sound absorbing acoustics

1. Introduction
to Acoustic
Measuring
Equipment
Klamath Falls and
Chiloquin, OR
September, 19 – 23,
2011
U.S. Geological Survey
TEchnical training in Support of
Native American Relations
(TESNAR) - 2011
Mark Uhrich, USGS, Portland, OR (mauhrich@usgs.gov)
Marc Stewart, USGS, Central Point, OR (mastewar@usgs.gov)
Glen Hess, USGS, Portland, OR (gwhess@usgs.gov)

2.  Ocean going boats used “speed logs” to measure speed of the
boat.
 “The first commercial ADCP, produced in the mid-1970’s,was an
adaptation of a commercial speed log” (Rowe and Young, 1979).
 1980s Doppler technology continue to involve
 Early 1990s ADCP become more widespread in the USGS and
other agencies.
 2012 Acoustic based instruments become the most common
instrument type used in the USGS (Flowtrackers and ADCPs)

3. 3
Much of the material in the presentation is
borrowed from USGS Hydroacoustic Classes

4. TRDI Rio
Grande
ADCP
SonTek
RiverSurveyor
TRDI
StreamPro
SonTek Flow
Tracker

5.  Ceramic transducers send and receive
pulses of sound
 Center transducer transmits the
sound, while the transducers on the
arms are receivers
 Location of velocity measurement is
called the sample volume
 Sample volume is located about 4
inches from the transmitting
transducer
 Measures velocity based on the
Doppler shift

6. Acoustic
Doppler
Current
Profiler
Sound Waves and the
Doppler Shifts are used to measure
Water Velocity
Profiles
WHAT IS AN ADCP?
LET’S LOOK AT THE NAME
“A” “D” “C” “P”

7. Crest
+
-
Trough
Water wave crests and troughs are points of
high and low water elevations.
Sound wave “crests” and “troughs” consist of bands of
high and low air or water pressure.
Trumpet ADCP Transducer

8.  Uses Doppler shift to measure water velocity
 The Doppler effect is the change in a sound's observed pitch (frequency)
caused by the relative velocities of the sound source and receiver.

9. fD = Doppler Shifted Frequency
fS = Source Frequency (frequency of ADCP)
V = Velocity of scatterers in water
C = Speed of Sound (dependent on water
char.)
fD = fS * V/C

10. V = fD / 2fS * C
V = water velocity =
scatterer velocity
Important
We assume that, on average, scatterer velocity
equals water velocity
Violation of this assumption will lead to errors
in water velocity computation.
Note: The 2 in the equation is result of two Doppler shifts, one
as the sound goes out and another as it returns

11. Water-velocity measurement is biased toward the fish velocity
Water
Fish:
Water
Stationary object:
Rock
Water-velocity measurement is biased toward zero

12.  A temperature error of 4o C or salinity error of 12 ppt will
result in a 1% velocity error
 The instrument must have an accurate temperature
sensor and must be configured for the correct salinity
 Rule of thumb: Specific conductance generally below 5000
uS/cm should not significantly affect C
 Policy: All acoustic instruments must have independent
temperature check
(within 2 degrees C)
V = (fD /2fS *)C
Important
Speed of sound (C) must be computed
accurately by the instrument.

13. Picture at
time T2
Picture at
time T1
S
V=S/(T2-T1)
 To measure velocity,
ADCPs listen to the
returns at two
separate times

14. Usually ADCPs use PHASE CHANGE to measure the speed,
instead of measuring the change in frequency of the wave (how
far the cars have travelled)
Phase is the fraction of
a wave cycle elapsed
relative to a point – or
when thought of as the
wheel on the left, how
much it has rotated

15.  Because we don’t know the direction the cars are traveling, we must account for both
positive and negative values (we measure a half rotation either direction)
 Set the time between pictures (lag) to optimize the tire rotation for the expected
speed– Longer time (lag) = more precise measurement
 Too long of time between picture (long lag) may cause the distance car travels to
exceeds a half rotation and result in a measurement error called ambiguity error
 Short time (lag) limits precision (increased random noise), but decreases possibility
of an ambiguity error
Picture 1 Picture 2

16.  Lag = Time between pulses in a ping
 Long lag = accurate measurements
 Long lag = low ambiguity velocity
 Exceed ambiguity velocity = ambiguity error
 Ambiguity error = inaccurate measurements
 Lag needs to be optimized based on maximum speed
 SonTek usually has short lags (no chance for ambiguity
errors but noisy – pictures close together and wheel
hasn’t turned much in lower velocites)
 TRDI – usually has longer lags that need to be adjusted
for conditions (less noise, but chance for ambiguity
errors if not adjusted correctly)

18. S
How well the two pictures can be aligned
If there is too long of a time between pictures, cars may be in different
locations relative to each other, or the to pictures could contain totally
different cars and the distance S may not be determined

20. Produce sound waves (pulses) and then listen to returning sound
waves
ceramic element protected with a urethane coating
ADCPs use the same transducer to both transmits and receive the
pulses.

22. 4 beams can be
resolved into: x, y, z
and error velocities
Since each Transducer only measures the velocity component parallel
to the beam, multiple transducers are needed
RiverRay forms 4 beams from
the single phased array
transducer
M9 only uses 4
beams at a time
to compute a
velocity

23. Homogeneous
(Low error velocity)
Non-homogeneous
(High error or D velocity)
?
The difference between
beam pair vertical velocities
is reported in software as
Error or D Velocity

24.  Error velocities should be
randomly distributed
 areas of high area error velocities
may occur when water is not
flowing at similar magnitudes and
direction in all beams (example:
turbulence)
 Error velocities may also be the
result of an instrument measuring
one beam velocity wrong
Behind
Bridge Pier

25.  How the pulses are transmitted into the water
and sampled can vary and be optimized for the
conditions
 This configuration is commonly called “water
mode”
 Some of the newer ADCPs automatically adjust
the configurations for the environment on the
fly
 Until recently the majority of ADCPs currently
in use must be set up prior to data collection

26. cell 1
cell 2
cell 3
echo echo echo echo
Transmitting
start end
Gate
1
Gate
2
Gate
3
Gate
4
Time
Blank
Bin 1
Bin 2
Bin 3
Bin 4
Distance
From
ADCP
cell 4
Blanking
A B C

27. Depth Cell

28. The faster the boat travels,
the faster the velocity of the water relative to the ADCP.

29.  ADCP’s can also measure the speed of the instrument
or boat by measuring the Doppler shift of a pulse off of
the bottom
 This is called bottom tracking and assumes that the
streambed is stationary
 Sediment transport on or near the streambed can affect
the Doppler shift of the bottom-tracking pulses, which
can result in the measured boat velocity being biased in
the opposite direction of the sediment movement. This
is referred to as a Moving Bed condition

30.  Bottom Track pulses are also used to measure
depth
 Typically 4 beam depths are averaged
 SonTek also can use a vertical beam dedicated
to depth

31. A single pass across the river is called a transect, a
discharge measurement is usually comprised of
multiple transects averaged together

32.  Measured Q = ∑(V x A)
 V = Velocity perpendicular to
boat path for the ensemble
 A = Depth Cell Size x Width
 Width = boat speed x time
since last valid ensemble
32
Assumption made: the measured boat and water velocities are
representative of the boat and water velocities since the last valid
data. The longer it has been since the last valid data, the greater
the error may be in this assumption
The above is equal to the cross product of the boat and water speed x
depth cell size and the time since last ensemble, which is how
software computes Q in a depth cell

34. Depth
Valid Ensembles
(Profiles)
Top (Estimated) Layer
Middle (Measured) Layer
Bottom (Estimated) Layer
Edges (Estimated)

35. Velocity cross product (X-value), ft2/s2
f1
f3
f4
f5
fn
f2
Measured
Estimated
Power fit
Free surface
Distance
from
the
bed
(Z),
feet
 Q in the top and
bottom unmeasured
areas is estimated for
each ensemble, based
on the measured data
 The typical method is
to use a power fit of
the measured data,
but other options are
available when this is
not valid

36. 0 0
(-) (+)
(-) (+)
Distance
from
the
bed
(Z),
feet
Unidirectional Flow Bi-directional Flow
Velocity cross product (f-value), ft2/s2
Distance
from
the
bed
(Z),
feet

37. Range
from
Bottom
Range
from
Bottom
Water Velocity Water Velocity
Depending on direction, wind can either cause the profile to bend
either way at the water surface
The magnitude of this may cause the standard power
fit to be a poor choice for top extrapolation, in this case the
software has options to only use data near the
surface for estimating the top Q

38. dm=last measured depth
dm
Measured by ADCP
Measured by User
L = distance from last
ensemble to edge of water
L
The averaged measured velocity is
multiplied by the averaged
measured depth, the measured
length, and finally by a coefficient
to account for the shape of the
edge (.35 for triangle and .91 for
square)
Average multiple
ensembles to get
an accurate depth
and velocity
MEASURE the
edge distance.
Vm = last measured velocity
Vm

39. From: Water Resources Investigations Report
00-4036. By K. M. Nolan and Shields
Online Training Class SW1271

40. Right Q
Left Q
Bottom Q
Middle Q
Total Q = Left Q + Right Q + Top Q + Bottom Q + Middle Q
Top Q

41. From: Water Resources Investigations Report
00-4036. By K. M. Nolan and Shields
Online Training Class SW1271
Reach - Straight and uniform for a
distance that provides for uniform
flow
Streambed - stable free of large
rocks, weeks or other obstructions
A poor cross section = poor
measurement regardless of the
accuracy of your point velocities

42.  Reach - Straight and
uniform for a distance
that provides for
uniform flow
 Streambed - stable free
of large rocks, weeks or
other obstructions
 A poor cross section =
poor measurement
regardless of the
accuracy of your point
velocities