[Acoustics] OSW Technical Memorandum 2004.04--Policy on the use of
the FlowTracker for discharge measurements.
Kevin Oberg
kaoberg at usgs.gov
Thu Jun 3 14:16:50 CDT 2004
FYI, the following memo was just issued by OSW and is available at
http://hydroacoustics.usgs.gov/memos/OSW2004-04.pdf . As a reminder other
OSW technical memos regarding Hydroacoustics are available at
http://hydroacoustics.usgs.gov/policy/index.html
Kevin
In Reply Refer To: June 2, 2004
Mail Stop 415
OFFICE OF SURFACE WATER TECHNICAL MEMORANDUM 2004.04
SUBJECT: Policy on the use of the FlowTracker for discharge measurements.
The purpose of this memorandum is to provide information and guidance
regarding the use of SonTek/YSI's FlowTracker for making wading discharge
measurements. More than 100 FlowTrackers have been purchased for use by
the U.S. Geological Survey (USGS) as of October 2003. Although the Office
of Surface Water (OSW) continues to conduct laboratory tests of these
instruments, we can state that the FlowTracker performs well in many stream
environments. Guidance on the general use of FlowTrackers is important,
especially for their use in low-backscatter environments and with
boundaries, such as rocks, logs, or large stream bedforms.
Background
The Water Resources Discipline (WRD) Instrumentation Committee (ICOM)
funded a project in the Indiana District, beginning in October 2000, to
guide the adaptation of hydroacoustic current meters for discharge
measurements in shallow, wadeable streams. SonTek/YSI designed the
FlowTracker to be used for making discharge measurements in wadeable
streams using a standard USGS wading rod. During the last 4 years, OSW
worked with SonTek/YSI to provide guidance on standard USGS
discharge-measurement methods and algorithms. WRD and SonTek/YSI personnel
also conducted a number of field and laboratory tests of the FlowTracker.
Tow-tank tests of the FlowTracker
In April 2002, two FlowTrackers were tested in a tow-tank at the USGS
Hydraulics Laboratory at Stennis Space Center, Mississippi. FlowTrackers
were tested at various tow-cart speeds in order to determine their
accuracy. These FlowTrackers met Price AA and the manufacturer's specified
accuracy limits for most tow-tank runs (
http://hydroacoustics.usgs.gov/reports/SEMpaper.pdf).
In November 2002, 31 FlowTrackers were tested in a tow-tank at the USGS
Hydraulics Laboratory at Stennis Space Center, Mississippi to ascertain the
performance of a large batch of instruments. FlowTrackers were tested at
eight tow-cart speeds, ranging from 0.10 to 2.99 feet per second (ft/s).
Unfortunately, data analysis indicated various problems related to the test
procedures. These problems included acoustic interference caused by
operating instruments simultaneously, residual currents in the tow-tank,
and possible test-equipment-induced noise. Testing hydroacoustic
instruments in a tow-tank is a difficult process and the USGS and other
organizations still are experimenting on how to properly conduct these
tests. A revised tow-tank test plan (
http://hydroacoustics.usgs.gov/downloads/flowtracker.field.test.pdf) was
drafted based on the November 2002 test experience. In March 2003, new
tests were conducted in the tow-tank based on the revised test plan.
Additional tests of the FlowTracker also were conducted in a jet tank at
the Hydraulics Laboratory. Data from both these tests currently are being
analyzed.
Acoustic Doppler Velocimeters (ADV), of which the FlowTracker is a special
adaptation, have been tested and shown to be an accurate instrument. ADV's
commonly are used in laboratory and field investigations in hydraulics. A
list of selected papers published in peer-reviewed journals that relate to
testing and evaluation of ADV's is provided in the "Selected References to
ADV Evaluations" section at the back of this memorandum. The results of
these and other papers provide confidence in the underlying technology that
is used in the FlowTracker for making velocity measurements.
FlowTracker results in low backscatter environments
Field tests at 29 sites in 2001 and 5 sites in 2002-03 have shown that the
discharge-measurement algorithms used by the FlowTracker are correct. For
the 2002-03 field tests, FlowTracker discharge measurements were within 5
percent of discharges measured with Price AA or pygmy meters, with mean
velocities ranging from 0.48 to 1.9 feet per second (ft/s), mean depths
ranging from 0.7 to 2.1 ft, and measured discharges ranging from 12 to 120
cubic feet per second (ft3/s). Whereas FlowTracker measurements compared
favorably with mechanical current meter measurements, all measurements were
made in streams with few boundary effects, and in streams that carried
enough suspended backscatter material to produce adequate FlowTracker
signal-to-noise ratios (SNR). Data from FlowTracker measurements made in
the last 6-9 months have indicated potential data-quality problems in
streams with low suspended backscattering material.
SonTek/YSI indicates that the SNR should be greater than 10 decibels (dB)
for optimal data quality (SonTek FlowTracker Operation Manual, Section
1.4.1), but FlowTrackers may collect reliable data when the SNR is lower.
Field and laboratory data analyses have indicated that velocity data are
unreliable when the SNR is below 4 dB.
In June 2003, SonTek/YSI released firmware version 2.4. It is recommended
that all FlowTrackers be upgraded to firmware version 2.4 by contacting
SonTek/YSI. This firmware version includes changes that may increase data
quality in some low backscatter environments. Information on obtaining the
firmware upgrade can be obtained from
http://www.sontek.com/e/0307a/ftupgrade.html or by contacting SonTek/YSI by
email at support at sontek.com or by phone at (858) 546-8327.
In July 2003, SonTek/YSI made a hardware upgrade available for FlowTrackers
that involves a new high-sensitivity receiver component. The manufacturer
states that this upgrade can improve the SNR by 5 to 10 dB, which may
dramatically improve the capability to work in low backscatter
environments. OSW has not yet tested FlowTrackers equipped with this
upgrade, but has plans for testing in the next 6-12 months. The hardware
upgrade is available from SonTek/YSI and is standard in all FlowTrackers
shipped since July 15, 2003.
FlowTracker results in streams with potential boundary effects
Recently reviewed FlowTracker discharge measurements indicate numerous
boundary effect problems. A boundary effect occurs when the sampling
volume for the FlowTracker includes some solid, stationary boundary such as
a cobble or boulder in the stream. These problems generally are more
common in streams with rough beds and at shallow depths. The volume of
water used by the FlowTracker to make velocity measurements (called the
sample volume) is approximately 0.25 cubic centimeters in size and is
located 10 centimeters (about 4 inches) away from the center transmitting
transducer on the FlowTracker (figure 1). If the FlowTracker sample volume
includes or is near a solid boundary, the velocity data will be corrupted
and biased low. The low bias is caused by the FlowTracker measuring a zero
velocity from the solid, stationary boundary for at least some part of the
sample volume.
Guidelines for use of FlowTrackers in making discharge measurements
FlowTracker users should take the following actions when making discharge
measurements:
1. Under appropriate conditions and use as described below, the
FlowTracker can be used in place of the mechanical Pygmy and Price AA
meters for wading discharge measurements. The same procedures and polices
for use of the mechanical meters apply to the use of the FlowTracker; these
include:
· Measurement site selection -- the measurement section should be
within a straight reach, where streamlines are parallel. The streambed
should be relatively uniform and free of numerous boulders, debris, and
heavy aquatic growth. The flow should be relatively uniform and free of
eddies, slack water, and excessive turbulence.
· Velocity sample time ? under normal measurement conditions, each
point velocity measurement should be sampled for a minimum of 40 seconds.
Under extreme flow conditions, such as rapidly changing stage, a shorter
sample time may be used to lessen the time needed to complete the discharge
measurement.
· Location of velocity observations in each vertical ? at depths below
1.5 feet, the 0.6-depth method should be used; at depths between 1.5 and
2.5 feet when a non-logarithmic velocity distribution may be present and
the 0.8 depth sample location would be more than 2 inches from any
boundary, the two-point method should be used; and at depths greater than
2.5 feet, the two-point method should be used. If a non-logarithmic
velocity profile is discovered when using the two-point method, the
three-point method should be used.
2. Prior to each field trip, or about once per week, perform a
FlowTracker ADVCheck (FlowTracker Operation Manual Section 6.5.4). The
ADVCheck should be logged to a file and archived. The name of the ADV
check file should be written on the measurement note sheet. Perform
FlowTracker field diagnostic procedures (FlowTracker Operation Manual
Section 3.3.2) prior to starting every measurement.
3. Monitor the FlowTracker SNR readings during the measurement for SNR
readings that are less than 4 dB. The FlowTracker will display a warning
at the end of a velocity measurement if the SNR for any beam is less than 4
dB. If the SNR is below 4 dB, try moving to a different measurement
section where backscatter may be higher. If a section with an acceptable
SNR cannot be located, the FlowTracker should not be used to make a
discharge measurement.
4. Avoid measurement sections with abrupt changes in bed topography.
These changes can result because of such things as large rocks or cobbles
in the measuring section. Abrupt changes in bed topography may cause
boundary effects leading to inaccurate velocity measurements. During the
measurement, velocities should be monitored for unrealistically low
velocities, and also for unusually large SNR values (a solid boundary
should cause an increase in the SNR). Be aware of the location of the
FlowTracker sample volume when measuring. The sample volume typically is
10 centimeters (about 4 inches) from the center transmitting transducer
(figure 1). Avoid placing the sample volume within 2 inches from any solid
boundary. Although this recommended placement does not mean that the
FlowTracker cannot be used closer to boundaries, nevertheless, extra care
should be taken in those situations.
5. Pay close attention to the flow angle reported by the FlowTracker.
The wading rod (with FlowTracker attached) always should be held
perpendicular to the tagline, so that the pulse generated by the
transmitter (see figure 1) is parallel to the tagline. Ideally, the
tagline should be set up in the cross section to be measured such that flow
is perpendicular to the tagline.
Flow angle, as calculated by the FlowTracker, is defined as the
direction of flow relative to the x-direction of flow, so that
(Embedded image moved to file: pic19169.pcx)
where, Vy is the velocity in the y-direction (parallel to the tagline)
and Vx is the velocity in the x-direction (perpendicular to the tagline)
used to calculate discharge.
The flow angle calculated by the FlowTracker can result from two
sources: (1) the flow is not perpendicular to the tagline, and (2) the
flow is perpendicular to the tagline but the wading rod is not being
held correctly relative to the tagline, as described above. Regarding
(1), some small angles and variation in the flow angle at a site is not
unusual. However, if large fluctuations of flow angles are reported,
measurements should be made at another section with more uniform flow.
Regarding (2), holding the FlowTracker such that it is skewed at any
angle relative to the tagline will result in a measurement of velocity
that is biased low. Small angles do not result in significant biases,
but because of these biases, users should be careful to minimize this
error. If the FlowTracker is held such that it is skewed at an angle of
approximately 8 degrees from the tagline, the measured velocity may be
in error by as much as 1% (assuming that flow is perpendicular to the
tagline).
Large variations in flow angles may be indicative of poor or
inconsistent alignment of the wading rod or poor site selection for the
measurement.
6. Each FlowTracker measurement is stored in a binary file (.WAD
extension) that a user can download to the computer and, using FlowTracker
software, extract four files containing the measurement data. These files
include a summary of the measurement (.DIS) with average subsection
velocities and SNR readings, and a "raw" data file (.DAT) that lists
1-second velocities and SNR readings. The effect of unrealistic velocities
may not be apparent in the average subsection velocities in the .DIS file.
Thus, users should review the 1-second velocity and SNR data in the .DAT
file if any of the following results are observed:
· Average subsection SNR readings in the .DIS files are less than 10;
· .DIS file average subsection velocities or SNR readings appear to be
inconsistent across the measurement section or unrealistic for site
conditions;
· SNR readings are observed to fall below 4 dB any time during a
discharge measurement;
· The .DIS file shows unrealistic subsection velocity spikes;
· The .DIS file shows subsection boundary variable values of 3 (poor);
· The mean standard error of velocity for the measurement shown in a
DIS file exceeds 5 percent of the mean velocity for the measurement; or,
· There is any reason to suspect data-quality problems (such as the
presence of irregular bedforms or other solid boundaries that could
interfere with the sample volume).
The FlowTracker software contains a velocity spike filter (number of spikes
filtered is displayed in the .DIS file); however, the spike filter may not
remove all bad or questionable velocity data from the subsection average.
Furthermore, boundary variables of 1 (best) or 2 (good) do not necessarily
mean that boundaries did not cause velocity biases in the measurement data.
Users should refer to the FlowTracker Operation Manual Section 1.4 and
become familiar with the detailed descriptions of SNR's, standard error of
velocity, boundary variables, and velocity spike filtering.
OSW is making available a program called DatView, written by Mike Rehmel of
the Indiana District, to facilitate the rapid review of FlowTracker data
contained in the .DAT file. At present, this program is considered Beta
software ? but appears to be stable. The program allows users to load
FlowTracker .DAT files and examine plots of measurement velocities and
SNRs. An example of a plot generated using DatView for a measurement with
low SNRs and associated velocity spikes is attached (figure 2). DatView can
be obtained at http://hydroacoustics.usgs.gov/download/DatView.exe.
Low SNR readings and velocity spikes can degrade the FlowTracker
measurement quality. For this reason, users should consider downgrading
the quality of measurements made in low-backscatter environments.
Summary
Test data collected to date indicate that FlowTrackers are accurate
instruments for making wading discharge measurements in many stream
environments. If a FlowTracker is used outside of its designed operating
environment, data can be compromised resulting in inaccurate measurements.
Until OSW completes field and laboratory testing, users of FlowTrackers
should make comparison measurements with Price AA current meters in various
stream environments per OSW memorandum regarding Acoustic Doppler
Velocimeter, dated October 4, 2001. A test plan for FlowTracker comparison
measurements can be obtained at
http://hydroacoustics.usgs.gov/downloads/FlowTracker.field.test.pdf.
OSW requests that any FlowTracker users share results of their experiences
and other information regarding FlowTracker use. Please send information
to Mike Rehmel, Indiana District, (msrehmel at usgs.gov), 317-290-3333 ext.
158 or to the USGS Hydroacoustics Work Group (hawg at simon.er.usgs.gov). OSW
is interested in receiving information regarding measurement comparisons,
problems identified with the instrument or software, possible hardware or
software enhancements, and other important issues to users. Field and
laboratory tests of these instruments are continuing and test results will
be communicated in future memorandums, by the OSW Hydroacoustics Web pages
(http://hydroacoustics.usgs.gov/), and through the Acoustics mailing list.
Questions concerning the information presented in this memorandum should be
directed to Kevin Oberg (kaoberg at usgs.gov, 217-344-0037 ext. 3004) or Mike
Rehmel
(msrehmel at usgs.gov, 317-290-3333 ext. 158). Specific questions concerning
the DatView program should be directed to Mike Rehmel.
/signed/
Setephen F. Blanchard
Chief, Office of Surface Water
Distribution Code: A,B,CD,OSW Staff all, District SW specialist, and FO
Selected References to ADV Evaluations
Kraus, N. C., Lohrmann A., and Cabrera R., 1994, New acoustic meter for
measuring 3D laboratory flows: Journal of Hydraulic Engineering, v. 120,
p. 406-412.
Snyder, W.H., and Castro, I. P., 1999. Acoustic Doppler Velocimeter
Evaluation in Stratified Towing Tank: Journal of Hydraulic Engineering, v.
125, no. 6, p. 595-603.
Song, T., and Graf, W. H., 1996. Velocity and turbulence distribution in
unsteady open-channel flows: Journal of Hydraulic Engineering, v. 122, no.
3, p. 141-154.
Voulgaris, G., and Trowbridge, J. H., 1998, Evaluation of the Acoustic
Doppler Velocimeter (ADV) for Turbulence Measurements: Journal of
Atmospheric and Oceanic Technology, v. 15, no. 1, p. 272-289.
Figure 1.?Schematic showing FlowTracker sampling volume and proximity to a
fixed boundary in the stream
(Embedded image moved to file: pic15724.pcx) (Embedded image moved
to file: pic11478.pcx)
Ideal FlowTracker measurement conditions FlowTracker measurement
illustrating conditions with
potential for interference from a
solid boundary
Figure 2.?Example Plot of FlowTracker .DAT File Using DatView
(Embedded image moved to file: pic29358.pcx)
*********************************************************************
Steve Blanchard Chief, Office of Surface Water
U.S. Geological Survey Phone: 703-648-5629
415 National Center Fax: 703-648-5722
Reston, VA 20192 Cell: 571-216-1423
sfblanch at usgs.gov
*******************************************************************
----------
Kevin Oberg
USGS - Office of Surface Water
Voice: 217.344.0037 extension 3004
Cell: 217.840.9739
http://hydroacoustics.usgs.gov/
More information about the Acoustics
mailing list