The Measurement of Bed Load Sediment Transport in Rivers and Estuaries using Stationary and Moving ADCP Methods (using workhorse, channel master and stream pro) David Williams AIMSDarwin Northern Territory Northern TerritoryAustralia
Bed load transport for Habitat change Port projects Navigation Bridge scour /deposition River form studies Sand / gravel extraction Littoral transport Engineering works Environmental management No bed load measurements in NT Rivers
Daly River Anson Bay 300 km
Occur in different locations with respect to the mouth Need for boundary conditions for sediment modelling
Helley-Smith 8055 US BL84 US TR2
Water ping Bottom ping fixed or moving bed
Water velocity Bed load velocity * -1 ADCP assumes a reference plane of a fixed bed Bed load movement will result in a perceived movement upstream Multiply by -1 to give correct direction Bed load rate = bottom track distance / time
Bed load transport follows moving bed tests with some variations Explained in the manuals Moving beds are not a problem
Frame deployed to measure the tides and the movement of the sediments along the sea bed
Tides, currents and bed load movement measured modelled
Measured versus modelled Based on the non-cohesive sediment model of Camenen and Larson (2006) Adapted by Wolanski and Williams
easy deployments from any vessel
using the loop technique using the parallel technique Want to keep it all simple No filtering and minimum post processing Results available after measurement
No dGPS Use bottom tracking No bedload – no misclose
Compass error Calibrate on site e.g. compass error 2 o CE = 100 * sin 2 = 3.5 metres Bed load error = CE / time
Bedload movement results in apparent upstream movement Distance made good Bed movement = dmg/time + - error CE / time
For ADCP loop gaugings the transects need to be surveyed at a slow a speed as is safely possible. This is because Bed load transport rates are generally of a low velocity (0.020 – m/s and are not uniform The longer the sampling time reduces the Compass Error component e.g. as in the previous example if the compass error (CE) over the cross section was 3.5 metres then A 5 minute survey would have 3.5/300 = 0.012m/s A 10 minute syrvey would have 3.5/600 = 0.006m/s A 20 minute survey would have 3.5/1200 = 0.003m/s Compass errors cannot be eliminated altogether but their effect can be reduced to negligible through longer sampling
Daly River Bedload transport study Pool habitats 10 km
Typical flood hydrograph
Bed movement Distance MG / Duration m/s
Bed movement Distance MG / Duration m/s
Bed load rate / m/s * ( m * 90 m) m 3 /s
Flow mean velAreawidth mean depth bed movement (m/s)qbqb bed load rate from equations ADCP Meyer PeterEinstein 42 Einstein 50 Ackers White r2r
Meyer Peter 0.65 Einstein 0.86 Ackers White 1.04 Slope Unit Bed load transport ? is all the sediment layer moving D 50 = 2 mm
GPS required (LengthBT – LengthVTG) / time = bed movement GPS VTG ADCP BT
gpsVTG adcpBT (BT – VTG) / Duration ( – ) / m/s
Bed load versus mean velocity
Bed load versus shear velocity
Bed load versus kinetic energy (½ * ρ * Q * v 2 ) Wil be exploring turbulent kinetic energy
HADCP’s used to continuously measure flow via velocity index or numerical techniques
At lower flows sand is still rolling
At low flow sand still moves and bedforms are made
Q = 36.1 m 3 /s A = 61.5 m 2 W = 64.8 m MV = m/s d = 0.94
HADCP
Sand wave movement during recession flows If mean depth changes by +/- 0.1 metres then mean velocity magnitude will change by 0.15 m/s Sand wave movement is ~ 4 m/25 days = 0.16 m/day Over 8 month dry season Sand waves move ~ 40 metres
Bed load versus mean velocity
Next moving on to cohesive sediment interactions with non cohesive
Conclusions Bottom mounted frame good for some estuarine and offshore sites Able to gather long time series Could be combined with underwater logger and transducer to get sediment thickness Moving boat techniques non intrusive and can gather a lot of information over a reach economically (Q, V, backscatter, bed load rate). Can compute bed shear, Turbulent Kinetic Energy, Reynolds numbers, Froude numbers, Manning’s ‘n’. HADCP measure bed movement and migration of waves downstream StreamPro used for bathymetry in shallow stream to measure bed forms StreamPro used in flume to better understand sand movement