Espresso: Open source software for the visualization of multibeam water column data
Article

Espresso: Open source software for the visualization of multibeam water column data

Innovative tool for seabed data exploration

Espresso is a free and open source software to visualize and analyse multibeam water column data. Its core feature is the capability to echo-integrate water column data vertically, allowing for the visualization ‘from above’ of georeferenced water column acoustic anomalies across multiple files. Originally developed at NIWA, Espresso is now open source, licensed under MIT, maintained internationally and available on GitHub. The software is coded in MATLAB and a compiled version is available for Windows.

Water column data visualization

Modern multibeam echosounders can record the acoustic echo returned by objects in the water column between the sonar and the seafloor. This ‘water column data’ can provide useful information about the presence, density, shape and temporal nature of features in the water column such as fish, gas seeps, aquatic vegetation, turbidity, shipwrecks or human-made structures. As a result, water column data is increasingly requested as an additional output of hydrographic surveys carried out under the guiding principle of ‘collect once, use many times’.

However, water column data comes with challenges. In particular, the data is difficult to store and dispatch due to its size, which is typically several orders of magnitude larger than bathymetry and backscatter data. The cause of this size disparity is that for any given ping and beam, there is one bathymetry value and one (or several) backscatter data value(s), but hundreds to thousands of water column data values, each corresponding to a different range from the sonar head to the seafloor and beyond. In other words, water column data is essentially a 3D dataset – varying in pings, beams and range – and this additional dimension leads to another challenge: to visualize and interpret multidimensional data, a visualization method must be chosen that eliminates some dimensionality, at the cost of introducing some inconvenience and/or ambiguity.

For example, water column data is most naturally visualized as a ‘wedge view’, where the values for each beam and each range of a single ping are displayed in the across-track plane (Figure 1a). This method effectively eliminates the ping dimension, so its inherent issue is that to visualize the entire dataset, one would need to go through every ping, one at a time, for every file. A less inconvenient visualization method is the ‘range-stack view’, where the signal for any given ping at a given range is averaged over all beams, which allows the visualization of many pings’ worth of data varying in range as a single image akin to that of a single-beam echosounder (Figure 1b). This method effectively eliminates the beam dimension, at the cost of causing acoustic anomalies to appear distorted and ambiguous. For example, two horizontal echoes on separate sides of the vessel would appear as a single vertical mark in this view. Moreover, to visualize an entire dataset, it is still necessary to go through many such range-stack images.

Figure 1A: Example of water column data containing echoes from gas seeps, visualized as a wedge view. (Data courtesy: Kongsberg EM710 data from the FOSAE-2015-BH03 survey in the Barents Sea, acquired as part of the Norwegian seafloor mapping programme MAREANO (Bøe et al., 2020))
 

A powerful but little-known visualization method is the ‘vertically echo-integrated view’, in which the 3D dataset is georeferenced, gridded and averaged vertically, which enables a 2D visualization ‘from above’ in the manner of bathymetry grids or backscatter mosaics (Figure 2). This method has the enormous advantage of allowing the display of several files’ worth of data in a single image, for efficient scanning and interpretation of broad regions of data. This approach essentially sacrifices the vertical dimension, for which the cost is ambiguity about the depth of an acoustic anomaly, but this is mitigated if the interval of depth, range or height above seafloor of the data to be vertically echo-integrated can be specified. Vertical echo integration has already proven useful for applications such as bubble vent localization (Urban et al., 2017; Mitchell et al., 2022) and mapping kelp density (Lucieer et al., 2023) but, to date and to our knowledge, it is not implemented in any of the few examples of commercial software available to visualize multibeam water column data.

Figure 1B: Example of water column data containing echoes from gas seeps, visualized as a range-stack view. (Data courtesy: Kongsberg EM710 data from the FOSAE-2015-BH03 survey in the Barents Sea, acquired as part of the Norwegian seafloor mapping programme MAREANO (Bøe et al., 2020))

Presentation and workflow overview

Espresso is a research software developed at NIWA between 2018 and 2021 to scrutinize multibeam water column data, including a capability for vertical echo-integration (Figure 3). Espresso is now open source and free to use under MIT licence, and available for download at https://github.com/alexschimel/Espresso. Espresso is coded in MATLAB, but releases are also compiled for Windows, which allows installation of the software as any standard Windows application without the need for a MATLAB licence. In this article, we summarize some of Espresso’s core features. For more information on its capabilities, please see its growing wiki here, which currently includes a quick start guide and a user guide (in development).

An Espresso session starts with converting and loading the desired raw data files. The raw data can be visualized, but Espresso offers a range of pre-processing options to remove or filter unwanted noise that may otherwise dominate the picture, especially in vertically echo-integrated view. These pre-processing options include 1) masking (that is, removing) portions of unwanted data, 2) radiometrically correcting the backscatter level, and 3) filtering the sidelobe artefacts (Figure 4). Several masking options are available, parameterizable and combinable, to remove a range of possibly unwanted data, such as close-range data (perhaps contaminated by wash-down bubble noise), outer beams, bad-quality pings, data making up the seabed echo (or data within a set distance above the seabed echo) or data beyond the minimum slant range. The sidelobe artefacts filtering option is the ‘Slant Range Signal Normalization’ algorithm described in Schimel et al. (2020).

Figure 2: Example of vertically echo-integrated view of water column data containing echoes from gas seeps, created and annotated with Espresso and exported to ArcGIS. The strong acoustic echoes produced by gas seeps are visible from above as ‘hot spots’ (bright yellow) relative to their empty water column surroundings (purple). (Data courtesy: Kongsberg EM302 data acquired from NIWA vessel RV Tangaroa over the Calypso Hydrothermal Vent fields in the Bay of Plenty, New Zealand (Lamarche et al., 2019; Spain et al., 2022))

Espresso can vertically echo-integrate all the water column data in the loaded files (minus data that was masked), but it is possible to limit this processing to data within a given depth interval, or within a given height-above-seafloor interval (Figure 4). This allows focusing the vertically echo-integrated view on certain objects of interest defined by their depth in the water column (such as the deep scattering layer) or by their height above the seafloor (such as aquatic vegetation). In both cases, the result of vertical echo-integration is a horizontal, georeferenced 2D grid, with water column acoustic anomalies shown as hot spots, distinct from the surrounding low-acoustic-energy background (Figure 4). Espresso operates this process on each file individually, but the resulting 2D grids can be blended in a single 2D mosaic, which can then be exported as a geotiff file for further analysis in GIS software (Figure 2).

Espresso also includes additional tools and visualizations to deepen the exploration of water column data. The standard ‘wedge view’ and ‘stacked view’ (in-range or in-depth) are also available (Figure 1). The accompanying bathymetry (and in some data formats, seafloor backscatter data) can also be gridded and displayed, allowing a check of whether water column data anomalies correspond to certain features of the seafloor. Finally, Espresso also includes geo-picking capabilities, enabling users to record the location of features of interest in the water column data as points or polygons, which can be augmented with information and exported in shapefile and text format.

Figure 3: An example screenshot of the main window of Espresso, showing the vertically echo-integrated view for multiple files with geo-picked gas flares (right), display options (top-left), and a range-stack view showing a gas flare (bottom-left). (Data courtesy: Kongsberg EM710 data from the FOSAE-2015-BH03 survey in the Barents Sea, acquired as part of the Norwegian seafloor mapping programme MAREANO (Bøe et al., 2020))

Applications and limitations

Espresso has been used for research by NIWA and NIWA partners for locating gas seeps (Turco et al., 2022) and benthic habitat mapping (Porskamp et al., 2022). In the first case, vertical echo-integration provided insights into the total area of gas seepage in proximity to the seafloor. In the second case, the vertical echo-integrated mosaic was used as an additional geographic layer in a machine learning algorithm, which resulted in increased accuracy in predicting kelp-dominated habitats. NIWA also routinely uses Espresso to visualize bathymetry and seafloor backscatter of newly acquired data for quality control. We foresee that Espresso could be used in various other applications, for example by hydrographers for routine examination and quality control of water column data, by marine biologists for fish school shape analysis, by marine conservationists for location of leaking offshore pipelines, by coastal scientists for turbidity plume tracking, or by marine engineers for examination of the footprint of submerged infrastructure.

Espresso was developed by researchers as a research tool, and thus has more limitations than a software created and maintained by professional developers for commercial use. First, it supports a limited number of multibeam data formats: mostly the Kongsberg .all/.wcd and .kmall/.kmwcd formats, with some support for the Teledyne .s7k format (SeaBat, Norbit systems). Moreover, the data processing in Espresso is often highly simplified, meaning that processed data does not have the same level of quality and positional accuracy as that of professional hydrographic software. More importantly, Espresso was coded in MATLAB and thus faces significant limitations in memory and speed, although considerable efforts were made to optimize the software for large-data handling (e.g. water column data is accessed via memory mapping) and computing speed (some processing steps use parallel computing on machines equipped with a compatible GPU).

Conclusion

Vertical echo-integration is a novel and useful visualization method for multibeam water column data, with a high potential for routine data examination and research. The open source Espresso software provides this visualization capability (and other features) to everyone and for free (under the terms of the MIT licence), thereby constituting a powerful complement to commercial software for the scrutinization and processing of multibeam water column data. The authors hope that the hydrographic community finds this tool useful. If you use Espresso in your work, please acknowledge the authors of this article. For citations, a peer-reviewed article is in preparation.

Figure 4: Overview of the core workflow of Espresso, from loaded raw data to the vertical echo-integration of individual files.

References

Bøe, R., Bjarnadóttir, L. R., Elvenes, S., Dolan, M., Bellec, V., Thorsnes, T., Lepland, A., & Longva, O. (2020). Revealing the secrets of Norway’s seafloor – geological mapping within the MAREANO programme and in coastal areas. Geological Society, London, Special Publications, SP505-2019–2082. https://doi.org/10.1144/SP505-2019-82

Lamarche, G., Le Gonidec Y., Lucieer V., Ladroit Y., Weber T., Gaillot A., Heffron E., Watson, Sy. & Pallentin A. (2019). Gas bubble forensics team surveils the New Zealand ocean. EOS Earth & Space Science News, 100. https://eos.org/science-updates/gas-bubble-forensics-team-surveils-the-new-zealand-ocean

Lucieer, V., Flukes, E., Keane, J. P., Ling, S. D., Nau, A. W., & Shelamoff, V. (2023). Mapping warming reefs—An application of multibeam acoustic water column analysis to define threatened abalone habitat. Frontiers in Remote Sensing, 4, 1–15. https://doi.org/10.3389/frsen.2023.1149900

Mitchell, G. A., Mayer, L. A., & Gharib, J. J. (2022). Bubble vent localization for marine hydrocarbon seep surveys. Interpretation, 10, SB107–SB128. https://doi.org/10.1190/INT-2021-0084.1

Porskamp, P., Schimel, A. C. G., Young, M., Rattray, A., Ladroit, Y., & Ierodiaconou, D. (2022). Integrating multibeam echosounder water‐column data into benthic habitat mapping. Limnology and Oceanography, 67, 1701–1713. https://doi.org/10.1002/lno.12160

Schimel, A. C. G., Brown, C. J., & Ierodiaconou, D. (2020). Automated filtering of multibeam water-column data to detect relative abundance of Giant Kelp (Macrocystis pyrifera). Remote Sensing, 12, 1371. https://doi.org/10.3390/rs12091371

Spain, E., Lamarche G., Lucieer V., Watson S., Ladroit Y., Heffron E., Pallentin A. & Whittaker, J.M. (2022). Acoustic predictors of active fluid expulsion from a hydrothermal vent field, offshore Taupō Volcanic Zone, New Zealand. Frontiers in Earth Science, 9, 785396. https://doi.org/10.3389/feart.2021.785396

Turco, F., Ladroit, Y., Watson, S. J., Seabrook, S., Law, C. S., Crutchley, G. J., Mountjoy, J., Pecher, I. A., Hillman, J. I. T., Woelz, S., & Gorman, A. R. (2022). Estimates of methane release from gas seeps at the Southern Hikurangi Margin, New Zealand. Frontiers in Earth Science, 10, 1–20. https://doi.org/10.3389/feart.2022.834047

Urban, P., Köser, K., & Greinert, J. (2017). Processing of multibeam water column image data for automated bubble/seep detection and repeated mapping. Limnology and Oceanography: Methods, 15, 1–21. https://doi.org/10.1002/lom3.10138

Hydrography Newsletter

Value staying current with hydrography?

Stay on the map with our expertly curated newsletters.

We provide educational insights, industry updates, and inspiring stories from the world of hydrography to help you learn, grow, and navigate your field with confidence. Don't miss out - subscribe today and ensure you're always informed, educated, and inspired by the latest in hydrographic technology and research.

Choose your newsletter(s)