Dual Use of a Sediment Mixing Tank for Calibrating Acoustic Backscatter and Direct Doppler Measurement of Settling Velocity (and Related Field Motivation and Observations) Grace Cartwright, Carl Friedrichs, and Paul Panetta Outline of Presentation: Motivation: Acoustic Doppler Velocimeter (ADV) Field Observations Sediment Mixing Tank: Acoustic Backscatter Calibrations Sediment Mixing Tank: Doppler Settling Measurements Independent field-based test of ADV-measured settling velocity
Motivation: Determine fundamental controls on sediment settling velocity and bed erodibility in muddy estuaries (X-rays courtesy of L. Schaffner) Study site: York River Estuary, VA This project aimed to determine the fundamental controls on settling velocity and bed erodibilty in muddy estuaries. The study area is the York River estuary in southeastern VA,. Previous work has shown that a physical-biological gradient is found along the York a great place to study the study the relationship between biology and overlaying hydrodynamics. In the upper to middle York, near the Clay Bank or the MUDBED Intermediate site, physical disturbances reduce macrobenthic activity and sediment layering is preserved (x-radiograph A and B in Figure 2.1). Biology abundance increases towards the mouth, so in the lower York, near Gloucester Point or the MUDBED Biological Site, bioturbation destroys physical layering (x-radiograph C in Figure 2.1) (Cartwright et al., 2011; Cartwright et al., 2009; Dickhudt et al. 2009; Friedrichs, 2009). Since 2006 ADV tripods have been deployed at both the Intermediate and Biologally dominated sites Physical-biological gradient found along the York estuary : -- In the middle to upper York River estuary, macrobenthos are seasonally overwhelmed by the floc-rich estuarine turbidity maximum, and sediment layering is often preserved. (e.g., Clay Bank – “Intermediate Site”) -- In the lower York and neighboring Chesapeake Bay, layering is generally destroyed by bioturbation, and abundant macrobenthos pelletize the mud as they feed (e.g., Gloucester Point – “Biological Site”) -- Acoustic Doppler Velocimeter (ADV) tripods provide long-term observations within a strong physical-biological gradient.
(Photos by C. Cartwright) Observations provided by a Sontek 5 MHz Acoustic Doppler Velocimeter (ADV) Sensing volume ~ 35 cmab ADV after retrieval ADV at deployment Suspended mass concentration (c) from acoustic backscatter when calibrated by pump samples. Bed stress (τb) based on burst-averaged near-bed Reynolds stress via τb = ρ <u’w’> ;estimating the reynolds flux from the velocity data (where ρ = water density, u’ & w’ = turbulent fluctuations in horizontal & vertical velocity). Bulk settling velocity (wsBULK ) Assume upward flux=downward gravitational settling Via a balance between upward settling flux and gravitational settling The proportionality between eroded mass and bottoms stress estimates the Depth-limited bed erodibility (ε) Effective Settling Velocity: (Photos by C. Cartwright) -- ADVs often provide quality long-term data sets despite extensive biofouling. -- ADVs can provide continual long-term estimates of: 3-D velocity (u,v,w) in ~ 1 cm3 sampling volume including ~ Hz turbulent fluctuations (u’,v’,w’) Suspended mass concentration (c) from acoustic backscatter including turbulent fluctuations (c’) Turbulent Reynolds Shear Stress, τ = ρ*<u’w’> Sediment Settling Velocity, ws = <w’c’>/c Elevation of seabed relative to tripod
Concentration Calibration Curves Significant scatter in suspended solids vs. acoustic backscatter relationship because of variations in suspended particle size, particle density, and response of individual ADVs Concentration Calibration Curves In-situ pump samples analyzed for total suspended solids Concentrations used to calibrate acoustic backscatter from deployed tripods All observations utilize Sontek 5 MHz ADV “Ocean” Model (Cartwright et al., 2009)
Lower Concentration Period at Biological (Gloucester Point) site (ADV height ~ 35 cm) Current (cm/sec) ~ 40 cm/s ~ 50 mg/l TSS (mg/liter) ~ 4 cm change Bed elev (cm) Days since December 4, 2006 (Cartwright et al. 2009)
<w'C'> = ws <C> Assume: (Fugate & Friedrichs, 2003) Upwards turbulent sediment flux Downwards gravitational settling = <w'C'> = ws <C> w' = vertical turbulent velocity, C' = turbulent concentration fluctuation < > = burst average, ws = sediment settling velocity, <C> = burst-average TSS Biological site (GP) ADV Data: <w'C'> (mm/s)(mg/liter) Slope = ws = <w’C’>/<C> = 1.5 mm/s <C> (mg/liter) -- Insensitive to ADV calibration for C , 50% change in calibration = 10% change in ws. (Cartwright et al. 2009)
Higher Concentration Period at Intermediate (Clay Bank) site (ADV height ~ 35 cm) ~ 40 cm/s Current (cm/sec) ~ 100 mg/l TSS (mg/liter) ~ 20 cm change Bed elev (cm) Days since February 27, 2007 (Cartwright et al. 2009)
Higher Concentration Period at Intermediate (Clay Bank) site (ADV height ~ 35 cm) ws = 0.77 mm/s ws = 0.55 mm/s ws = 0.20 mm/s Current (cm/sec) Current (cm/sec) <w'C'> vs. <C> <w'C'> vs. <C> <w'C'> vs. <C> TSS (mg/liter) ws = 0.80 mm/s Bed elev (cm) <w’C’> vs. <C> Days since February 27, 2007 (Cartwright et al. 2009)
Biological site Intermediate site Intermediate site Biological site Seasonal Variability in sediment settling velocity (ws ) and erodibility (e) is observed at the Intermediate Site. Biological site Ws ~ 1.5 mm/s Intermediate site Ws varies from ~ 0.5 mm/s to ~ 1 mm/s 3- day Mean Ws from fits to <w’C'> = ws<C> using ADVs 2 1.5 1.0 0.5 Ws (mm/s) 3-day mean of erodibility (e) using ADVs and Gust erosion chambers 6 5 Long term data for bulk settling velocity and erodibilty from ADV’s shows seasonally variability in Ws and erod at the Intermediate site. It looks like we have two separate Regime occuring here. Top figure shows Ws—noisy but stable and higher at biological site, at the intermediate site. Settling velocity seems to go from ~ 0.5 mm/s during this Regime 1 up to Regime 2 Bottom figure shows erodibility- again we see a stable value at biological site-this times its lower And now we see higher erodibility corresponding to the lower Ws during regime 1 and lower erodibility corresponding to lower settling velocities during Regime 2. So we wanted to figure out why? 4 Intermediate site ε varies from ~ 3 kg/m2/Pa (Regime 1) to ~ 1 kg/m2/Pa (Regime 2) ε (kg/m2/Pa) 3 2 Biological site Generally < 1 kg/m2/Pa 1 Cartwright et al., 2009
-- Can controlled lab experiments provide insight? Seasonal Variability in sediment settling velocity (ws ) and erodibility (e) is observed at the Intermediate Site. Biological site Ws ~ 1.5 mm/s Intermediate site Ws varies from ~ 0.5 mm/s to ~ 1 mm/s 3- day Mean Ws from fits to <w’C'> = ws<C> using ADVs 2 1.5 1.0 0.5 Ws (mm/s) Questions: -- Are these ADV-based estimates of sediment concentration and settling velocity accurate and reliable? -- Can controlled lab experiments provide insight? 3-day mean of erodibility (e) using ADVs and Gust erosion chambers 6 5 Long term data for bulk settling velocity and erodibilty from ADV’s shows seasonally variability in Ws and erod at the Intermediate site. It looks like we have two separate Regime occuring here. Top figure shows Ws—noisy but stable and higher at biological site, at the intermediate site. Settling velocity seems to go from ~ 0.5 mm/s during this Regime 1 up to Regime 2 Bottom figure shows erodibility- again we see a stable value at biological site-this times its lower And now we see higher erodibility corresponding to the lower Ws during regime 1 and lower erodibility corresponding to lower settling velocities during Regime 2. So we wanted to figure out why? 4 Intermediate site ε varies from ~ 3 kg/m2/Pa (Regime 1) to ~ 1 kg/m2/Pa (Regime 2) ε (kg/m2/Pa) 3 2 Biological site Generally < 1 kg/m2/Pa 1 Cartwright et al., 2009
Dual Use of a Sediment Mixing Tank for Calibrating Acoustic Backscatter and Direct Doppler Measurement of Settling Velocity (and Related Field Motivation and Observations) Grace Cartwright, Carl Friedrichs, and Paul Panetta Outline of Presentation: Motivation: Acoustic Doppler Velocimeter (ADV) Field Observations Sediment Mixing Tank: Acoustic Backscatter Calibrations Sediment Mixing Tank: Doppler Settling Measurements Independent field-based test of ADV-measured settling velocity
(a) VIMS Sediment Mixing Tank, with suspended sampling tubes highlighted, (b) example placement of ADV in chamber, with pump circulation outlets highlighted (Photos by C. Cartwright) Acrylic portion of tank is 32 cm x 32 cm x 1.5 m, with bottom 0.5 m tapering in toward the 44 liter/minute pump inlet. Flow is returned 25 cm below the top of the tank through 4 circulation outlets. Sediment concentration is sampled at 1 m/s through sampling tubes.
(Photos by C. Cartwright) Lab calibration with quartz sand: -- Commercially available, predominantly quartz sand was divided into size classes using sieves with mesh diameters of 63, 75, 90, 106, 125, and 150 microns. Example of 63 micron component Example of 125 micron component (Photos by C. Cartwright)
Sediment Mixing Tank: Acoustic Backscatter Calibrations (With ADV mounted mid-chamber) Quartz Sand Only calibrations (For each phi size) A successive series of sand was added to the chamber Acoustic backscatter collected for 10 minutes (10Hz) and averaged. For each concentration a water sample was pulled from mid-chamber Samples were dried and weighed for suspended mass concentration. Mixed Sediment and Mud Only calibrations (For three sites) (This section of experiment part of Newbill, 2010) Analysis methodology similar to Sand Only calibrations Did calibrations with natural muddy sediment collected from 3 sites Claybank (CB) Channel (~1 % sand) Claybank (CB) Shoal (~20 % sand) Ferry Point ( FP) Shoal (~10% sand) Repeated calibrations with Mud Only portions <63 microns Claybank (CB) Shoal Ferry Point ( FP) Shoal Sampling Tubes (Photo by C. Cartwright)
Sediment Mixing Tank: Acoustic Backscatter Calibrations (With ADV mounted mid-chamber) Quartz Sand Only calibrations (For each phi size) A successive series of sand was added to the chamber Acoustic backscatter collected for 10 minutes (10Hz) and averaged. For each concentration a water sample was pulled from mid-chamber Samples were dried and weighed for suspended mass concentration. Mixed Sediment and Mud Only calibrations (For three sites) (This section of experiment also part of Newbill, 2010) Analysis methodology similar to Sand Only calibrations Did calibrations with natural muddy sediment collected from 3 sites Claybank (CB) Channel (~1 % sand) Claybank (CB) Shoal (~20 % sand) Ferry Point ( FP) Shoal (~10% sand) Repeated calibrations with Mud Only portions <63 microns Claybank (CB) Shoal Ferry Point ( FP) Shoal Sampling Tubes (Photo by C. Cartwright)
Sediment Mass Concentration (log10 mg/liter) Mixed Mud & Sand Sediment Mass Concentration (log10 mg/liter) Mud only Solid Lines Sand Only ADV Backscatter (counts) -- ADV backscatter systematically increases with concentration for any one sediment type. -- For quartz sand, ADV backscatter systematically increases with sand grain diameter. -- Natural mud responds less strongly than sand, but response is site specific. -- Mud + Sand is also site specific and doesn’t respond as a sum or average of the two.
Dual Use of a Sediment Mixing Tank for Calibrating Acoustic Backscatter and Direct Doppler Measurement of Settling Velocity (and Related Field Motivation and Observations) Grace Cartwright, Carl Friedrichs, and Paul Panetta Outline of Presentation: Motivation: Acoustic Doppler Velocimeter (ADV) Field Observations Sediment Mixing Tank: Acoustic Backscatter Calibrations Sediment Mixing Tank: Doppler Settling Measurements Independent field-based test of ADV-measured settling velocity
Sediment Mixing Tank: Doppler Settling Measurements -- For each sand size, a grid of vertical velocity ADV measurements were collected with 10 min at each point cm/s (Photo by C. Cartwright) ADV mounted above Circulation outlets 2-D map of vertical Doppler velocity (+ = upward, - = downward) (Example for 125 micron case) -- Vertical velocity measurements are total velocity of sand + water, i.e., ws + <w>. -- Horizontally-integrated flow associated with water along, <w>, must add up to zero. -- So the spatially averaged sum of ws + <w> must be the sediment settling velocity, ws . -- However, the spatial coverage of the tank by the ADV is incomplete.
Map of vertical Doppler velocity (+ = upward, - = downward) cm/s (Example for 125 micron case) a) Measured flow cm/s -- Spatially averaged sum of ws + <w> must be the sediment settling velocity, ws . -- However, the spatial coverage of the tank by the ADV is incomplete. -- Radially symmetry of tank can be used to extrapolate and interpolate flow before averaging to solve for settling velocity, ws . c) Interpolated flow b) Measured flow
Answer: Rapid Sediment Analyzer (RSA) What independent “true” sand settling velocity can we compare Doppler measurements to? Answer: Rapid Sediment Analyzer (RSA) Balance connected to computer Sediment drop and start button Thermometer to measure water temp. Settling tube filled with water Computer records weight and settling time Need to explain what I did, either on this slide or next one Metal plate connected to balance (~150 cm from sediment top) (Photo by C. Cartwright)
Rapid Sediment Analyzer (RSA) Example output from 106 micron sieve Mean Ws = 1.310 ±0.063 cm/sec (Photo by C. Cartwright)
Sediment Mixing Tank: Doppler Settling Measurements RSA Ws and Individual Flow Fit Ws comparison RSA Ws and Global Flow Fit Ws comparison Doppler velocity fits (cm/s) Ws from individual Doppler velocity fits (cm/s) Ws from global Ws from Rapid Sand Analyzer (cm/s) Ws from Rapid Sand Analyzer (cm/s) Individual Flow Fit a) Regression of distance from center vs. <w> b) Circular fit of regression Global Flow Fit Use result from average of all best-fit slopes of velocity vs. radial distance for all sizes, since slope (i.e., spatial distribution of flow) is entirely due to water, not sand.
Dual Use of a Sediment Mixing Tank for Calibrating Acoustic Backscatter and Direct Doppler Measurement of Settling Velocity (and Related Field Motivation and Observations) Grace Cartwright, Carl Friedrichs, and Paul Panetta Outline of Presentation: Motivation: Acoustic Doppler Velocimeter (ADV) Field Observations Sediment Mixing Tank: Acoustic Backscatter Calibrations Sediment Mixing Tank: Doppler Settling Measurements Independent field-based test of ADV-measured settling velocity
<w'C'> = ws <C> This field method was the motivation – But it couldn’t be used in the mixing tank! (The spatial variation in concentration couldn’t be resolved in the tank) Assume: (Fugate & Friedrichs, 2003) Upwards turbulent sediment flux Downwards gravitational settling = <w'C'> = ws <C> w' = vertical turbulent velocity, C' = turbulent concentration fluctuation < > = burst average, ws = sediment settling velocity, <C> = burst-average TSS Biological site (GP) ADV Data: <w'C'> (mm/s)(mg/liter) Slope = ws = <w’C’>/<C> = 1.5 mm/s <C> (mg/liter)
<w'C'> = ws <C> This field method was the motivation – But it couldn’t be used in the mixing tank! (The spatial variation in concentration couldn’t be resolved in the tank) Assume: (Fugate & Friedrichs, 2003) Upwards turbulent sediment flux Downwards gravitational settling = <w'C'> = ws <C> w' = vertical turbulent velocity, C' = turbulent concentration fluctuation < > = burst average, ws = sediment settling velocity, <C> = burst-average TSS Question: -- It there a reliable, independent measure of ws we can use in the field? Biological site (GP) ADV Data: <w'C'> (mm/s)(mg/liter) Slope = ws = <w’C’>/<C> = 1.5 mm/s <C> (mg/liter)
Period of strongest flow Comparison of ADV measurements to camera measurements of from York River estuary from October 6th 2012 (just last week): Promising!
Dual Use of a Sediment Mixing Tank for Calibrating Acoustic Backscatter and Direct Doppler Measurement of Settling Velocity (and Related Field Motivation and Observations) Grace Cartwright, Carl Friedrichs, and Paul Panetta Outline of Presentation: Motivation: Acoustic Doppler Velocimeter (ADV) Field Observations Sediment Mixing Tank: Acoustic Backscatter Calibrations Sediment Mixing Tank: Doppler Settling Measurements Independent field-based test of ADV-measured settling velocity