Integrating optical data with sediment transport models

Optical suspended-solids data can reveal when sediment is moving, but it does not directly measure bedload. An optical turbidity sensor records light scattering or attenuation, while bedload is the mass of grains transported through saltation, rolling and sliding along a channel or seabed. Converting the first measurement into the second therefore requires hydraulic data, particle properties and a sediment-transport equation. The resulting bedload estimate supports dredging plume monitoring, environmental research, defence applications and OEM systems that need a defensible measure of near-bed sediment flux.

This integration is strongest when the optical instrument is treated as an independent observation that constrains a physically based model, rather than as a direct proxy for bedload. A practical workflow combines site-specific calibration, flow-derived bed shear, grain-size information and sediment-transport equations. For Australian projects, the approach must also accommodate episodic stormwater, tidal re-suspension, wind and waves, dredging activity, cohesive river sediment and the variable supply of material from catchments, dunes and eroding banks.

Why optical concentration differs from bedload flux

Concentration and bedload flux have different dimensions and physical meanings. Total suspended solids, usually reported in milligrams per litre, describe the mass of particles suspended in a sampled volume. Bedload flux, commonly expressed as kilograms per square metre per second or kilograms per metre per second, describes mass crossing a bed boundary. An optical sensor at one elevation and position cannot integrate the entire moving layer, and the relationship between a suspended concentration and bedload flux changes with flow depth, velocity, turbulence and sediment supply.

Particle properties create another limitation. Most optical turbidity instruments respond most strongly to fine silt and clay because their size produces strong light scattering. Coarse quartz sand may contribute substantially to bedload while remaining almost invisible to an optical sensor. The measured signal can therefore represent wash load accompanying a transport event rather than the sand actually moving along the bed. Conversely, bed material abrasion can generate fines that make the optical response increase without indicating a comparable increase in bedload.

A sound integration uses the transport equation as a physical estimate and the optical trace as evidence about event timing, sediment availability and near-bed conditions. During rising flow, an optical increase may mark the arrival of sediment from upstream. During falling flow, the signal may persist because fines take longer to settle than coarse grains. These patterns help identify erosion, transport, exhaustion and deposition phases, but they must be interpreted with flow and particle data to avoid attributing every turbidity pulse to bedload.

Building a sediment-specific optical calibration

A general conversion from nephelometric turbidity units to total suspended solids should not be assumed to apply across sites. The relationship changes with particle size, shape, colour, refractive index and mineralogy. Collect paired optical readings and gravimetric samples over the expected concentration range, with enough coverage of clear-water, moderate-flow and storm conditions. Samples are normally filtered, dried and weighed to determine suspended-solids concentration. Australian testing should use a NATA-accredited laboratory where contractual defensibility is important, and the sampling method should remain consistent through calibration and validation.

The background material on optical sensing technology explains how scattering and attenuation differ and why sensor configuration matters. Raw signal, instrument settings and site metadata should be retained rather than storing only converted values. A site-specific rating curve can then be expressed as concentration against signal, turbidity or another instrument output. Its uncertainty will widen outside the sampled range, making concentration limits and missing-data flags important for later model enforcement.

Calibration also needs an assessment of non-sediment effects. Air bubbles, suspended organic matter, changing water temperature, salt-bearing coastal water, sensor fouling and biological growth can alter the optical response. Duplicate readings, clean-water blanks, periodic sensor cleaning and comparison with a reference instrument help separate these effects from genuine sediment changes. Vertical placement should also be considered because settling creates strong concentration gradients near the bed and at the surface.

Adapting bedload equations to field evidence

Bedload equations express transport as a function of excess bed shear, grain size, particle density and bed conditions. The Shields parameter is commonly written as (\theta=\tau/[(\rho_s-\rho)gd]), where (\tau) is bed shear stress, (d) is representative grain diameter and (\rho_s) and (\rho) are sediment and water densities. The excess stress (\theta-\theta_c) compares the applied shear with the critical value required to move grains. The Meyer-Peter-Müller equation is widely used for sand, while Engelund-Hansen and Soulsby-van Rijn formulations provide alternatives for other well-sorted or mixed-sediment conditions.

The selected equation should match the material being transported. Sand-sized particles generally suit bedload relationships more closely than cohesive silt and clay. For muddy or mixed environments, suspended and bedload pathways may interact so strongly that separate transport components cannot be estimated reliably. Fine sediment may also be supplied from the banks or upstream rather than the local bed. The distinction between wash load and bed material load must therefore be established before interpreting the model output.

Optical observations can update or test the equation through a data-assimilation step. One option is a Bayesian or Kalman filter in which the transport equation provides the prior bedload flux and the optical measurement supplies a likelihood term based on a site-specific response model. Another is an event-based regression between calculated transport and measured concentration. The response model must allow for hysteresis, settling delays and changing calibration rather than forcing a direct linear relationship. Clarifications about instrument output and calibration are also available in the measurement FAQs.

Deploying through variable Australian conditions

Deployment design should match the transport process being investigated. A sensor installed near the bed is more relevant to coarse bedload conditions, whereas a surface instrument may better describe a dredging plume or catchment export. Sensors should be positioned to avoid wakes, sediment shadows, direct sunlight, bubbles and unstable mounting. Flow velocity and depth are needed to estimate shear stress, and wave or tidal conditions may require a combined-current shear formulation. Before field installation, bench agitation can test the effect of sensor position and flow; the stirred-tank placement guide provides a practical reference for that work.

Australian catchments produce sharply contrasting optical records. Brisbane River and tributary flows can rise rapidly during summer storms, carrying organic material and mineral sediment that create short-lived blackwater events and sewer-overflow signals. In Melbourne, dredging in Port Phillip Bay can produce a near-surface plume that must be separated from natural tidal re-suspension or vessel-generated turbulence. Around Port Adelaide, wind-driven resuspension, tidal salinity and episodic urban-drain discharge can alter optical readings independently of the grains moving along the bed. These processes require event classification, not just a single turbidity-to-solids conversion.

Procurement and compliance are part of the deployment environment. Equipment imported into Australia must be correctly declared through Australian Customs, using the appropriate tariff classification and meeting applicable GST requirements. Australian buyers in the water-quality and sediment-monitoring market also need local calibration advice, replacement parts and clear documentation for state environmental authorities and port-consent conditions. The former D & A Instruments product line is now supported by Campbell Scientific for contact and product-management information, which is particularly relevant when a specification has been carried over from legacy D & A documentation.

Practical checks and reporting outputs

The controls that matter before accepting an automated bedload estimate are straightforward:

• Pair optical readings with gravimetric samples and analyse the representative grain-size distribution.
• Derive bed shear from measured hydraulics, including current, wave and tidal contributions where relevant.
• Check the optical trace for fouling, bubbles, sensor drift and concentrations outside the calibration envelope.
• Validate at least one transport event against bedload traps, sediment collections or an independently surveyed sediment budget.

A defensible reporting package should distinguish measured and modelled quantities:

• Report bedload flux in mass-per-width-per-time units, with the selected equation and its assumptions.
• Publish optical total suspended solids separately from inferred bedload and identify the calibration period.
• Quantify uncertainty from concentration calibration, grain size, shear stress, sensor position and model structure.
• Flag storm drainage, dredging, weed removal, vessel activity or sensor servicing that could affect interpretation.

Used together, the optical trace and sediment-transport equations produce a more defensible account of sediment mobility than either source can provide alone. The optical record resolves the timing and concentration of sediment pulses, while the equations provide the physical conversion from hydraulic forcing to bedload flux. For Australian monitoring programmes, that combination creates an auditable sediment budget suitable for catchment management, dredging operations, coastal engineering and marine environmental monitoring.