In Reply Refer To: December
7, 2001
Mail
Stop 415
OFFICE
OF SURFACE WATER TECHNICAL MEMORANDUM 2002.03
SUBJECT: Release of WinRiver
Software (version 10.03) for Computing Streamflow from
Acoustic Profiler Data
The
purpose of this memorandum is to announce the availability of WinRiver 10.03
for collecting and processing data collected with RD Instruments (RDI) acoustic
profilers (Broadband, Rio Grande, Workhorse, and others). The WinRiver software
was developed by RDI, in cooperation with the USGS, as a part of a Cooperative
Research and Development Agreement (CRADA). The use of WinRiver software by
USGS offices was announced in Office of Surface Water Technical Memorandum
No. 2000.03.
The
WinRiver software continues to be updated and improved. Various new features
are included in this release of WinRiver and many known problems have been
fixed. Lists of new features, changes
to existing features, and bugs fixed are provided in Attachment A. At least one known bug remains in WinRiver
10.03. Attachment B documents this bug,
as well as two other bugs that appear to have been fixed. All USGS users who collect streamflow data
must upgrade to WinRiver 10.03. The
latest version of the software is available to USGS personnel who own and use
ADCP’s manufactured by RDI at http://hydroacoustics.usgs.gov/downloads/setupex.exe.
The
following sections contain information about two major enhancements in WinRiver
10.03, the addition of a configuration wizard for configuring the ADCP and the
addition of data-screening tools. In
the following sections, italicized text refers to menu options or buttons in
the WinRiver software.
The configuration wizard is a tool that can be used to set up an ADCP for data collection. The goal of the configuration wizard is to guide the user through the ADCP setup process and reduce the need to remember all direct commands. The OSW hopes that the addition of this feature will minimize some of the common mistakes that occur when setting up ADCP’s for discharge measurements.
The
configuration wizard is invoked in Acquire mode by selecting Configuration
Wizard under the Settings menu or by pressing Shift-F6. An example of the configuration wizard is
shown below.
When
the Run Wizard button is pressed, the wizard verifies that the required
fields have been entered by the user and uses these values to generate a
configuration file for the measurement.
The appropriate direct commands are specified, based on rules supplied
by the USGS to RD Instruments, including depth cell size, number of depth
cells, ambiguity velocity, and water mode.
The wizard also scales the chart properties accordingly based upon the
entered values. Although the software has been tested and we are reasonably
confident that the configuration wizard will work well, users always should
check to make sure that the direct commands are appropriate. In particular, users should pay attention to
the Configuration Wizard Warnings generated after running the wizard.
After
running the configuration wizard, pressing Save As will save the
configuration file as a WinRiver configuration file. Pressing OK will apply the wizard-generated settings to
the currently loaded configuration file.
The user then can select Save or Save As from the File
menu to save the configuration file.
After the configuration wizard window is closed, the configuration file
settings created by the wizard can be viewed in the Configuration Settings window. The direct commands generated by the
configuration wizard are listed under the Commands tab of the Configuration
Settings window in the second column labeled Wizard Commands. Wizard
Commands can override Fixed Commands and User Commands can override Wizard
Commands.
The
capability of WinRiver 10.03 to screen data in software provides a better
approach to processing data that should result in more accurate measurement of
discharge. Previously all screening was done in the firmware of the ADCP. The new data-screening switches and
threshold variables are found in the Configuration Settings window,
under the Processing tab.
The
first set of screening tools is for the acceptance or rejection of 3-beam
solutions. In the past, users determined
whether to allow 3-beam solutions using the EX command. This determination was an irreversible
decision that applied to both water-track and bottom track data that had to be
made before data collection. With this new screening tool all raw data can be
collected by setting the EX command (EX10111) to accept 3-beam solutions (which
is now the default in WinRiver) and then accepting or rejecting 3-beam
solutions in WinRiver after the user reviews the data. Three-beam solutions
often indicate that a fish or other object has interfered with one beam, which
can cause the sound reflected by that object to contaminate the neighboring
beams. Thus, simply throwing out the
bad beam and accepting a 3-beam solution can cause errors in the measured
velocity. In most situations, 3-beam solutions should be rejected and the
default setting for water-track is to reject 3-beam solutions. The cause of
3-beam solutions in bottom-track data is different and it generally is
acceptable to use 3-beam solutions for bottom tracking, and the default setting
is to accept 3-beam solutions for bottom tracking.
The
next set of screening tools allows the rejection of data in individual bins
based on the error velocity and (or) vertical velocity in that bin. Ambiguity errors and areas of highly
non-homogeneous flow cause high values in the error and vertical velocity
magnitudes. Contour plots of the error or vertical velocity with the
appropriate color scale can be used to identify data that obviously are
different from the surrounding data.
Likewise, time-series plots of the boat speed or water speed, and the
stick ship-track plot, can show unreasonable spikes. Prior to the release of
this version, no mechanism was available for removing data containing obvious
ambiguity errors or data with high error velocities because of non-homogeneous
flow. By appropriately setting the
error and vertical velocity thresholds, WinRiver 10.03 will mark these data as
invalid and compute discharge in these areas from neighboring valid data. By
default, the thresholds are set very high, which for practical purposes turns
off the screening. Reasonable values for the error velocity thresholds are a
function of the standard deviation of the velocity measurement in a single bin,
which is dependent on the frequency of the ADCP, the bin size, and the
ambiguity velocity setting for mode 1 data. The objective of screening is to
reject only the obviously bad data and because the errors are normally
distributed, a threshold of 3 to 4 times the standard deviation of the velocity
measurement is reasonable (see table 1).
Table
1. Reasonable error velocity thresholds.
[WM
refers to the WM command used to set the water mode in RDI profilers. WS refers to the WS command used to set the
depth cell size in RDI profilers.]
ADCP
Frequency (kHz) |
Water Mode [WM] |
Depth Cell Size (m) [WS] |
Standard Deviation[1]
(ft/sec) |
Reasonable Error Velocity
Threshold (ft/sec) |
300 |
1 |
1.00 |
0.59 |
1.8 |
5 |
0.20 |
0.03 |
0.1 |
|
8 |
0.20 |
0.86 |
2.5 |
|
600 |
1 |
0.50 |
0.59 |
1.8 |
5 |
0.10 |
0.03 |
0.1 |
|
8 |
0.10 |
0.86 |
2.5 |
|
1200 |
1 |
0.25 |
0.59 |
1.8 |
5 |
0.10 |
0.02 |
0.1 |
|
8 |
0.10 |
0.81 |
2.5 |
Setting
the threshold for the vertical velocity is not as straightforward as the error
velocity threshold. The vertical velocity threshold only should be used if the
user has a reasonable knowledge of the vertical velocities in the measurement
section and (or) observes patterns or banding in the vertical velocity contour
plot. The vertical velocity threshold
then can be set based on the user’s judgment to remove obviously erroneous
data.
The
last screening tool allows users to apply a fish-detection algorithm in the
WinRiver software. When using WinRiver,
by default, a firmware fish detection algorithm is implemented by sending the
direct command, WA50. However, in the
WinRiver software users now can select different values for intensity amplitude
for detecting and removing data that potentially are contaminated by echoes
from fish. This feature has the
advantage of collecting all the raw data and screening the data in
post-processing rather than having the instrument reject the data based on the
configuration setting.
Because
screening data by use of these thresholds is a new feature, we expect that
guidance regarding their use will change as we gain experience. Additional information regarding the use of
thresholds will be made available through the OSW ADCP Web pages (http://hydroacoustics.usgs.gov/).
J.
Michael Norris
Acting
Chief, Office of Surface Water
Attachments (2)
Distribution Code: A, B, FO, PO
Changes made to existing features
[1] The standard deviation for mode 1 measurements shown in this table are computed assuming a mode 1 ambiguity velocity (WV) of 170 centimeters per second.