|
publications > poster > tools for technicians; lessons learned in index-velocity site selection and ADVM setup
Tools for Technicians; Lessons Learned in Index-Velocity Site Selection and ADVM SetupStephen Huddleston, Mark Zucker, Jeff Woods, Carrie Boudreau, Shane Ploos, and Christian Lopez Introduction
Hydrologists and engineers have spent countless hours and resources developing hydroacoustic instrumentation and methods, their efforts are lost if the technology is applied incorrectly. A technical focus is needed to ensure high quality data collection with ADVMs and ADCPs. What good is all the theory, research, and development packed into the head of an acoustic transducer if you are measuring the velocity of tree roots on the stream bank? How can we ensure that high quality data are collected and minimize uncertainty? Reviewing multiple index-velocity sites have brought up a few critical issues that are solved by simply reviewing data provided by acoustic instruments and making appropriate adjustments; some of these issues are presented here. ADVM Site SelectionSite selection is probably the single most important aspect in the implementation of an ADVM. Many issues can be avoided by simply constructing a station in a suitable site for gaging. Although one may be constrained by the goals of a project, it is important to select the "best" location possible before installing instrumentation. In order to develop an accurate index velocity rating an understanding of stream dynamics is important prior to site implementation. Ultimately a straight channeled cross section with uniform flow is ideal. In tidal locations flow reversals will occur and velocity may be horizontally or vertically stratified. Reconnaissance is crucial to determine flow characteristics at any given location. Although it may be impossible to catch all conditions present, a series of measurements will provide a general idea of what is going on in the channel. It is important to continuously monitor data (salinity if applicable and temperature profiles) as this may indicate the presence of stratification or bi-directional flow. Depending on site selection and instrumentation it may be possible to minimize the impact of stratification on your velocity index. In shallow streams and rivers be sure to account for water level changes when installing side looking ADVMs to preserve an appropriate aspect ratio (range / distance to nearest boundary). Aspect ratios of 15-20 are noted as being acceptable by SonTek (Argonaut-SL System Manual, 2007) however the office of surface water (OSW) suggests a conservative maximum of 10 (OSW Index-Velocity Quick Sheet; ADVM Cell Size). Shown in figure 1A-D are screenshots of velocity vectors at various depths as well as a projected velocity contour from ADCP measurements conducted at North River (022908205) in Monroe County, Florida on May 15, 2007. The graphs clearly indicate the presence of both vertical and horizontal flow patterns observed with the ADCP. The width of North River is roughly 275 ft while the depth ranges from roughly 4 to 7 feet. The current ADVM set up allows for a sampling volume of 42.65 ft. Given these constraints, no single ADVM set up is able to capture an index velocity that is truly representative of flow. While it may be possible to install a second ADVM for multi-parameter rating development, the costs, maintenance, and research permitting must be considered. Remember, reconnaissance is key! Instrument DiagnosticsOnce an ADVM is installed, various diagnostic tools are available to evaluate the quality of data being collected. If using SonTek ADVMs, advanced user settings such as powerping are available. Powerping data is an internal diagnostic parameter collected every 100 samples which provides a profile of signal strength for each beam versus the range of the system (Argonaut-SL System Manual, 2007). Diagnostics plots are accessed in ViewArgonaut Processing Menu with a Ctrl+D keystroke. Once an acoustic pulse is sent from the transducer head, sound will be scattered and signal strength will decrease with increased distance. In the diagnostic profile, a theoretical signal decay curve is shown with a gray line. Given a clear path signal strength for each beam should follow the decay curve. Spikes in signal strength can indicate boundary interference such as trees/roots, the stream bed, stream surface, stream bank, or obstacles along the stream bottom which will affect the quality of data collected. If using SonTek software the instrument cell end will be given by a vertical blue dashed line in the diagnostic profile. OSW guidelines recommend that the cell end be at least 0.10*D from any boundary where D is the distance from the ADVM to a boundary (OSW Index-Velocity Quick Sheet; ADVM Cell Size). Presented below are two diagnostic plots from two South Florida streams. First, a powerping profile of Shark River (252230081021300) is given (Fig. 2A) showing spikes in the signal strength around 25 feet indicating that the beams are being obstructed or reflecting off of the surface and bottom. Reflection off of the opposite bank is not possible as it is far beyond the range of the ADVM installed, a SonTek SL 1500 kHz side looking ADVM. Multi-cell data collected at the same time as the powerping diagnostic highlights the effect of the interference on the velocity data (Fig. 2B). Clearly cells 3 and 4 are affected by the interference in the acoustic path. This example illustrates an aspect ratio that is too large. In this situation the average mean low water level could have been used when setting the depth of the ADVM to preserve an acceptable aspect ratio and avoid the boundary interference encountered. In the second example, a powerping profile from East Creek (251152080370900) is given (Fig. 2C) and indicates a properly deployed ADVM. Spikes at around 17 and 20 feet indicate the presence of the opposite bank. In this diagnostic profile, the cell end is around 15 feet shown by the vertical dashed blue line; an acceptable cell end given the distance to the opposite bank.
Velocity InterferenceThe effect of correcting for velocity interference on the index-velocity rating at Shark River (252230081021300) is discussed below (Fig. 3A and Fig. 3B). Two index-velocity ratings were developed to demonstrate the importance of data review, field observations, ADVM program adjustments, and the effects they can have on an index-velocity rating performance. The first rating example utilized 120 discharge measurements collected between April of 2004 and March of 2008 using the average of all four velocity cells (CE=14 meters including those with boundary interference) for rating development (Fig. 3A). The second rating example utilized the same set of 120 measurements using only the average of velocity cells 1 and 2 (CE=5.6 meters) where no interference was in the path of the acoustic beams (Fig. 3B). This adjustment resulted in a higher R2 value (0.93 vs. 0.98) and a lower standard error (0.28 vs. 0.13) for the rating using Cells 1 and 2 only. ADCP measurements collected on May 12, 2004 resulted in a percent difference of roughly 71% using the all cells rating as compared to a percent difference of roughly 29.5% for the rating using only cells one and two (Fig. 3A and 3B). We acknowledge that additional ADCP measurements are needed to improve index-rating and further document the relevant flow characteristics. We speculate the flow from a tributary stream located upstream of the ADVM as well as differences in the flow dynamics of the incoming and outgoing tide may be inducing bias as well. Further investigation is warranted to quantify the effects of the tributary stream including the development of a break rating for the Shark River index site. A portion of the ADVM setup is provided below showing multi-cell settings for the Shark River ADVM.
Selecting an Bottom Mounted over a Side Looking ADVMAn example of a projected velocity contour plot from an ADCP discharge measurement collected at Card Sound Canal (CSC, 251181608232200) in southeast Miami-Dade County in Florida (Fig. 4A) is provided to highlight the presence of bi-directional flow. From the surface to approximately 2.5 feet positive flow is observed. From 2.5 feet to the bottom of the channel slightly negative to near zero flow is observed. Bi-directional flow is encountered regularly at this site due tidal influences from nearby Barnes Sound and Little Card Sound as well as head differences between the upstream wetland and the coast. From October 1, 2003 to November 28, 2004 a side looking ADVM was deployed on the sloping right bank of the canal with limited success. With frequent stratified flow, salinity, and temperature, the side-looking ADVM was unable to capture a representative index of mean channel velocity (Fig. 4C). On November 29, 2004 a bottom mounted ADVM (Fig. 4B) was installed resulting in an improved index-rating for CSC (Fig. 4D). For additional information on the use of bottom mounted ADVMs please refer to the following poster located on the Coastal Gradients Project web page (http://sofia.usgs.gov/publications/posters/nut_loadCSC/). To reiterate, a series of reconnaissance measurements may assisted the streamgager on selecting a bottom mounted ADVM rather than a side-looking ADVM.
For more acoustic tips, see Jeff Woods' poster, "Salvaging Sontek ADVM Data Using a One Beam Solution for a Bad Transducer." Thanks to Carrie Boudreau for creating the rating plots for Card Sound Canal Any use of trade, product, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government
Poster displayed at 2008 National Data Conference, sponsored by CHIDER, Tunica, MS June 16-20, 2008 Related information: SOFIA Project: Freshwater Flows to Northeastern Florida Bay |
||||||||||||||||||||||||||||||
U.S. Department of the Interior, U.S. Geological Survey
This page is: http://sofia.usgs.gov/publications/posters/tools_tech/index.html
Comments and suggestions? Contact: Heather Henkel - Webmaster
Last updated: 25 November, 2008 @ 04:53 PM (KP)
