Calibrating Suspended-Solids Sensors With Grab Samples In The Field

Optical suspended-solids sensors provide continuous information where laboratory testing alone cannot. A probe can record changes caused by dredging, stormwater, tidal currents, bank erosion, mine dewatering, or construction activity at intervals of seconds or minutes. The instrument usually measures light scattering or attenuation, then reports turbidity in NTU or FNU. Environmental managers, however, often need suspended-sediment concentration (SSC) in milligrams per litre.

The relationship between turbidity and SSC is site-specific. Clay, silt, sand, organic matter, particle colour, grain size, and sensor geometry all affect the optical response. An in-situ calibration using carefully matched grab samples converts the sensor’s output into a defensible estimate of solids concentration under the conditions that matter at the monitoring location.

Why Site Calibration Is Necessary

A factory calibration establishes that the instrument responds correctly to a reference standard. It does not predict the concentration of sediment in a particular estuary, river, dredge plume, or groundwater investigation. Two waters with the same SSC can produce different turbidity values if one contains fine pale clay and the other contains dark organic particles or coarse mineral grains.

The calibration curve can also change as sediment sources change. A dredging campaign at the Port of Brisbane may expose compacted marine clay, while a later storm event carries weathered catchment soil into the same reach. Seasonal flow in northern Australia can create an equally pronounced difference between dry-season background water and wet-season runoff. For this reason, the most useful relationship is based on samples collected while the sensor is operating in the target water body.

A grab sample is a physical sample of water collected at a known time, depth, and location. Its laboratory SSC result becomes the reference value for the sensor reading recorded at the same moment. Several paired observations across the expected concentration range are used to fit a regression equation. The result may be linear, logarithmic, or segmented, depending on the optical behaviour of the particles and the quality of the data.

Before starting, define the reporting purpose. A compliance program may require a conservative alert threshold, while a research project may require the most accurate estimate across a broad range. Confirm whether the final data should be reported as SSC, turbidity, or both, and document the units, sampling depth, averaging interval, and time zone used by the data logger.

Preparing The Sensor And Sampling Plan

Inspect the sensor before deployment and record its serial number, firmware, optical configuration, cable length, and measurement settings. Clean the optical windows with a soft, lint-free material and check for scratches, biological growth, trapped sediment, or air bubbles. A clean-water reading can identify an obvious drift problem, but it should not be treated as a full field calibration.

Install the probe where it sees representative water rather than a stagnant pocket or a direct jet. In a dredging application, this may mean positioning it within the plume while avoiding contact with the seabed, cutter head, discharge pipe, or floating debris. The principles described in optical sensing principles are useful when deciding how optical geometry and plume variability may influence the deployment.

Choose sampling conditions before collecting the first bottle. The plan should cover background water, moderate concentrations, and the highest practical concentration expected during the project. If the sensor will be used to trigger an alert, collect several samples around the proposed trigger level. A curve based only on clear water and one extreme event will be weak in the range where decisions are being made.

Prepare clean, labelled sample bottles, a field notebook or electronic form, a cooler with ice, gloves, a depth sampler if required, and a reliable way to record time. In Australia, remote teams may travel by four-wheel-drive or small boat and rely on solar-powered loggers and intermittent communications. Download data locally when possible, because a delayed telemetry link should not prevent the field crew from matching samples to sensor measurements.

Collecting Matched Water Samples

Allow the sensor to stabilise before taking a sample. Record the instantaneous reading and a short average, such as the mean over 30 to 60 seconds, depending on how quickly the water is changing. Note water depth, tide stage, flow condition, weather, recent rainfall, visible plume characteristics, and any nearby activity. If the instrument output is fluctuating rapidly, collect duplicate samples or use a longer averaging period, but retain the raw time series.

The sample must represent the same water measured by the optical path. Collect it as close as practical to the sensor and at the same depth. A bottle lowered from the surface is unsuitable if the probe is near the bed and the water column is stratified. When conditions are hazardous or the sensor is fixed at depth, a sampler attached to a weighted line, pole, or winch can provide a better match. Avoid disturbing bottom sediment while retrieving it.

Fill the bottle according to the laboratory’s instructions, leaving the required headspace and avoiding unnecessary shaking. Label it with a unique sample number, date, local time, site, depth, sampler, and preservation details. Chain-of-custody records are especially important for regulatory work. Samples should be chilled promptly and delivered within the laboratory’s specified holding time; suspended particles can settle, flocculate, or adhere to the container if the sample is left warm for too long.

Collect enough paired observations to describe natural variability rather than simply repeating the same condition. Ten samples may be a starting point for a simple exploratory curve, but a robust operational calibration commonly needs more observations spread across the measurement range. Replicates help estimate laboratory and sampling variation. If the site contains distinct sediment populations, such as riverine material and dredged seabed sediment, record the event type and consider separate calibration groups.

Turning Laboratory Results Into A Calibration Curve

The laboratory should determine SSC using a documented gravimetric method. A typical process involves mixing the sample thoroughly, filtering a measured volume through a pre-weighed filter, drying the retained material to a specified temperature, cooling it in a desiccator, and weighing it again. The mass increase divided by the filtered volume gives concentration, commonly expressed in mg/L. The laboratory should report method detection limits, replicate results, and any sample condition that could affect accuracy.

Pair each laboratory result with the sensor value from the correct timestamp. Do not match a bottle to the nearest available reading if the plume was moving quickly. Correct time synchronisation between the logger, field clock, sampler, and laboratory chain-of-custody record is essential. If the sensor records a rolling average, calculate the corresponding average from the raw data rather than using a single instantaneous value.

Plot sensor output against laboratory SSC before choosing an equation. Look for curvature, clusters, outliers, and changing scatter at higher concentrations. A simple model might be:

SSC = a × sensor reading + b

If the relationship bends, a logarithmic or polynomial model may fit better, but mathematical complexity should have a physical reason. An equation that fits one dataset extremely well can perform poorly outside that dataset. Avoid forcing the intercept through zero unless the measurements demonstrate that this is appropriate and the monitoring purpose supports it.

Use residuals and validation samples to test the curve. Residuals are the differences between laboratory concentration and the concentration predicted by the sensor. Check for bias across low, medium, and high values, rather than relying only on R². Keep some paired samples separate from the fitting process, then use them to assess prediction error. Report the number of samples, concentration range, equation, units, fit statistics, validation error, and date of calibration.

At high concentrations, optical sensors can become non-linear because the water blocks or scatters too much light. At very low concentrations, the signal may approach the instrument’s noise floor. If the operating range is broad, a two-range calibration or a warning that results are approximate at the extremes may be more honest than one equation applied everywhere.

Managing Drift And Field Uncertainty

A calibration curve remains valid only while the water and instrument behave similarly. Inspect the sensor regularly for biofouling, sediment coating, scratches, cable damage, and loose mounting hardware. In warm coastal Australia, marine growth can develop quickly, while inland systems may experience fine sediment deposition after flood pulses. Wipers, copper components, protective guards, or scheduled cleaning can reduce fouling, but each measure should be checked for interference with the optical measurement.

Recalibrate when sediment appearance changes, the sensor is moved, the optical head is replaced, the deployment depth changes substantially, or quality-control checks show a persistent bias. For long dredging projects, collect occasional verification grab samples even after the initial curve has been accepted. A stable relationship during calm conditions does not guarantee accuracy during a spring tide, flood peak, or unusual discharge.

Site work should also account for Australian environmental requirements. A monitoring program near the Great Barrier Reef may be governed by project approval conditions and Queensland requirements, while a New South Wales dredging or construction project may need to satisfy consent conditions and NSW Environment Protection Authority expectations. The federal Environment Protection and Biodiversity Conservation Act 1999 can apply where nationally protected matters are affected. ANZG 2018 water-quality guidelines are widely used as a technical reference, but the project approval or regulator’s condition determines the enforceable requirement.

Data interpretation benefits from local context. Tidal ports such as Sydney Harbour and Fremantle can show repeated concentration changes unrelated to construction, while tropical sites around Darwin may shift rapidly between dry-season clarity and wet-season runoff. In remote groundwater work, a profiler may pass through unconsolidated sediments with changing optical conditions; profiler deployment guidance can help with positioning and sediment disturbance controls.

Practical Controls For Reliable Results

A defensible calibration is supported by simple controls that make field observations traceable. Keep the original sensor output, laboratory certificates, sample photographs where useful, deployment sketches, cleaning records, and calibration calculations together. Do not overwrite raw values with converted SSC data. Store the equation and version number in the data-processing workflow so later users can identify which curve produced each result.

For equipment specifications, application notes, and supporting documentation, the technical downloads area provides a useful reference point. Product-management and contact information for the D & A Instruments range is now supported by Campbell Scientific, so project teams should confirm current configuration and support arrangements before ordering replacement parts or planning a new deployment.

The following practices improve confidence in a grab-sample calibration:

• Collect paired samples across the full expected SSC range, with extra observations near operational or regulatory thresholds.
• Match sample depth, location, and collection time to the sensor measurement as closely as field conditions allow.
• Record tide, rainfall, flow, dredging activity, sediment appearance, and any unusual disturbance beside every sample.
• Use clean bottles, chilled transport, documented chain of custody, and a laboratory gravimetric method suitable for suspended solids.
• Validate the fitted equation with independent samples and inspect residuals rather than relying on R² alone.
• Retain raw optical data, laboratory results, calibration coefficients, units, and revision dates in the project records.
• Repeat verification sampling after fouling, maintenance, relocation, major weather events, or a noticeable change in sediment source.

When these controls are followed, an optical probe becomes more than a turbidity indicator. It becomes a site-specific monitoring instrument with a documented relationship to suspended solids, capable of supporting operational decisions, environmental reporting, and long-term trend analysis in Australian marine and freshwater settings.