Sound Velocity Errors in Hydrographic Surveying: Causes, Detection and Prevention

Sound Velocity Errors in Hydrographic Surveying: Causes, Detection and Prevention

Sound velocity is one of the most important environmental inputs in hydrographic surveying. A multibeam echosounder may be correctly installed, calibrated and operated, yet still produce distorted bathymetry when the sound-speed information is inaccurate, outdated or incorrectly applied.

These errors are especially noticeable toward the outer beams of a multibeam swath. If they are not detected early, they can affect crossline comparisons, feature detection, surface models, volume calculations and the final confidence assigned to the survey.

This article explains how sound travels through seawater, where sound-velocity errors enter the measurement chain, what the typical warning signs look like and which checks survey teams should perform offshore.

What determines sound velocity in seawater?

Sound speed in seawater is controlled mainly by:

Temperature: Warmer water normally increases sound speed.

Salinity: Higher salinity generally increases sound speed.

Pressure or depth: Sound speed generally increases as pressure increases with depth.

Ocean water is rarely uniform. Rivers, rainfall, currents, tides, solar heating and mixing can create layers with different temperatures and salinities. As an acoustic pulse crosses these layers, its direction changes through refraction.

A multibeam system must account for this bending when calculating the position of each sounding. The farther a beam travels from nadir, the more sensitive its computed position becomes to an incorrect sound-speed model.

Two sound-velocity inputs with different jobs

It is important to distinguish between the sound speed measured at the transducer and the sound-velocity profile through the water column.

Surface sound speed

A surface sound-speed sensor is normally installed close to the multibeam transducer. Its value is used by the echosounder for beam steering and receive-beam forming.

If this value is incorrect, delayed, unstable or not reaching the sonar, the beam angles themselves can be wrong. A good profile cannot fully correct a serious beam-steering error introduced during acquisition.

Sound-velocity profile

An SVP, CTD or moving-vessel profiler measures sound speed at different depths. The acquisition or processing software uses this profile for ray tracing: calculating how each acoustic path bends from the transducer to the seabed.

An old profile, a cast taken in unrepresentative water or a profile applied to the wrong lines can produce systematic depth and horizontal-position errors across the swath.

How sound-velocity errors affect multibeam data

For a near-vertical beam, most of the acoustic travel is close to nadir, so the effect may be small. Outer beams travel a longer slant range and cross the water-column layers at larger angles. Small modelling errors are therefore magnified toward the edges of the swath.

Common effects include:

Curved or distorted seabed on an otherwise flat area

Outer beams that bend consistently upward or downward

Mismatch between adjacent survey lines

Poor agreement between main lines and crosslines

Features appearing at different depths or horizontal positions in opposing lines

Increased uncertainty as beam angle increases

A surface that looks acceptable near nadir but separates at the swath edges

Surveyors often describe the characteristic cross-section as a smile or frown. However, the direction of curvature alone should not be used as the final diagnosis. The result depends on the true water-column structure, the incorrect profile used, sonar geometry and software convention. Compare the profile, line direction, overlap and independent checks before deciding on the cause.

Where errors enter the measurement chain

Sound-velocity problems do not always begin with a poor cast. The complete chain should be checked.

1. Sensor measurement errors

The probe may have calibration drift, contamination, air bubbles, poor response time or a damaged temperature, conductivity or pressure sensor. A winch-speed problem can also reduce the quality of a cast.

2. Incorrect cast location or timing

A technically good cast may not represent the water through which the vessel is surveying. This can happen near river outfalls, fronts, tidal boundaries, dredging plumes or areas with strong daily heating.

3. Incomplete depth coverage

Stopping a cast well above the seabed forces the software to extrapolate the lower part of the profile. That assumption may be unsuitable where a deeper layer or thermocline is present.

4. Data-format or import problems

Depth and sound-speed columns can be misidentified, units can be wrong, headers can be read incorrectly, duplicate depths may be present, or the software may reject part of the profile without a clear warning.

5. Wrong profile selection

Operators may accidentally apply a cast from the wrong day, location or tide. Automated selection may also use unsuitable time or distance limits.

6. Surface sound-speed problems

The surface sensor may be fouled, installed in aerated flow, affected by bubbles, frozen at one value or mapped to the wrong input. A difference between the surface sensor and the top of the cast can be valid, but a persistent unexplained difference requires investigation.

7. Processing mistakes

Data may be ray-traced twice, processed with an unintended default profile, or exported before the correct profile is applied. Changes to the profile database can also create inconsistent results between processing sessions.

How to identify a sound-velocity problem

No single display proves that sound velocity is the cause. A reliable diagnosis combines several observations.

Review swath cross-sections

Inspect clean, reasonably flat seabed in the across-track view. Look for smooth, systematic curvature that increases with beam angle. Random scatter is more likely to indicate noise, motion, aeration or bottom-detection problems.

Compare reciprocal or overlapping lines

Adjacent lines should agree in their overlap. If the outer beams disagree while the near-nadir soundings match, refraction is a strong possibility. Reciprocal lines are particularly useful because the same seabed is observed from opposite sides of the swath.

Examine crosslines

Crosslines provide an independent geometric comparison. Plot the differences by depth, beam angle, time and location. A systematic relationship with across-track distance can reveal sound-speed modelling errors.

Compare casts

Overlay the current and previous profiles. Check whether a thermocline has moved, the near-surface layer has changed, or the new cast is inconsistent with local conditions. Do not compare only the average velocity; profile shape is critical.

Monitor the surface sensor

Trend the real-time surface sound-speed value. Sudden steps, long flat periods, excessive noise or disagreement with the top of a recent cast can indicate a sensor or data-link problem.

Test the profile in processing

Reprocess a short, representative section using the most appropriate nearby casts. If the line agreement improves consistently without masking another problem, the original profile selection was probably unsuitable. Keep a record of every test and do not overwrite the raw acquisition data.

Checks to perform offshore

Before acquisition

Confirm that the profiler and surface sensor are within calibration.

Inspect sensors for fouling, damage and trapped air.

Verify time, position, depth and sound-speed fields in the cast file.

Confirm the acquisition system is receiving a live and plausible surface value.

Check which profile is active and how automatic profile selection is configured.

Record weather, tide, water depth and expected environmental variability.

During a cast

Take the profile in a representative location without creating unnecessary operational risk.

Lower the probe at a controlled rate and monitor the live trace where possible.

Reach close enough to the seabed to characterise the full water column while maintaining safe clearance.

Watch for spikes, loops, dropouts or unrealistic gradients.

Record cast time, coordinates, maximum depth and operator comments.

During acquisition

Monitor surface sound speed continuously.

Review swath profiles and line overlap instead of relying only on the plan view.

Run timely confidence checks or crosslines.

Repeat casts when water conditions, tide stage, location or data behaviour changes materially.

Document which cast applies to each acquisition period.

During processing

Confirm that each line uses the intended profile and ray-tracing method.

Compare main-line overlaps before aggressive cleaning.

Analyze crossline statistics by beam angle and depth.

Separate sound-velocity artefacts from patch-test, motion, timing and tide errors.

Preserve a traceable record of profile changes and reprocessing decisions.

Avoid confusing sound-velocity errors with other problems

Several faults can produce similar-looking surfaces:

Possible problem

Typical clue

Sound-velocity profile error

Smooth error that usually grows toward the outer beams and may vary with time or area

Roll alignment error

Opposite vertical bias on port and starboard, repeated consistently on reciprocal lines

Timing or latency error

Position-dependent mismatch that becomes more visible with vessel speed or changing seabed slope

Heave or motion problem

Short-period vertical movement or line-wide noise linked to vessel motion

Tide or vertical-reference error

Similar vertical shift across most or all of the swath

Bottom-detection problem

Irregular or noisy soundings linked to weak returns, vegetation, soft sediment or incorrect sonar settings

The correct approach is to test one hypothesis at a time. For example, if only the outer beams separate and the difference changes after applying a better profile, sound velocity is more likely than a constant roll bias.

Common mistakes to avoid

Using one cast for an entire long survey: Water-column conditions can change with position, tide and time.

Taking casts only by schedule: Fixed intervals are useful, but observed environmental or data changes should also trigger a new cast.

Ignoring the surface sensor: A good SVP does not replace a valid real-time sound speed at the transducer.

Assuming the newest cast is always best: The closest cast in time may be on the other side of a water mass boundary.

Cleaning away curved outer beams: Editing may hide the symptom without correcting the measurement model.

Applying an SVP without checking it: Always verify position, time, units, depth coverage and profile shape.

Blaming every swath mismatch on refraction: Patch-test, timing, motion and vertical-reference errors must also be considered.

Failing to record profile usage: Without traceability, later reprocessing and quality assurance become difficult.

A practical troubleshooting workflow

When a sound-velocity error is suspected:

Stop and review a clean section before collecting more questionable data.

Check the live surface sound speed for plausibility and stability.

Confirm the correct profile is selected and correctly formatted.

Compare recent casts and examine changes in the water column.

Inspect line overlaps, reciprocal lines and crosslines by beam angle.

Acquire a new full-depth profile in representative water if conditions allow.

Reprocess a short test section and compare the result objectively.

Check competing causes such as roll, latency, motion and tide.

Document the finding, corrective action and affected acquisition interval.

Re-run affected lines when the required accuracy cannot be achieved reliably through valid reprocessing.

Conclusion

Sound velocity is not simply a value added during processing; it is part of the complete multibeam measurement chain. The surface sensor supports correct beam steering, while the water-column profile supports accurate ray tracing. Both must be measured, transferred, selected and applied correctly.

The most effective quality control combines environmental awareness with systematic data checks. Monitor the surface value, collect representative full-depth profiles, review outer-beam behaviour, compare overlaps and crosslines, and preserve a clear record of every profile used. Detecting a sound-velocity problem offshore is far less costly than discovering it after the vessel has left the survey area.

Suggested excerpt: Sound-velocity errors can bend multibeam swaths, create line mismatches and reduce bathymetric accuracy. Learn how to identify the warning signs, test the measurement chain and prevent refraction-related problems offshore.

#Hydrographic#Survey#Sound Velocity #SVP #Multibeam #Echosounder #Bathymetry #Quality #Control #Offshore #Survey