Air bubbles in submerged optical sensors: mitigation techniques
Submerged optical instruments provide rapid measurements of turbidity, suspended solids, and sediment movement in rivers, lakes, estuaries, and coastal waters. Their value depends on a stable optical path: light must travel through the water, interact with particles, and return to a detector in a predictable way. Air bubbles can disrupt every part of that process.
A bubble is more than an empty space in the sample. Its gas-water boundary strongly refracts and reflects light, creating a scattering signal that may be much larger than the signal produced by suspended sediment. A short-lived bubble cloud can therefore appear as a sudden sediment plume, an instrument fault, or an implausible concentration spike.
Effective control begins before deployment. Sensor location, orientation, mounting, cleaning, sampling rate, and quality-control procedures all influence how often bubbles reach the optical window. The best mitigation strategy combines mechanical prevention with data interpretation rather than relying on software correction alone.
Why bubbles distort optical measurements
Optical turbidity sensors and suspended-solids probes generally estimate water quality from the behavior of emitted light. Nephelometric instruments measure light scattered at a defined angle, while optical backscatter systems detect light returned from particles near the sensing volume. In both designs, the measurement assumes that the scattering objects are primarily the particles being studied.
Air bubbles violate that assumption. The refractive-index difference between air and water produces intense reflections at the bubble surface. A single bubble crossing the beam can cause a sharp positive excursion, while a cluster can generate a broad and irregular response. Depending on bubble size, position, and sensor geometry, the signal may saturate, oscillate, or briefly fall if the bubble blocks the receiver.
Bubbles can also alter the apparent relationship between turbidity and suspended-solids concentration. A calibration developed with mineral particles cannot reliably convert a bubble-contaminated optical response into mass concentration. This is especially important in dredging plume monitoring, where a false increase may trigger an unnecessary operational response or conceal the timing of a real sediment event.
How bubble interference appears in data
The most recognizable signature is a spike that rises and falls faster than the surrounding hydrodynamic process could reasonably change. Bubble interference often produces several consecutive peaks, irregular high-frequency noise, or a signal that returns to baseline as soon as the sensor shifts position. In contrast, a genuine sediment plume usually has a spatial or temporal structure related to flow, vessel movement, discharge, tide, or bed disturbance.
The pattern depends on the source. Wave breaking and surface turbulence may create intermittent bubbles near the upper water column. Propellers, thrusters, pumps, and intake structures can produce a persistent cloud. Rapid lowering of a probe can trap air around the optical face, while a poorly designed mounting frame can create a wake that directs bubbles across the measurement volume.
Comparing channels can improve diagnosis. If optical turbidity rises sharply but pressure, conductivity, acoustic measurements, or a second optical path remain stable, bubbles become more likely. A sudden change in the ratio between two optical wavelengths may also indicate bubbles or fouling rather than a change in particle concentration. These checks should be treated as evidence, not as universal rules, because real water-quality events can affect multiple sensors at different speeds.
Sensor placement and mechanical controls
The first control is to keep the sensing face away from active bubble sources. Avoid placing an instrument directly behind a propeller, beneath a discharge, beside an intake, or in the recirculating wake of a vessel or frame. In rivers, examine eddies, hydraulic jumps, spillways, and zones where falling water entrains air. A modest change in depth or horizontal position can reduce bubble exposure more effectively than complex post-processing.
Orientation also matters. A sensing window pointed upward can collect bubbles as they rise, particularly when the instrument is stationary. A vertical or slightly downward-facing arrangement often allows bubbles to pass away from the optical path, although the correct angle depends on the body shape and local flow. The mounting should avoid cavities and sharp shoulders that shed vortices directly toward the window.
Use a streamlined frame with open water flow around the instrument. Protective guards must prevent impact without creating a bubble trap. Cables should be secured so they do not flutter across the sensing zone or form loops where bubbles accumulate. In moving surveys, keep the instrument clear of the vessel’s boundary layer and verify that lowering speed does not drag air down the mooring line.
A clean optical window is essential because films, sediment deposits, and biological growth change the same signal that bubbles affect. Wipers or mechanical cleaning systems can reduce fouling, but they do not eliminate entrained air. Cleaning intervals should be based on site conditions, and any cleaning event should be recorded because it can create a temporary signal disturbance.
| Mitigation approach | Best use | Main benefit | Limitation |
|---|---|---|---|
| Repositioning the sensor | Propeller wakes, intakes, wave action | Removes the instrument from the bubble source | May change the sampled water mass |
| Downward or sideward orientation | Rising bubbles near fixed installations | Encourages bubbles to leave the optical face | Geometry must be tested in real flow |
| Streamlined mounting | Mobile surveys and moorings | Reduces vortex shedding and bubble capture | Guards can create new turbulence |
| Mechanical wiping | Sites with biological or sediment fouling | Restores a clean optical path | Does not prevent bubble entry |
| Shorter measurement intervals | Intermittent bubble passage | Helps identify transient spikes | Produces more data and noise |
| Median or persistence filtering | Automated monitoring | Suppresses isolated excursions | Can remove genuine short events |
| Reference sensors or channels | Research and compliance systems | Supports event discrimination | Adds cost and integration effort |
Sampling design and signal processing
Sampling frequency should reflect both the expected water-quality changes and the speed of bubble interference. A slow logger can average a brief bubble event into a misleading moderate concentration. A faster logger preserves the shape of the disturbance and makes it easier to identify. Recording raw or minimally processed measurements alongside reported values is useful when data quality must later be audited.
Robust statistics are often preferable to a simple arithmetic mean. A short rolling median can reject isolated spikes, while a trimmed mean can reduce the influence of extreme values during a reporting interval. Persistence rules are another option: accept a high value as a real event only if it remains above a threshold for a defined duration or appears consistently across adjacent samples.
Filters require restraint. A real sediment pulse can also be brief, especially close to a dredge head, outfall, or storm front. Set the window length and threshold using field observations rather than selecting an aggressive filter that makes the record look smooth. Preserve the unfiltered series, document every rule, and flag rejected or suspect observations instead of silently replacing them.
Where possible, combine optical data with instrument diagnostics. A bubble event may coincide with unusually high variance, signal saturation, abrupt changes in detector balance, or a mismatch between redundant optical channels. A quality flag can then distinguish “high turbidity” from “possible bubble interference.” This distinction is valuable for environmental reporting, adaptive dredging control, and long-term trend analysis.
Selecting the right optical configuration
Different optical technologies respond differently to particle size, concentration, color, and bubble presence. Nephelometric turbidity sensors are commonly used for standardized turbidity measurement, whereas optical backscatter instruments are often selected for suspended-sediment monitoring and field profiling. Understanding the sensing geometry helps engineers anticipate how an air-water interface may enter the measurement volume. A useful sensor comparison guide can support this initial evaluation.
The instrument should match the application rather than simply the nominal concentration range. A dredging plume survey may require rapid response and strong resistance to high sediment loads. A groundwater or freshwater profiler may prioritize compact form, low power consumption, and stable performance during vertical movement. Marine installations may need pressure-rated housings, antifouling measures, and mounting hardware designed for currents and wave-driven motion.
Calibration and validation should include the actual deployment environment. Collect water samples during calm and turbulent periods, during representative sediment concentrations, and when bubbles are visibly present if possible. Compare laboratory solids analysis with the optical signal, while marking intervals affected by air. A calibration that excludes contaminated observations is more defensible than one that forces bubble-driven readings into the sediment relationship.
The technical downloads provide a practical place to review available documentation, specifications, and application information before configuring a monitoring system. Product-management and support information for the D & A Instruments line is provided through Campbell Scientific, which can help users identify current options and integration details.
Validation in moving water
A tank test can reveal whether bubbles strike the window, but it rarely reproduces the turbulence of a working vessel, tidal channel, spillway, or dredging site. Field validation should therefore include controlled changes in position and orientation. Move the sensor away from the suspected source, rotate it incrementally, and compare the resulting signal while other conditions remain as constant as possible.
For mobile surveys, log vessel speed, heading, depth, pitch, roll, and proximity to machinery. A bubble plume may appear only at a particular heading or throttle setting. In a fixed station, record water level, wave conditions, discharge, pump status, and maintenance events. These contextual variables can explain short anomalies that would otherwise be misclassified as sediment changes.
Independent observations strengthen the diagnosis. Visual inspection from a vessel, underwater video, acoustic backscatter, dissolved oxygen, and duplicate optical instruments can each provide useful evidence. No single auxiliary measurement is perfect, but agreement among several indicators makes automated quality flags more reliable.
Deployment records should include the sensor model, optical path, mounting angle, cleaning method, calibration date, depth, and local flow conditions. When a time series is later used for compliance, research, or operational decisions, this metadata shows whether a sudden increase is physically plausible and whether the instrument was exposed to a known bubble-generating condition.
A practical field checklist
Bubble mitigation works best as a repeatable operating procedure. Before installation, identify every likely source of entrained air and decide how the sensor will be protected without isolating it from the water being measured. During commissioning, inspect the live signal while changing orientation and flow exposure rather than accepting the first convenient mounting position.
Data handling should be equally deliberate. Establish quality flags before collection begins, retain raw measurements, and define when an observation is reported as suspect rather than converted into a concentration. The following practices are suitable starting points for many submerged optical monitoring systems:
- Install the sensor away from propellers, pump outlets, falling water, hydraulic jumps, and vessel wakes.
- Use a streamlined, open mounting frame and secure cables so they cannot direct bubbles toward the optical window.
- Test a sideward or downward-facing orientation, then verify that it does not compromise the intended sampling zone.
- Record at a rate that preserves short disturbances and apply documented median, persistence, or variance-based quality checks.
- Validate optical results with water samples, duplicate sensors, or another independent measurement during representative flow conditions.
Long-term programs should review flagged data alongside maintenance and weather records. If anomalies increase after a mounting change, cleaning interval, seasonal temperature shift, or equipment replacement, revisit the physical installation before modifying the algorithm. A recurring bubble problem is usually easier and cheaper to solve at the sensor than in the database.
Reliable submerged measurements depend on treating the optical path as part of the measurement system, not as an isolated component. Apply the placement, mounting, validation, and data-quality practices described here to reduce false turbidity events and protect the integrity of sediment and water-quality records. Review the available D & A Instruments documentation and involve Campbell Scientific support when selecting, integrating, or maintaining equipment for a demanding marine or freshwater deployment.