of computation of
apparent bed velocity. Instrument error constitutes
the majority of the error (Rennie et al . 2002). Raw
bed velocities are computationally very noisy, and
must be averaged. The error of the bottom track
velocity for a mobile bed is the same order of mag-
nitude as that for water velocity (Rennie & Millar
2007). Measurements taken from moving boats use
the inherent averaging of kriging to reduce error
(Rennie & Millar 2004; Rennie & Church 2007).
Another limitation of apparent bed velocity compu-
tation is that the technique needs calibration for each
site. The calibration is a function of the bed-load
sediment size and the operating parameters of the
ADCP, and can be infl uenced by near-bed suspended
transport (water bias). The ADCP requires manual
deployment, and can be purchased for about four-
fold the price of a turbidimeter.
Bottom track velocity is calculated using proprie-
tary fi rmware. Improvements to the fi rmware used
to determine apparent bed velocity would be helpful.
The spectrum of returned echoes could be used to
determine the range of velocities contributing to the
signal instead of estimating a spectral peak from the
autocovariance function to represent an apparent
average velocity.
Apparent bed velocity measurement using an
ADCP is a fast and non-intrusive surrogate technique
for computing bed-load transport. One major advan-
tage of using an ADCP to characterize bed-load
transport rates is the ability to measure the spatial
distribution of relative bed-load transport. From a
more general perspective, because quantifi cation of
bed-load transport is typically diffi cult and problem-
atic even in sand-bed rivers, any surrogate means for
providing quantifi ably reliable sand bed-load data is
desirable.
u
30
h
*
u
=
ln
(7)
κ
k
s
where u is the velocity at h ; h is the elevation above
the bed; u * =
τρ
is the shear velocity;
τ
is the bed
shear stress;
is the von Karman
constant (0.41); and k s is the bed roughness.
Signifi cant variations existed in the shear velocity
distributions mapped in Sea Reach, a sand-bed estua-
rine distributary of the Fraser River, Canada. (Rennie
& Millar 2004). Both the near-bed water velocities
and the depth-averaged water velocities were corre-
lated with the apparent bed velocities for spatial lags
up to about 10 m. Similarly, areas with high shear
velocity matched those with high apparent bed veloc-
ities. High shear velocities were found to stretch
from the upper left side to the lower right side of
the reach.
Velocity distributions were produced for a 5.5-km-
long gravel-bed reach of the Fraser River, Canada,
about 150 km upstream from the river mouth (Rennie
& Church 2007). Vertical velocity profi les, averaged
over a width of 7.7 m, were fi tted to the log law to
calculate the shear velocity. Apparent bed velocities
were interpolated by kriging onto a 25-m grid to yield
the spatial distribution. The distributions of fl ow
depth, depth-averaged water velocity, and shear
velocity were generated likewise. The distributions
for depth, depth-averaged water velocity (Fig. 2.7a),
shear velocity, and apparent bed velocity (Fig. 2.7b)
were very coherent. Maximum values of shear stress
were found in the deep bend pools of the thalweg just
downstream from areas of fl ow convergence. Areas
of fl ow separation and over shallow point bars had
lower shear stress. Apparent bed velocity matched
bed shear except in a deep pool adjacent to a rapidly
eroding bank, where highly turbulent fl ow existed.
This pool was located downstream from the river's
confl uence with a major side channel. The highest
apparent bed velocities were measured here with the
erosion due to high 3-dimensional turbulence in a
region of fl ow separation. The shear velocity, which
is calculated from mean velocity profi les, was not
estimated to be high at this location.
ρ
is the fl uid density;
κ
Search WWH ::




Custom Search