MBES as a practical tool for seepage impact assessments

Recent research demonstrates that multibeam echosounder (MBES) water column data deserves closer attention from the hydrographic community, particularly for the remote estimation of underwater gas emissions. Although these acoustic estimates may initially appear coarse and require expertise in acoustic calibration, they enable cost-effective classification of seepage sites and scale efficiently to large areas with numerous bubble streams. This article explains how these measurements became possible and how they were validated.

Gas release from the seabed occurs in many forms, ranging from natural methane seepage to leakage from abandoned wells and subsea infrastructure. Multibeam echosounders (MBESs) that record water column data are proven tools for detecting marine gas bubble releases. They are used by scientists and professionals alike to map natural seeps and assess leaks from pipelines, wells or (future) carbon storage sites. The good area coverage and widespread availability on survey vessels makes them a practical choice for such investigations.

However, while detecting bubble release (Figure 1) is important, it is only the first step in gas seepage investigations and is often insufficient for decision-making. The real questions are: How much gas is being released? Does the seepage matter? Are further investigation or actions of any kind required? And, if so, where should efforts be focused?

The importance of quantifying gas release

Quantitative information enables prioritization: determining which sites matter most, which leaks warrant further investigation, and how emissions evolve over time. To answer these questions, data is traditionally obtained through direct in-situ measurements using remotely operated vehicles (ROVs) (Figure 2) or seabed landers. These methods can be accurate, but they are also expensive, operationally complex and do not scale well to investigate hundreds or even only dozens of bubble streams.

Multibeam echosounders, on the other hand, provide good area coverage and are already used to detect the seepage. Using them for quantification as well would minimize the need for these expensive operations. This would enable faster and more flexible gas emission monitoring, which could also be carried out from smaller vessels at significantly lower costs.

From plume counting to estimating gas flow

Gas bubbles are clearly visible in multibeam water column data as vertical columns of enhanced backscatter (which is also shown in Figure 1), known as acoustic plumes. These plumes are typically straightforward to identify, but their interpretation has been mostly qualitative (i.e. counting, describing size and/or shape).

It has long been established that the collective acoustic backscattering of a column of gas bubbles can be used to estimate the gas flow of that plume if the statistical distribution of released bubble sizes is known. This inversion of the acoustic signals is based on a fundamental principle called echo integration, which is widely used in fisheries science to estimate fish stocks. However, applying echo integration to multibeam measurements of gas plumes is not straightforward. The overlapping and chaotic arrangement of acoustic samples smears the signal over multiple pixels, making it challenging to combine these measurements without introducing biases from factors like survey speed, depth or beam angle.

To address this, a new principle that integrates acoustic values on a three-dimensional voxel grid and is referred to as ‘echo grid integration’ was developed at GEOMAR, Germany. Simulations confirmed that the echo grid-integrated acoustic energy from a bubble stream scales linearly with the volumetric gas flow passing through horizontal layers of the water column. The main practical challenges of this approach are assumptions about stable bubble size distributions across varying flow rates and a requirement for acoustically calibrated multibeam systems.

Calibration challenges

Acoustic calibration means establishing a reliable reference for backscatter levels, which can vary due to differences between systems, transducer aging or changes in water temperature. The echo grid integration method enables quantitative comparisons of bubble streams from the same acoustic survey, even without calibration. For instance, it can determine that one bubble stream may release twice as much gas as another, assuming similar bubble sizes. Calibration, however, extends this capability by allowing comparisons between bubble stream measurements from different surveys and systems. Furthermore, calibration against an absolute reference enables the conversion of relative acoustic measurements into absolute gas flow estimates.

Several practical calibration approaches are available, though no single, universally simple method exists yet. Previous work combined multibeam data with simultaneous measurements from a calibrated singlebeam echosounder by targeting the same bubble stream or by using layers of high background turbidity. Another approach involves using seafloor reference areas where backscatter levels are stable. The most straightforward method is direct calibration against known gas flow rates from controlled bubble releases, although this requires the capability to deploy such a source in the field.

Bubble size uncertainty

While calibration addresses system-related variability, another critical factor in acoustic inversion is bubble size. Bubble sizes represent the largest source of uncertainty in acoustic measurements because acoustic backscattering does not increase linearly with bubble volume. As a result, gas sources with larger bubbles produce less signal per unit of released gas volume. To address this, reasonable limits for minimum and maximum bubble sizes must be estimated using cameras, passive acoustic hydrophones or literature values.

If bubble sizes are not measured directly, the resulting uncertainty can be significant and may reach a factor of two or more. However, even with this uncertainty, acoustic methods provide more accuracy than simple visual classification. With literature-based estimates for realistic bubble sizes, it is typically possible to constrain the flow to the right order of magnitude. For example, a weak seep releasing 50ml/min might be estimated between 20 and 150ml/min, but this uncertainty interval remains sufficient to distinguish it from bubble streams that release 200ml/min or more. With sufficient demand for developing practical methods, these uncertainties could be reduced further in the future. One option would be the remote estimation of bubble size distributions using broadband or multifrequency acoustic devices.

Validation through controlled field experiments

To move beyond theoretical feasibility, the method was tested in a field experiment carried out in collaboration with TNO and NIOZ in the Dutch North Sea (Figure 3). A seabed-deployed bubble generator produced gas plumes of consistent bubble size with adjustable flow rates. Independent measurements provided ground-truth reference values for the actual gas release. During these experiments, bubble plumes were surveyed using a ship-mounted multibeam echosounder, the Kongsberg EM2040, and processed using the quantitative integration framework. The resulting flow rate estimates were compared directly with the known release rates.

Preliminary results showed a clear and reproducible linear relationship between integrated multibeam backscatter and gas flow rate. For the tested conditions, deviations were on the order of 10%. These experimental results provide strong empirical evidence that modern multibeam systems are capable of quantitative gas measurements under realistic survey conditions. Ongoing research at TNO and NIOZ is currently validating the bubble backscatter modelling by comparing measurements from two different bubble sizes released by this lander. Once verified, the achieved calibration will be part of a larger work focused on estimating the gas flow from abandoned wells in the Dutch North Sea.

Conclusion

Water column data represents a valuable and often underused information source. In the right context, it can provide quantitative answers to questions that were previously considered outside the scope of acoustic surveys (Figure 4). Quantitative multibeam approaches are particularly attractive where spatial coverage and efficiency matter. Large areas can be screened for emissions and temporal changes can be monitored. An important aspect is that the most interesting and important sites can be identified before deploying more expensive in-situ techniques.

The discussed method is not yet a push-button solution and requires careful planning, calibration and expert interpretation. When these conditions are met, however, it delivers substantial gains in spatial coverage, efficiency and cost reduction for marine bubble seepage assessments. Future advancements will further improve the accessibility of quantitative multibeam measurements, potentially enabling seepage assessment as part of routine hydrographic operations.

Further reading

Urban et al. (2017), Processing of multibeam water column image data for automated bubble/seep detection and repeated mapping, Limnol. Oceanogr. Methods, 15(1), 1-21, doi.org/10.1002/lom3.10138

Urban et al. (2023), Echo grid integration: A novel method for preprocessing multibeam water column data to quantify underwater gas bubble emissions, Limnol. Oceanogr. Methods, lom3.10552, doi.org/10.1002/lom3.10552

Urban et al. (2023), Combining multibeam and singlebeam echosounders for quantifying gas bubble release from the seafloor, Limnol. Oceanogr. Methods, 24(2), e10722, doi.org/10.1002/lom3.10722

De Bruin et al. (2025), Methane leakage from abandoned wells in the Dutch North Sea, Mar. Pet. Geol., 171, 107184, doi.org/10.1016/j.marpetgeo.2024.107184

Presentation by Peter Urban et al. (2025), Quantifying gas bubble seepage using multibeam echosounders: a scalable workflow for monitoring, Methane Emissions in the North Sea Symposium, Utrecht, the Netherlands, www.nlog.nl/media/3613

Poster by Laurens van der Marel et al. (2025) Calibration of multibeam echosounder bubble plume measurements using a controlled bubble plume generator, presented at the Methane Emissions in the North Sea Symposium, Utrecht, the Netherlands, 13-14 November 2025, https://www.nlog.nl/media/3613