Designing a Multi-Sensor Array for Vertical Turbidity Profiles

A stratified lake can hold several very different water-quality conditions within a few metres. Clear, warm water may sit above a colder layer, while fine clay, organic particles, or resuspended sediment remain concentrated near the bed. A single turbidity sensor records only one point in that vertical structure, so it can miss the layer that matters most to a water manager, researcher, or dredging contractor.

A multi-sensor array turns isolated readings into a depth-resolved profile. Several optical turbidity instruments are mounted at known elevations on a fixed frame, cable, buoyed line, or profiling platform. Their measurements can reveal the position of a sediment plume, the movement of a mixing boundary, or the depth at which inflow water travels through a reservoir.

Designing a multi-sensor array for vertical profiling of turbidity requires more than placing instruments at equal intervals. The array must reflect the lake’s morphology, likely sources of suspended solids, seasonal thermal behaviour, deployment depth, maintenance access, and the way data will be used. A sound design also distinguishes optical turbidity from suspended-solids concentration, since the relationship between the two depends on local particle characteristics.

Start With The Lake’s Vertical Structure

The first design step is to identify the processes creating vertical variation. Thermal stratification commonly produces an epilimnion near the surface, a metalimnion or thermocline through the transition zone, and a colder hypolimnion below. Turbidity can be highest in any of these layers, depending on whether the main source is surface runoff, inflowing water, algal material, or sediment disturbed from the bottom.

A bathymetric map helps place the array where it will answer a useful question. A deep central basin may be appropriate for long-term lake monitoring, while the mouth of a tributary may better capture a sediment pulse after rainfall. Near an intake, the important objective may be to establish whether a turbid layer is approaching the withdrawal depth. In a shallow arm, wind-driven resuspension can make the lower part of the water column more important than the thermocline.

Australian reservoirs illustrate why site context matters. In the Murray–Darling Basin, inflows can carry fine sediment after intense catchment rainfall, while long dry periods may leave a stable water column and concentrated material near the bed. A drinking-water storage such as Lake Eildon can require a different layout from a shallow coastal lake affected by wind and tidal exchange. In tropical Queensland, seasonal inflow events and high water temperatures may produce rapid changes in mixing and biological fouling.

Before selecting sensor depths, collect temperature, conductivity, depth, and turbidity profiles during contrasting conditions. A handheld sonde or boat-mounted profiler can show where gradients actually occur. Repeat the survey after a storm, during turnover, and in calm weather. These reconnaissance profiles prevent an array from being optimised for a single day.

Select Sensor Positions That Resolve Change

Equal spacing is easy to install, but it is rarely the best scientific arrangement. Place sensors more closely where turbidity is expected to change rapidly: around the thermocline, above the bed, near a tributary intrusion, or at a known intake horizon. Wider spacing can be used in relatively uniform surface water and deep clear layers. The result may be a 1 m interval through the active zone and 3–5 m intervals elsewhere, subject to the lake depth and project objectives.

The lowest sensor should sit high enough above the bed to avoid direct interference from settled material and the near-bed boundary layer. The exact clearance depends on bottom roughness, wave action, mooring movement, and the expected resuspension depth. A sensor mounted too close to soft sediment can produce persistent high readings when its optical window becomes coated or when its beam intersects a moving bedload cloud.

The uppermost instrument should remain submerged through expected water-level changes, or its position should be adjustable. Reservoirs in Australia can experience substantial drawdown, particularly during drought or water-supply operations. A fixed frame that works in winter may place its top sensor too near the surface in summer, while a cable array may need depth marks and a revised mooring position as storage levels change.

A practical array might contain six to ten sensors: closely spaced units through the thermocline and likely plume layer, with broader spacing above and below. For deeper research sites, a hybrid arrangement can combine fixed sensors with an intermittent vertical profiler. Fixed instruments provide continuous time series, while the profiler checks whether the chosen levels still represent the evolving water column.

Useful Placement Checks

• Locate the thermocline and major turbidity gradients before final installation.
• Keep the lowest optical path clear of the bed and moving sediment.
• Allow for drawdown, wave motion, and seasonal changes in lake level.
• Place at least one sensor near the operational depth of an intake or outlet.

Match The Optical System To The Measurement

Turbidity sensors estimate the scattering or attenuation of light caused by particles in water. Suspended-solids sensors use related optical principles, yet a turbidity value in NTU or another instrument-specific unit is not automatically equivalent to milligrams per litre. Quartz, clay, silt, algae, and organic particles scatter light differently, and particle size distribution can change during a flood or mixing event.

Choose an optical configuration that suits the expected range and water matrix. A low-range sensor may resolve clear background conditions well but saturate in a dense sediment plume. A higher-range instrument can capture storm inflows but may offer less sensitivity at very low concentrations. For a wide dynamic range, select sensors with appropriate gain settings, multiple measurement paths, or a deployment strategy that places different sensor types at likely high- and low-concentration levels.

Sensor orientation matters in a vertical array. Optical windows should face away from the mooring line and other components that could cast shadows or create local turbulence. Cross-arm mounts can hold each instrument clear of the central cable. The mounting hardware should resist rotation, protect connectors, and prevent one sensor’s bubbles or disturbed sediment from affecting another.

The optical sensing technology used by the instrument family should be considered alongside the particle population, expected fouling, and required response time. In an OEM or research installation, document the optical geometry, calibration method, output units, and temperature limits so that readings from different depths remain comparable. If the project spans freshwater and estuarine sites, retain site-specific calibration information rather than assuming one conversion equation will work everywhere.

Build Calibration And Quality Control Into The Design

Calibration should combine laboratory checks with field verification. Before deployment, inspect each sensor using clean water and stable reference materials, then record serial number, calibration date, firmware, output scaling, and zero response. Use the same procedure across the array. Small offsets between sensors can look like real layers when displayed as a vertical contour plot.

Field samples are essential for converting optical response into suspended-solids estimates. Collect water at several depths during clear, moderate, and highly turbid conditions. Analyse total suspended solids using a documented gravimetric method, then develop separate relationships where particle composition or concentration range changes. A single regression may be adequate for operational warning thresholds, but research applications often require uncertainty estimates and a record of sample conditions.

Temperature and pressure should be recorded with enough accuracy to confirm sensor depth and interpret stratification. Conductivity can help identify inflow intrusions, saline influence, or groundwater contributions. Time synchronisation is equally important: if the sensors sample at different moments, a fast-moving plume may appear as a false vertical pattern. Use a common clock, consistent logging intervals, and a clear rule for missing or delayed records.

Quality control flags should identify values affected by cleaning, deployment, cable movement, air exposure, saturation, or communications failure. Review both raw measurements and derived concentrations. A sudden spike at one depth may be a genuine sediment event, but it may also indicate a loose mount, a scratched window, or a trapped bubble.

Plan Mooring, Power And Communications

A vertical array must stay in position while allowing water to move naturally around each sensor. A tensioned line with a bottom weight, subsurface float, and surface marker can provide a stable reference, although currents and waves still create tilt. In high-energy locations, a rigid frame or streamlined cage may reduce movement. Record the actual depth and tilt of each instrument rather than relying solely on nominal mounting distances.

Materials should suit long freshwater exposure, ultraviolet radiation, and biofouling. Marine-grade fasteners, suitable plastics, sacrificial protection where needed, and sealed connectors reduce failure risk. In Australian lakes, warm seasons can encourage algal growth and biofilm formation, while remote sites may make a simple mechanical cleaning schedule more realistic than frequent specialist servicing. Wipers, copper components, antifouling guards, or replaceable optical caps can be evaluated during a pilot deployment.

Power calculations must include sensor consumption, logger operation, telemetry, cold-weather performance, and reserve capacity. Solar panels may work well on an exposed buoy but become less reliable when shading, bird fouling, or low winter sun is significant. Store data locally even when cellular telemetry is available. Many inland sites have patchy coverage, and a useful time series should survive a communications outage.

The support model is part of the engineering decision. Campbell Scientific now provides product-management and contact information for the D & A Instruments range, while the technical FAQ resources can help clarify legacy equipment, operating details, and application questions. This matters for Australian projects where procurement, spares, field servicing, and data integration may involve different organisations across a state water utility, university, council, or OEM.

Deployment And Maintenance Records

• Record sensor depth, frame geometry, serial numbers, and clock settings at installation.
• Photograph optical windows, connectors, mounts, and the surrounding site before lowering the array.
• Schedule cleaning and verification around likely algal growth and storm seasons.
• Keep local copies of raw data, calibration files, maintenance notes, and quality flags.

Turn Measurements Into A Useful Profile

A vertical array becomes valuable when its readings are interpreted together rather than viewed as separate sensor traces. Display turbidity against depth and time using a contour plot or coloured profile, then overlay temperature, conductivity, lake level, rainfall, inflow, wind, and operational events. This can show whether a turbidity layer is sinking, spreading, trapped at a density interface, or being lifted by mixing.

Use depth coordinates that reflect the actual water surface and sensor movement. For a fixed frame, calculate elevations from surveyed reference points and lake level. For a cable array, combine pressure or position data with the mooring geometry. Even a small vertical error can shift a sharp turbidity boundary onto the wrong layer, particularly near an intake or thermocline.

Sampling frequency should match the process being observed. Fifteen-minute readings may be sufficient for slow seasonal changes, while storm inflows, dredging plumes, or gate operations may require one-minute data. Apply sensible averaging only after retaining high-resolution raw records. A median or short rolling filter can reduce bubbles and electronic noise, but excessive smoothing may erase the arrival of a narrow sediment pulse.

Interpretation improves when the array is paired with targeted water sampling and occasional full-depth profiles. A weekly field visit may verify fouling and collect calibration samples, while an event-triggered survey can document a flood plume. In a working reservoir, alerts can be based on turbidity at several depths, the rate of change, or the thickness of a high-turbidity layer rather than on one absolute threshold.

The final dataset should distinguish measured optical response from inferred suspended-solids concentration. Report units clearly, retain the site-specific calibration equation, and state the uncertainty associated with particle composition and sensor condition. That discipline makes the array useful for environmental research, dredging plume monitoring, reservoir operations, and long-term assessment of changing sediment behaviour.