Environmental Engineering Reference
In-Depth Information
SSC levels approach 0.1 g/L. Attenuation due to pres-
ence of sediment can be accounted for following
Flammer (1962). A coeffi cient,
is to assume the theoretical value for the slope,
B
,
equal to 0.1 and determine an appropriate value of
intercept,
A
= log
10
(SSC
measured
) - 0.1
RB
.
Limitations of the acoustic technique are well
described in the literature (e.g. Reichel & Nachtnebel
1994; Hamilton
et al
. 1998). One critical limitation
is the fact that it is not possible to differentiate
between concurrent changes in SSC and PSD (without
suffi cient calibrations) when using a single-frequency
instrument, as changes in both SSCs and PSDs can
result in a change in the backscatter signal strength.
In addition, there is an appropriate acoustic fre-
quency for a given PSD. Errors in estimates of SSC
will increase if a substantial fraction of the suspended
material includes particles that are too large or too
small for a response by a given frequency. For these
reasons, techniques or instruments that utilize more
than one acoustic frequency are preferable to single
frequency methods. Several applications of multi-
frequency instrumentation have successfully charac-
terized both SSC and mean particle size (Hay &
Sheng 1992; Crawford & Hay 1993; Thorne
et al.
1996; Topping
et al
. 2007).
Finally, an alternative approach for segregating
size fractions using a single acoustic frequency has
been developed by Topping
et al.
(2006, 2007) on
the Colorado River at Grand Canyon, Arizona, USA.
This approach segregates the silt-clay and sand com-
ponents of the suspension by taking advantage of the
fact that silt-clay tends to dominate acoustic attenu-
ation whereas sand tends to dominate backscatter.
Side-looking ADCPs are mounted on the river bank
that profi le across the river width; after removing the
two-way transmission losses, the slope of the back-
scatter profi le yields the attenuation coeffi cient,
which is strongly correlated with silt-clay SSC, while
the acoustic backscatter is strongly correlated with
sand SSC. The potential to segregate “wash load”
from “bed material suspended load” in sand-bedded
rivers warrants future testing of this methodology in
a wider range of environments.
ζ
, is defi ned as:
{
}
+
(
2
2
)
ζ
=−
K
(
γ
1
)
S
S
2
++
(
γ
τ
)
K a
43
p
6
⎣
⎦
(6)
where:
K
= 2
is the particle or aggregate wet
density divided by the fl uid density;
π
/
λ
;
γ
τ
= 0.5 +
9/(4
β
a
p
);
S
= [9/(4
β
a
p
)][1 + 1/(
β
a
p
)];
β
= [
ω
/2
v
)]
0.5
;
ω
is the kinematic
viscosity of water, in stokes. The two-way attenua-
tion from suspended particles, 2
= 2
π
f
,
f
is frequency in Hz; and
ν
α
s
in decibels per
centimeter, is equal to (8.68)(
)(SSC), where SSC is
dimensionless (1000 ppm = 0.001) and 8.68 is the
conversion from nepers to decibels. The fi rst term in
eqn. 6 is the attenuation from viscous losses and the
second term is the attenuation from scattering losses.
An alternative form for the scattering loss compo-
nent can be found in Richards
et al.
(1996).
From a practical standpoint, it is not necessary to
know the source level, nor is it typically feasible to
measure all the characteristics of suspended material
required to directly model target strength (Thevenot
et al
. 1992; Reichel & Nachtnebel 1994). Therefore,
following the derivation of Thevenot
et al.
(1992),
eqn. 3 is cast in terms of relative backscatter,
RB
=
RL
+ 2
TL
. After appropriate substitutions, the
sonar equation can be written in the desired form in
terms of SSC and relative backscatter as:
ζ
SSC
=
10
(
−
01
.
K
+
01
B
)
(7)
2
where:
K
2
is a parameter that includes terms for
source level, target strength, ensonifi ed volume, and
mass of suspended material.
The theoretical parameters
A
=
0.1
K
2
and
B
= 0.1
are appropriate for an SSC of uniform particles of
the same mass and other properties. For a distribu-
tion of particles in the fi eld, agreement with the theo-
retical values is experimentally checked by regression
of
RB
with measured estimate of SSCs at the same
location. Thevenot
et al
. (1992) determined the coef-
fi cient
−
0.1
K
2
to be equal to 0.97 and 1.43 for labo-
ratory and fi eld calibrations, respectively. They
determined values for the coeffi cient multiplying
RB
to be 0.077 (laboratory) and 0.042 (fi eld). Thus eqn.
7 can be used to compute a time series of SSC from
ADCP ABS at any distance from the acoustic trans-
ducer where valid backscatter data are available once
appropriate transmission losses and slope and inter-
cept values are determined. An alternative approach
−
1.2.5.2 Example fi eld application
A multi-instrument, multi-frequency system has been
established at the USGS streamgage on the Colorado
River at Grand Canyon, Arizona, USA, to produce
data from which continuous SSCs and SSLs can be
computed (Topping
et al
. 2007). The system uses
