Estimating suspended-sediment particle size with optical sensors

Suspended sediment is rarely uniform. A river plume may contain fine clay, mineral silt, organic fragments and coarse sand, with the mixture changing as flow, tide and bed disturbance vary. Measuring concentration alone can therefore hide important information about transport, settling behaviour, dredging impacts and the fate of contaminants attached to particles.

Multi-wavelength optical sensors provide a practical way to estimate these changes in particle size distribution (PSD). By measuring how suspended material interacts with several wavelengths of light, an instrument can extract spectral patterns associated with particle size classes. The result is an indirect estimate rather than a direct image of every grain, so reliable interpretation depends on calibration, sound optical design and knowledge of the local water and sediment.

Why particle size matters in suspended water

Particle size affects how long sediment remains in the water column, how far it travels and where it eventually settles. Fine clay and silt can stay suspended through tidal cycles and move well beyond a dredging site, while coarser particles often settle close to the disturbance. Size also influences the transport of nutrients, metals, hydrocarbons and other substances that may adhere to mineral surfaces or organic matter.

A conventional turbidity value indicates the water’s optical cloudiness, usually in nephelometric turbidity units, but it does not uniquely identify the sediment mixture causing that response. Two samples can have the same turbidity while containing very different proportions of clay and sand. Suspended-solids concentration has a similar limitation: it describes mass per unit volume, not the distribution of particle diameters within that mass.

For Australian monitoring programmes, this distinction is important in waterways such as the Murray–Darling Basin, where changing flows can mobilise fine bank sediment, and in coastal Queensland, where dredging plumes may affect seagrass, reefs or aquaculture areas. A PSD estimate can help distinguish a short-lived coarse resuspension event from a persistent fine-particle plume that remains detectable after the visible disturbance has dispersed.

How multiple wavelengths reveal size classes

Optical sensing works by transmitting light into the water and measuring the light scattered or attenuated by particles. The response depends on particle diameter, concentration, mineral composition, shape, refractive index and the geometry of the transmitter and detector. At a single wavelength, these effects are strongly mixed. Measurements at several wavelengths provide additional structure for separating them.

The wavelength response can be expressed as a spectral slope or as ratios between signals. Small particles generally interact differently with shorter wavelengths than larger particles, although the exact relationship depends on whether the particles are much smaller than, similar to or much larger than the wavelength. When particle dimensions approach the wavelength of the light, scattering becomes especially sensitive to size and composition. A multi-wavelength system can use these changes to estimate broad fractions such as fine, medium and coarse suspended material.

The measurement is usually treated as an inverse problem. A forward optical model predicts the sensor response for a proposed PSD, and an inversion algorithm searches for the distribution that best matches the observed signals. Models may draw on Mie scattering for simplified spherical particles, empirical relationships derived from field samples, or more advanced descriptions for irregular mineral grains. Because the inversion can be underdetermined, reporting a few robust size classes is often more defensible than presenting a highly detailed grain-size curve.

A useful instrument may combine several optical channels with different wavelengths, viewing angles or path lengths. Wavelength diversity improves sensitivity to size, while angular diversity can add information about scattering behaviour. In practice, the best design depends on whether the application requires rapid plume tracking, long-term river monitoring, process control or an OEM sensor integrated into a larger hydrology platform.

Calibration turns optical response into a useful estimate

Calibration must connect optical signals with independently measured particle properties. Water samples collected across the expected range of conditions can be analysed in a laboratory using laser diffraction, settling methods, sieve analysis or microscopy. The laboratory method should match the intended reporting convention, because a volume-based PSD and a mass-based PSD can tell different stories about the same sample.

The samples should cover changes in mineralogy, organic content, concentration and particle size, rather than representing a single calm-weather condition. A sensor calibrated only with fine estuarine mud may perform poorly when exposed to freshly disturbed sand or highly organic material. For Australian sites, calibration may need to account for seasonal river discharge, tropical storm runoff, dry-season concentration changes and local dredge spoil characteristics.

Concentration has a major influence on optical measurements. If particle loading rises, detectors may saturate, multiple scattering may become significant and the relationship between signal and mass may cease to be linear. At very low concentrations, background light, detector noise and coloured dissolved organic matter can dominate. Calibration data should therefore define a valid operating range and identify conditions where the PSD estimate should be flagged rather than accepted automatically.

Field validation is equally important. A practical programme can compare sensor data with depth-integrated samples, bottle samples taken at several locations, and laboratory PSD results. Sampling should include different depths and flow stages because particles can stratify vertically. In an Australian estuary, for example, a surface reading during a flood tide may not represent the near-bed material moving during a spring-tide resuspension event.

Separating sediment effects from instrument effects

Optical signals are influenced by the sensor’s measurement geometry as much as by the particles. Nephelometric instruments commonly measure light scattered at a defined angle, whereas optical backscatter systems detect light returned from a volume in front of the probe. These approaches have different sensitivities, calibration behaviours and responses to particle size. A clear explanation of the sensor differences is useful when selecting data for a PSD model or comparing records from different instruments.

Window fouling, bubbles, ambient sunlight and biological growth can create spectral changes that resemble a shift in sediment composition. Wipers, anti-fouling measures, shaded installation and sensible deployment orientation reduce these errors, but they do not remove the need for quality control. Diagnostic channels, reference measurements and time-based checks can help identify an optical signal caused by the installation rather than by suspended solids.

A robust processing workflow should examine signal stability, concentration range and relationships between wavelengths. Sudden changes in one channel may indicate bubbles or fouling, while a coherent change across all channels is more likely to reflect a genuine plume. Cross-checking with water level, current velocity, rainfall and site observations can add valuable context. In ports such as Brisbane or Gladstone, vessel movements and propeller wash may explain short bursts of coarse material that would otherwise be mistaken for a calibration failure.

Sensor placement also affects the meaning of the result. A probe mounted near the bed may capture resuspension and bedload-related particles, while a surface deployment may focus on the dispersing portion of a plume. Fixed stations, profiling packages and autonomous platforms each produce different spatial samples. The installation should be designed around the management question, whether that is compliance at a boundary, plume mapping around a dredger or continuous sediment transport measurement.

Applying PSD estimates in Australian monitoring

Dredging is a prominent use case for multi-wavelength sediment sensing in Australia. Ports, navigation channels and coastal construction sites often need near-real-time information about plume movement. A sensor package can provide frequent measurements near dredging equipment and at sensitive receptors, helping operators relate observed concentrations and size fractions to tidal currents, vessel activity and dredge cycles.

Environmental researchers can also use optical PSD estimates to study catchment erosion, wetland exchange and coastal sediment budgets. Along the Great Barrier Reef, distinguishing fine, persistent material from rapidly settling coarse sediment can improve assessments of how river discharge and coastal works influence water clarity. The estimate is most valuable when combined with current measurements, bathymetry, rainfall records and occasional physical samples rather than treated as a standalone measurement.

Defence and maritime users may require compact systems that operate for long periods with limited access. Remote sites around Western Australia, the Northern Territory and offshore facilities can impose high demands on power consumption, telemetry, corrosion resistance and maintenance intervals. A multi-wavelength instrument must continue to produce traceable data through heat, biofouling, suspended organic matter and rapidly changing weather, including conditions associated with tropical cyclones.

For OEMs and Australian integrators, the sensor output may be incorporated into a buoy, profiling package, autonomous vehicle or larger water-quality station. Consistent interfaces, documented calibration coefficients and clearly defined data flags are essential when the optical head is combined with third-party electronics. The local market also values serviceability: replaceable windows, accessible cleaning procedures and support through an established instrumentation provider can matter as much as the nominal optical specification. Product and application information, including technical FAQs, can help engineers assess terminology, legacy D & A Instruments equipment and current support arrangements through Campbell Scientific.

Interpreting results with appropriate limits

A PSD derived from multi-wavelength signals should be presented with uncertainty and operating limits. The estimate may be strong for distinguishing a fine-dominated plume from a coarse-dominated event while remaining unreliable for resolving adjacent size classes. Reporting confidence intervals, quality flags and the calibration range prevents the data from appearing more precise than the physics allows.

Particle shape and composition deserve particular attention. Flat clay minerals, angular quartz grains, organic fibres and biological particles can produce different scattering patterns even when their nominal diameters are similar. Coloured water can attenuate wavelengths unevenly, and a mixed population can generate a response that does not match any simple single-material model. Site-specific laboratory samples remain the best way to identify these ambiguities.

Data interpretation should also distinguish optical diameter from settling diameter. Optical methods respond to how a particle interacts with light, while settling behaviour depends on density, shape and hydrodynamic drag. A porous organic floc may have a large optical dimension but settle slowly, whereas a dense mineral grain can settle quickly. Converting optical size classes into transport or deposition predictions therefore requires additional assumptions or independent settling measurements.

Used within those limits, multi-wavelength optical sensing provides a valuable bridge between simple turbidity monitoring and intensive laboratory particle analysis. It offers high-frequency information about changes in suspended material, supports faster responses to dredging and storm events, and can improve understanding of sediment movement across rivers, estuaries and marine environments. The strongest monitoring systems combine spectral measurements with careful calibration, physical sampling and site knowledge.