Turbidity & suspended solids instrumentation, historically based in Port Townsend, WA Product line now supported through Campbell Scientific, Inc.
A bold yellow angular logo mark set against a deep dark background, sharp and modern
D & A Instruments
Turbidity monitors & water-quality instrumentation

Building Reliable Sediment Calibration Curves

Suspended-solids sensors estimate the concentration of particles in water by measuring how those particles interact with light. The instrument may report turbidity, backscatter, or a related optical signal, while the environmental result is often needed as milligrams per litre of total suspended solids. A calibration curve connects those two measurements.

Using natural sediment standards makes that connection more representative of field conditions than relying only on manufactured particles or a generic turbidity reference. Sediment from the monitoring site carries the mineral composition, grain-size distribution, particle shape, and color that influence the sensor response. A carefully prepared series of natural suspensions can therefore improve the usefulness of a suspended-solids measurement for dredging, rivers, reservoirs, estuaries, and other marine or freshwater applications.

The process requires more than mixing soil into water and fitting a line. Sample preparation, concentration determination, settling behavior, optical geometry, and statistical validation all affect the final curve. A sound method also documents when the calibration applies and when it should be repeated.

Why Natural Sediment Changes Sensor Response

Optical suspended-solids sensors do not respond to mass directly. They respond to the amount and direction of light scattered, absorbed, or reflected by particles within the measurement volume. Two suspensions with the same mass concentration can produce different signals if their particles have different sizes, shapes, refractive properties, or colors.

Fine clay can generate a strong optical response because it remains dispersed and presents a large total surface area. Coarser sand may settle quickly and produce a weaker or less stable signal at the same measured mass. Dark organic particles can absorb light, while pale mineral grains may scatter it efficiently. Flocculated particles also behave differently from individual grains because their effective size and internal structure change the optical path.

For this reason, a factory calibration or a curve based on a commercial reference material may not predict field suspended-solids concentration accurately. Natural sediment standards should ideally come from the site, the same watershed, or a closely comparable sedimentary environment. For broad monitoring programs, separate curves may be needed for distinct sediment sources or seasons.

The distinction between turbidity and suspended solids is important. Turbidity is an optical property, commonly expressed in nephelometric units, whereas suspended-solids concentration is a mass concentration. A sensor can provide a repeatable turbidity signal while still requiring site-specific conversion to milligrams per litre.

Preparing A Representative Sediment Standard

Begin by collecting enough sediment to represent the material likely to be transported during the monitoring event. A composite sample from several locations or depths can reduce the influence of an unusually coarse or fine deposit. Record the collection location, date, water conditions, sediment appearance, and any visible organic content. Remove large debris, vegetation, shells, and stones that are not part of the suspended fraction being measured.

Drying is convenient for weighing, but it can alter aggregation and organic coatings. If the objective is to reproduce field behavior, gently disaggregate the material and document the preparation method. Some laboratories use wet sediment preserved in its native water, while others use oven-dried material that is later rehydrated. Either approach can work if it is applied consistently and reported with the calibration results.

Prepare a concentrated stock suspension with filtered site water, deionized water adjusted to suitable ionic strength, or another defined carrier medium. The carrier should be selected deliberately. Water chemistry affects clay dispersion, flocculation, and settling, so a curve produced in low-conductivity laboratory water may not transfer directly to brackish or mineral-rich field water.

Calculate stock concentration from a dry-mass measurement rather than assuming that a volume of wet sediment has a known solids content. Mix the stock thoroughly, withdraw a known volume, filter or evaporate it using a validated gravimetric method, dry the retained material to constant mass, and calculate:

[ C = \frac{m_s - m_f}{V} ]

where (C) is suspended-solids concentration, (m_s) is the mass of the filter or container plus dried solids, (m_f) is the tare mass, and (V) is the processed sample volume. Replicate determinations reveal whether the stock is homogeneous enough for calibration work.

Designing The Calibration Series

Use a concentration range that covers the expected environmental measurements, including the lower detection region and the upper range likely during storms, dredging, or resuspension. At least five or six concentration levels are usually more informative than two-point calibration. Include a blank or very low concentration sample to evaluate baseline signal and possible contamination.

Before each measurement, resuspend the natural sediment using a defined mixing procedure. A mechanical stirrer, recirculation loop, or controlled agitation can be used, but the method must reproduce the physical conditions during sensor exposure. Record mixing duration, stirrer speed, vessel dimensions, water temperature, and the time between mixing and measurement.

Settling is often the largest source of inconsistency. Measure each standard immediately after homogenization, or establish a fixed delay that reflects the intended field application. For a sensor installed in flowing water, a continuously mixed tank or recirculating flume may provide a better simulation than a beaker. If the sensor is intended for quiescent profiling, settling behavior should be characterized rather than eliminated.

Collect multiple readings at each concentration and inspect their spread. The sensor should be positioned at a consistent depth and orientation, with enough clearance from the vessel wall and bottom to avoid reflections. Keep bubbles away from the optical window, since air can create large transient signals. Rinse the sensor between standards and measure the sequence in a randomized or alternating order when drift is possible.

A calibration series should include a separate validation set when practical. Standards used to calculate the curve demonstrate fit; independent standards demonstrate predictive performance. Reusing the same observations for both purposes can make the curve appear more reliable than it is.

Selecting A Model And Checking Fit

Plot measured sensor output against gravimetrically determined suspended-solids concentration. The appropriate model depends on the sensor, sediment, and concentration range. A linear relationship may be suitable over a restricted range:

[ C = aS + b ]

where (S) is sensor output, (a) is the slope, and (b) is the intercept. Some optical systems become nonlinear at high concentration because multiple scattering, light attenuation, or detector saturation limits the response. In that case, a piecewise linear, polynomial, or physically informed nonlinear model may be more appropriate.

Do not select a model only because it produces the highest coefficient of determination. Examine residuals, confidence intervals, replicate variability, and error at environmentally important concentrations. A high (R^2) can coexist with a meaningful bias at low concentration or a curved residual pattern at the upper end.

The intercept deserves attention. A nonzero intercept may reflect electronic offset, background turbidity in the carrier water, residual sediment, or an inappropriate blank. It should not be forced to zero unless the instrument physics and validation data support that decision. Report the calibration range and avoid extrapolating beyond it without additional evidence.

Calibration feature What to examine Why it matters
Concentration range Blank through expected peak conditions Prevents routine extrapolation
Replicate spread Standard deviation or coefficient of variation Shows mixing and measurement repeatability
Regression form Linear, segmented, or nonlinear response Accounts for optical saturation and curvature
Residual pattern Bias across low, middle, and high levels Reveals errors hidden by a single fit statistic
Independent validation Predicted versus gravimetric concentration Tests real predictive performance
Sediment description Grain size, color, mineralogy, organic content Defines where the curve can be applied
Recalibration trigger New sediment source, sensor service, or drift Keeps the conversion defensible

For instruments used in dynamic conditions, compare calibration residuals with the accuracy needed for the application. Dredging plume management may require reliable detection of changes and threshold exceedances, while a research study may require tighter mass-concentration accuracy. The acceptable error should be defined before the curve is approved.

Connecting Laboratory Results With Field Conditions

A laboratory curve is useful only when its physical assumptions resemble the deployment. In a river, particles may be sorted vertically, with coarse material near the bed and finer material in the upper water column. In an estuary, salinity changes can promote flocculation and alter optical behavior. During dredging, freshly disturbed sediment may differ from deposited material because the plume contains a changing mixture of particle sizes.

Sensor placement therefore matters. Compare readings from the calibrated instrument with discrete water samples collected near the sensing volume. Analyze those samples using the same gravimetric definition used for the standards. Sampling too far away or at a different depth can create apparent calibration errors that are actually caused by spatial variability.

When suspended solids are measured alongside hydrologic or groundwater observations, the same discipline applies to depth, timing, and quality control. For projects involving hydraulic gradients, vertical profiling context can help frame how depth-resolved measurements should be interpreted and documented. A profile that combines concentration data with changing flow conditions should preserve the exact sensor depth and sampling time.

Field verification should include low, typical, and high conditions where possible. Collect paired discrete samples during calm periods and during the events that matter most operationally. If field samples show a consistent offset or changing slope, investigate sediment composition, fouling, bubbles, water chemistry, and flow regime before editing the curve.

Managing Drift, Fouling, And Data Quality

Optical windows can accumulate biofilm, fine sediment, grease, or air bubbles. These conditions may increase baseline signal, reduce transmitted light, or create erratic readings. Establish a cleaning and inspection schedule based on deployment conditions, not only calendar time. Record cleaning events so that abrupt changes in the data can be distinguished from genuine environmental variation.

Electronic drift and lamp or detector aging can also change the response. Keep a reference standard or verification suspension for periodic checks, while recognizing that a reference material is a performance check rather than a replacement for site-specific natural sediment calibration. A verification result outside the control limits should trigger investigation and possible recalibration.

Time-series screening should preserve the original data. Flag spikes, dropouts, flat-lined values, and impossible rates of change rather than silently deleting them. The spike detection guidance provides a useful example of how quality-control logic can be documented for environmental profiler data; equivalent rules can be adapted to suspended-solids records.

Maintain calibration metadata with the processed data file. Include sediment source, preparation method, carrier water, concentration calculations, sensor serial number, firmware or configuration, optical path details, regression coefficients, valid range, validation error, and date of approval. If the instrument is moved to a new site, treat the existing curve as a hypothesis until field checks support its continued use.

Practical Controls For A Defensible Curve

A reliable calibration program is built around repeatability and traceability. The following controls help keep a natural-sediment curve useful over the life of a monitoring project:

Keep the raw sensor output and original laboratory mass results alongside the final regression coefficients. If a future review identifies a problem with one standard, the curve can be reassessed without repeating the entire project. Version-controlled calibration files are especially important when several instruments are deployed across different waterways.

For product-specific documentation, installation questions, and current support arrangements, the support information can help identify the appropriate technical and product-management resources. Campbell Scientific supports the D & A Instruments product line, including instrumentation used for suspended-solids monitoring, hydrology, environmental research, and OEM integration.

Natural sediment calibration is most valuable when treated as an ongoing measurement process rather than a one-time spreadsheet exercise. Build the curve from representative material, verify it against independent samples, and watch for changes in sediment character and sensor condition. Teams that document those decisions can convert optical measurements into suspended-solids data with a clear, defensible connection to the waterway being monitored. Identify the sediment source, establish the calibration range, and begin with a verification plan before the next deployment.