1. SonTek FlowTracker® Basics of Operation
AN introduction to discharge measurements with a flowtracker
GEOTech Environmental Equipment training center
Denver, Colorado
April 23rd-25th, 2013

2. FlowTracker Handheld ADV®
Presentation Outline
History of the FlowTracker
Basic FlowTracker Operating Principles
Firmware Features Overview
Discharge Uncertainty

3. History of the FlowTracker
A direct descendant of the SonTek ADV® - invented in 1993 and used for hydraulic research in laboratories
ADVs feature:
Extremely high velocity precision based on pulse-coherent Doppler processing
Full 3D velocity measurement
25 Hz sampling
0.25 cc sampling volume remote from instrument
4 (manual) velocity range settings
Outstanding shallow water and low- velocity capability
1000+ systems in laboratory use today

4. History of the FlowTracker
Low power CMOS-based electronics platform (late ’90s) provided mechanism for compact packaging and battery operation
AutoVelocity processing (invented in 2000) eliminated need for separate velocity range scales – facilitated adaptation for field use
Market demand to make a tool for discharge measurement in shallow streams

5. History of the FlowTracker
Assigned as a USGS ITASS project to adapt ADV for use as a discharge measurement instrument from wading rods
Some funding and cooperation from USGS Indiana (Scott Morlock) and Maryland Office (Gary Fisher)
FlowTracker prototype being tested against Price AA - ~2000
Traditional ADV attached to wading rod as proof of concept
Mention of the USGS does not constitute endorsement

6. History of the FlowTracker
History of upgrades and improvements:
First production release (2001)
High sensitivity receiver improvements (2002)
3 m probe cable option (2003)
Deluxe two-piece wading rod and case (2005)
Software v (2009)
Firmware v (2012)

7. Basic Operating Principles
ADVs are classified as Bi-static Doppler Current Meters
Central acoustic transmitter
2 or 3 acoustic receivers
For 2D or 3D probes
Remote sampling volume
10 cm from probe tip
2D or 3D velocity
Full resolution of flow direction
Acoustic
transmitter
Acoustic
Receiver 2
Acoustic
Receiver 1
Sampling Volume
6mm dia
by 9mm height
Sampling volume is fixed at 10 cm from transmitter

8. ADV Doppler Processing
Pulse coherent processing
Best performance of any Doppler technique
Fast response time
Wide velocity range
Automatic velocity range
Adapts operation based on flow speeds
Excellent performance for flows to (4.5 m/s / 15 ft/s)
Unbeatable low flow performance
Flows less than (1 cm/s / 0.03 ft/s)
Accuracy 1% of measured velocity in 1 second

9. FlowTracker Doppler Calibration
Velocity Depends on 3 Factors
Measured Doppler shift
Fundamental to design of Doppler processing algorithms
Never changes
Probe geometry
Factory calibration for each probe
Calibration can only change with physical damage
Diagnostic software (ADVCheck/Beam Check) easily verifies probe integrity
Sound speed
Internal temperature sensor for automatic compensation
User input salinity for salt or brackish water applications

10. Basic Operating Principles
Pulse coherent processing
Transmit 2 pulses separated by time TLAG
Measure phase of return signal
Phase change divided by TLAG gives Doppler shift
Maximum velocity determined by TLAG
Shorter TLAG gives higher maximum velocities
Velocity noise level changes with TLAG
Longer TLAG gives lower noise levels
Adjust TLAG to give sufficient maximum velocity and lowest noise level

11. Basic Operating Principles
Traditional ADV velocity range
User selects fixed velocity range to set TLAG
Requires knowledge of flow conditions to optimize performance
Automatic velocity range
Uses multiple pulse pairs with different TLAG settings
Based on results from one TLAG the algorithm determines if the next TLAG can be used without error
Uses the longest possible TLAG for given flow conditions, giving the lowest possible noise in velocity data
Maximum velocity is 4.5 m/s
Optimizes performance from less than 0.1 cm/s to 4.5 m/s

12. Adaptation of ADV into FlowTracker product for discharge applications
Handheld controller
AA Battery power
Readily fixes to top-setting wading rods
2m (3m optional) probe cable
2D ADV probe (3D optional)

13. External Power/Communication Connector
System Configuration
Handheld controller
Custom keypad
LCD display
Batteries (8 AA)
4 Mb Data recorder
Complete data collection and discharge software
LCD Screen
Keypad
Probe Cable
External Power/Communication Connector
FlowTracker Probe
Three probe types:
2D side looking
2D/3D side looking
3D down looking

14. Battery Requirements Operates from 8 AA batteries Typical battery life
Alkaline, NiMH or NiCad
Standard rechargeable batteries can be used (user supplied or available from SonTek)
Typical battery life
Alkaline: 25 hours continuous operation
NiMH: 15 hours continuous operation
NiCad: 7 hours continuous operation
Monitoring battery capacity
Access battery voltage from keypad interface
Shows estimated remaining capacity for all battery types

15. System Specifications
Probe configuration
2D side looking standard
2D/3D side looking and 3D down looking optional
Additional sensor
Temperature (±0.1° C / ±0.2° F)
Environmental
Operating temperature (-20° to 50° C / 0° to 120° F)
Storage temperature (-20° to 50° C / 0° to 120° F)
Data Quality Annunciation
Acoustic signal strength as SNR (signal to noise ratio)

16. Case Study --Tow Carriage Testing
Velocity offset: (0.15 cm/s / ft/s)
Performance exceeds 1% accuracy specification
Total of 71 tow carriage runs
Flow angles to ±40°
Best fit slope: *Cart speed

17. Typical Field Velocity Data
Raw Velocity Data
Typical Field Velocity Data
Sample field data
Mean velocity (45.7 cm/s / 1.50 ft/s)
Standard deviation of 1 second data (3.7 cm/s / 0.12 ft/s) – about 8% of mean velocity
Fluctuations are real variations in velocity
Standard error of 40 second average (0.6 cm/s / 0.02 ft/s)

18. Averaging Time FlowTracker burst sampling
Collects fixed length record of velocity at each station
User specified averaging time from 10 to 1000 seconds
Recorded data includes raw 1 second velocity, mean velocity, temperature, and extensive quality control data
Specifying averaging time
Function of the environment
FlowTracker provides standard error data in real time to determine the accurate of mean velocity data

19. Basic Operating Principles
Menu driven interface for discharge measurement
Specify setup parameters
Perform system diagnostics
Data collection run
“Set” keys to specify station location, water depth, measurement method, etc.
Measure to start data collection
Quality control data shown with each measurement (measurements can be repeated if desired)
Next and Previous Station keys to scroll through completed station data

20. Basic Operating Principles
FlowTracker Handheld-ADV technique when used with a wading Rod and tag line
Wading rod must be held perpendicular to tag line
FlowTracker reports true angle of flow – no estimation required

21. Midsection Method (ISO)
Firmware Features
Mid-Section Method
Midsection Method (ISO)

22. Firmware Features
Mean-Section Method

23. Firmware Features
Japanese Method

24. Firmware Features
0.6
0.2/0.8
0.2/0.6/0.8
Arrows denote ADV probe positions on wading rod

25. Firmware Features Kreps 2 5-point Multipoint
Arrows denote ADV probe positions on wading rod

26. Velocity Methods Equation: how stations are combined for discharge
Method: how mean station velocity is determined
Supported methods: 16 total
Displayed methods
Can remove those that will never be used
Special cases
Method NONE
Internal islands
No measurement possible, use adjacent stations
Method INPUT V
No measurement possible, user input velocity
Method MULTI PT
Any number of measurements at any locations
Integrated mean velocity

27. Velocity Methods Method Measurement Locations Mean Velocity Equation
0.6
0.6 * depth
Vmean = V0.6
0.2/0.8
0.8/0.2
0.2 / 0.8 * depth
Vmean = (V V0.8) / 2
.2/.6/.8
.8/.6/.2
0.2 / 0.6 / 0.8 * depth
Vmean = (V *V0.6 + V0.8) / 4
Ice 0.6
0.6 * effective depth
Vmean = 0.92*V0.6
Ice 0.5
0.5 * effective depth
Vmean = 0.89*V0.5
Ice 2/8
Ice 8/2
0.2 / 0.8 * effective depth
Kreps 2+
Kreps 2-
0.0 (near surface)
0.62 * depth
Vmean = 0.31*V *V0.62
5 Point+
5 Point-
0.2 / 0.6 / 0.8 * depth
1.0 (near bottom)
Vmean =
(V *V *V *V0.8+ V1.0) / 10
Multi Pt
Any number of points at user specified depths
Integrated velocity average

28. Smart QC What is the goal? Best possible discharge measurement
What is needed to do this?
Verify instrument operation
Evaluate all data used for discharge calculation
Automatic warnings for a variety of QC data
How well did it work?
Discharge uncertainty

29. Is the FlowTracker working properly?
Smart QC
Is the FlowTracker working properly?
Auto QC Test
Run at the start of every measurement.
Place the probe in moving water, well away from any underwater obstacles
Takes <30 seconds, data analyzed automatically
Automated version of BeamCheck PC software

30. Four Test Parameters Noise level SNR
Smart QC
Four Test Parameters
Noise level
SNR
Peak location
Peak shape

31. Smart QC Why run BeamCheck from a PC
if you do an Auto QC with each measurement?
Running BeamCheck manually is one of the best ways to learn about the FlowTracker
BeamCheck allows subjective evaluation, particularly valuable for the peak shape criteria
Experienced users can use the BeamCheck in the software to achieve the same results as the FlowTracker Auto QC plots
Beam Check is found in SonUtils4

32. Evaluates all data used for discharge calculation
Smart QC
Evaluates all data used for discharge calculation
Data entry
Location
Consistent spacing, stations entered in order
Depth
No radical changes in depth
Measurement practices %Q
Is any one section greater than 10% of total discharge?
Flow Angle
Is flow angle > 20°?
May be real at some measurement sites

33. Smart QC SNR (Signal-to-Noise Ratio)
Function of how much suspended material is in the water
Must be greater than 4 dB for reliable operation
Both beams should be roughly the same (<10dB difference)
Most stations in a cross section should be about the same (<10 dB different from mean SNR for all stations)
Standard deviation of SNR > 5 dB
Spikes
All acoustic systems will see some spikes in velocity data
These are automatically filtered out
Large numbers of spikes (> 10% of samples) is a problem
Aerated water
Interference from submerged object

34. Smart QC σV : Standard error of velocity
Measures variation in velocity during the averaging period
Varies with the environment
Higher in high velocity or highly turbulent environment
SmartQC automatically adjusts criteria to the environment
σV threshold: largest of following 3 values
Fixed value (0.01 m/s / 0.03 ft/s)
Mean of all stations
Percentage of velocity
Unusually high σV could be
Aerated water
Interference from underwater obstacle
Real for localized turbulence

35. Smart QC QC data reviewed and warnings issued
At the end of each measurement
When End Section pressed (all measurements reviewed)
All criteria can be modified or disabled
Data entry errors
Location
Station spacing changes dramatically
Station location is out of order
Depth
Large depth change compared to adjacent stations
Boundary QC
Warning before data collection if FAIR or POOR
Reposition probe and try again

36. Smart QC Boundary QC Acoustic interference from underwater obstacles
If possible, FlowTracker adapts operation to avoid interference
If a warning given, re-position probe and try again

37. Smart QC Smart QC Summary For most warnings If problem persists
Evaluate probe location, consider moving probe
Repeat measurement
If problem persists
Check probe operation (Auto QC, BeamCheck)
Maybe there is not a problem
Higher turbulence
High measurement angle
Aerated water
Locally high sediment load

38. Auto QC Test Auto QC Test = BeamCheck In Firmware
Prompted at the start of every file
Takes <30 seconds
Can be run at any other time, open data file or not
Procedure
Put probe in moving water and press start
Automatic Analysis
Noise, SNR, peak shape, peak level,
What to do if warnings?
Try again, possibly different probe location
Run BeamCheck on PC

39. QC Menu Supplemental data QC settings Discharge settings
The QC menu is active essentially anytime during a measurement
Supplemental data
QC settings
Discharge settings
Change averaging time
Raw velocity display
Auto QC test

40. Setting QC Criteria Setup Parameters Menu QC Settings
QC Settings Menu (Option 4)
Discharge Settings Menu (Option 5)
Also accessible from QC Menu
Set any value to 0 to disable
QC Settings
SNR Threshold (dB)
σV Threshold (m/s or ft/s)
Spike Threshold (%)
Discharge Settings
Max Section Discharge (%)
Max Depth Change (% of comparison depth value)
Max Location Change (% change in spacing)
Max Velocity Angle (°)

41. Additional Firmware Features
Data export
BeamCheck
Recorder download
International language support
Windows 2000/XP/Vista compatibility

42. Software: Data Export HTML discharge report Open / view multiple files
Batch processing options
Modifications to DIS file for new features and better integration with databases
Column headers (SUM, DIS)

43. Discharge Report: Page 1-2

44. Discharge Report: Page 1-3
HTML-based reporting
Fully customizable
Easily translates to 7 different languages

45. Discharge Uncertainty
Background
Two Uncertainty Calculations
ISO Uncertainty Calculation
Statistical Uncertainty Calculation
Why 2 Calculations? Comparison of Results
Displaying Uncertainty Results

46. Discharge Uncertainty
What is uncertainty?
How accurate is the discharge measurement
Most agencies record a subjective measurement quality
This provides a quantitative measure
Why does uncertainty matter?
Data are used for other analysis such as Stage/Discharge or Velocity Index ratings.
Uncertainty propagates through these ratings.

47. Discharge Uncertainty
ISO Standard 748
International standard for discharge measurement procedures
Includes a discharge uncertainty calculation
Estimated uncertainty for all measured variables (depth, width, velocity)
Other sources of uncertainty
Limited number of stations across the river
Velocity at limited number of depths at each station
Above values are based on extensive studies of many different rivers
Well documented formula using readily available data
Similar method and results to Sauer and Meyer (USGS)

48. Discharge Uncertainty
Statistical Uncertainty Calculation:
Developed by researchers at the U.S. Geological Survey
Tim Cohn, Julie Kiang, and Robert Mason
New technique, limited publications and field experience
Also called “Interpolated Variance Estimated (IVE) method
Fundamentally Different Approach:
ISO: Physical properties of the measurement and calculation
Statistical: Use adjacent measurements to estimate uncertainty

49. Discharge Uncertainty
Statistical Uncertainty Calculation
Discharge calculation assumes linear change in depth/velocity between stations
Estimate data at each vertical by interpolating adjacent data
Uncertainty is measured value minus estimated value

50. Discharge Uncertainty
24 Measurements
Discharge to 9 m3/s
Velocity 0.01 to 0.5 m/s
ISO
2.4 to 8.4% (all files)
2.4 to 4.3% (removing 1 file)
Statistical
2.1 to 19% (all files)
2.1 to 15.2% (removing 1 file)
ISO Uncertainty
Statistical Uncertainty (%)

51. Discharge Uncertainty
Why 2 Different Calculations?
ISO calculation
International standard, well documented
Critical parts from statistical averages from many different rivers
Not responsive to varying data from a specific site
Statistical method
Appears very effective
Responds to conditions and variations in data
Limited publication and field experience

52. Discharge Uncertainty
Software (Post-Processing)
Overall Uncertainty
Contribution of Each Source
TotalQ m3/s RatedQ m3/s
Difference -2.8%
0=Exit or Enter=More
Firmware (Real-Time)
Overall Uncertainty
Largest Source
Q Uncertainty 3.1% Largest Source
Velocity
0=Exit or Enter=More

53. Questions?