In This Issue
Fall Bridge on Ocean Exploration and its Engineering Challenges
September 18, 2018 Volume 48 Issue 3
This issue is dedicated to the engineering methods used to enhance understanding of the world’s oceans.

Mapping the World's Oceans

Thursday, September 20, 2018

Author: Larry A. Mayer

Mapping is fundamental for exploring, navigating, engineering, exploiting, protecting, and understanding the world. Through the great advances of modern remote sensing technology, it is now relatively simple to image and map the one-quarter of the earth’s surface that is readily visible to optical sensors on aircraft or satellites. A 5-year-old using Google Earth can interactively visualize almost any place on the Earth’s surface at a meter-scale resolution with just a few mouse clicks. Detailed three-dimensional depictions (Digital Terrain Models) of the Earth’s surface are easily downloadable and offer critical information for a range of applications. With the addition of multispectral satellite imagery, information derived from remote sensing enables characterization of the nature of the surface (e.g., forests, deserts, crops, etc.).

Unfortunately, the electromagnetic waves used for Earth imaging cannot penetrate more than a few tens of meters through seawater (depending on water clarity). Thus for the three-quarters of the Earth’s surface that lies beneath the oceans other sensors are necessary to map the ocean depths and to image subsurface ocean processes.

A Brief History of Ocean Mapping

For thousands of years, a weight at the end of a rope (or wire), called a lead line (figure 1 inset), was the only means to determine ocean depth, measuring only a tiny spot of the seafloor. In shallow water (tens of meters), a lead line measurement was reasonably accurate and could be completed in a few minutes, but with increasing depth, accuracy decreased and the collection of a single depth measurement could take hours. Each sounding was positioned by the best technology of the day, typically celestial navigation, which, in general, has an accuracy of ±2 nautical miles (Defence Council 2011). The resulting charts showed sparsely spaced soundings with contours or isobaths representing lines of constant depth interpolated between the few measured soundings (figure 1). At best they were broad approximations of what might lie below.

Figure 1 

Remarkably, the lead line was the primary tool to measure depth until the early 20th century, when the sinking of the Titanic led to early experiments with acoustic sources to detect objects in the ocean and, eventually, to the development of echo sounders (D’Amico and -Pittinger 2009). Echo sounders generate an acoustic pulse that propagates from a transducer (typically magnetostrictive or piezoelectric materials) that oscillates when a voltage is applied to it.

Unlike electromag-netic waves, acoustic waves propagate easily in ocean waters. At frequencies of approximately 15 kHz or less, they can propagate the full range of ocean depths (>11,000 m), reflecting off the seafloor (due to the contrast in acoustic impedance between seawater and the seabed) and back to the echo sounder. If the speed of sound in seawater is known (nominally 1,500 m/sec and easily measurable), the two-way travel time of the acoustic pulse can be converted to an accurate measurement of depth.

Figure 2 

Single beam echo sounders form a single beam of sound from the source. Their typical beam width is 15–30 degrees, resulting in an area of ensonification on the seafloor (the beam footprint) with a diameter between one half and one times the water depth (figure 2). Thus in 4,000 m of water, a typical single beam echo sounder would ensonify an area on the seafloor with a diameter of 2,000–4,000 m. The echo from the seafloor represents the shallowest point in the beam footprint but, with no angular resolution in the beam, its position can only be assumed to come from directly below the ship (figure 2). The resulting bathymetric information is thus a spatially averaged (over the scale of the beam footprint) representation of the true depth of the seafloor.

By the end of World War II single beam echo -sounders had been improved to provide rapid, but laterally averaged, measurements of seafloor depths. By the end of the 20th century, the National Geophysical Data Center (now the National Center for Environmental Information, NCEI), a repository for global bathymetric data, had approximately 70 million single beam depth soundings, allowing the production of generalized maps depicting interpolated contours between sparse ship tracks to indicate bathymetric trends. Today the NCEI database has more than 96 million single beam soundings.[1]

Current Technologies

Toward the end of the 20th century two great advances were made in seafloor mapping: the development of techniques to use satellite altimetry to predict seafloor bathymetry and the evolution of multibeam sonar technology from classified military applications to the academic and commercial communities.

Satellite Altimetry

Bathymetry can be derived from satellite altimetry through the fact that the ocean surface deviates from mean sea level in response to the excess (seamounts) or absence (trenches) of mass created by deep-sea -topography. Satellite altimeters measure the -topography of the ocean surface and, by relating the altimeter measurements to real bathymetric measurements, Smith and Sandwell (1997) demonstrated an approach to predict depth measurements for the global ocean. The bathymetric maps produced by satellite altimetry were revolutionary, providing an unprecedented view of seafloor topography and tremendous insight into -tectonic-scale processes. However, such maps are limited in achievable resolution (altimetry resolution is in bands from 20 km to 200 km, whereas interpolated maps are produced with pixel resolution of 1–12 km; Smith and Sandwell 1994, 1997) and fall far short of the resolution achievable with modern multibeam sonars.

Multibeam Echo Sounders

Unlike single beam sonars, multibeam sonars use two arrays, a transmitting array oriented in the direction of vessel transit (along-track) and a receiving array orthogonal to the transmit array (a Mills Cross or Mills T). The transmit array produces a fan-shaped acoustic pulse that ensonifies a swath of seafloor that is wide (-typically 150°) in the across-track direction and very -narrow (typically 1°) in the along-track direction. The receive array forms a number (up to several hundred) of individual beams that are narrow in the across-track direction (typically on the order of 1°) and a bit wider in the along-track direction. The result is the simultaneous high-resolution mapping of hundreds of small patches of the seafloor based on the intersection of the transmit and the receive beams, across the entire swath (figure 3).

Figure 3 

Like all acoustic systems, multibeam sonars must balance the tradeoffs between propagation range (which decreases with increasing frequency due to higher attenuation at higher frequencies) and resolution (which increases with higher frequency due to shorter pulse lengths and increased bandwidth). Small, high--frequency (>400 kHz) systems with arrays of a few 10s of cm in length are capable of centimetric resolution of seafloor features but with propagation ranges of a few hundred meters or less. Large (many meters long) arrays operating at lower frequencies (~12 kHz) are capable of propagating to full ocean depth, but with lateral resolution of 10s to 100s of meters depending on water depth and sonar beam width (e.g., a 1° × 1° multibeam sonar would ensonify a 70 m diameter patch of seafloor in 4,000 m of water).

Figure 4 

Despite these tradeoffs, the resolution achievable by multibeam sonars is significantly higher than that achievable by satellite altimetry or single beam sonars and, when combined with modern satellite navigation, has revolutionized the ability to map and image the seafloor (figure 4). The resolution, density, and coverage of multibeam data have offered new perspectives of seafloor topography that have yielded many insights into seafloor and ocean processes (e.g., sediment dynamics and the origin and creation of oceanic crust), led to new engineering applications (e.g., cable and pipeline surveys, offshore platform installations), improved navigation safety, enhanced maritime heritage studies, and enabled numerous national security applications.

Figure 5 

Most multibeam sonars also simultaneously record the amplitude of the seafloor return that, when appropriately processed and geo-referenced, produces an image of seafloor backscatter that can be related to the roughness or composition of the seafloor (figure 5). More recently, multibeam sonars have begun to collect information on scattering targets in the water column, with applications in fisheries science, physical oceanography, and most importantly the location and quantification of natural and man-made gas seeps (figure 5; Stranne et al. 2017; Weber et al. 2014).

While most multibeam sonars are deployed on surface vessels (typically low-frequency, deepwater systems on large vessels and high-resolution shallow water systems on small, near-shore vessels), small, high-resolution multibeam sonars can also be deployed on remotely operated vehicles (ROVs) or autonomous underwater vehicles (AUVs). When operated close to the seafloor they are capable of producing very high (centimetric) resolution maps in deep water, though only over small areas and at very slow speeds.

How Much of the Seafloor Is Mapped?

How much of the ocean has been mapped with modern sonar technology?

The international General Bathymetric Chart of the Oceans (GEBCO) has, for more than 100 years, been collecting and distributing publicly available -bathymetric data for the world’s oceans (Hall 2006).[2] The most recent GEBCO compilation (GEBCO_2014[3]) is a gridded (approx. 1 × 1 km) product that incorporates single and multibeam sonar data on a backdrop of satellite altimetry (figure 6).

Figure 6 

An evaluation of the GEBCO database shows that only 18 percent of all grid cells in the compilation contain any bathymetric data and only 9 percent contain modern multibeam sonar data (Weatherall et al. 2015). Thus at a 1 × 1 km scale, approximately 82 percent of the seafloor has never had a direct measurement of depth. The sparsity of high-resolution seafloor mapping data became painfully obvious during the search for Malaysia Airlines flight MH370: the existing bathy-metric dataset was so inadequate that the searchers had to conduct surface ship–deployed multibeam sonar surveys in order to know what depth ratings were needed for the AUVs that were necessary to resolve the wreckage.

The Quest to “Fill the Gaps”

Recognizing the poor state of knowledge of ocean depths and the critical role of such knowledge in understanding and maintaining the planet, the Nippon Foundation–GEBCO Seabed 2030 Project has been established to facilitate efforts to promote the mapping of the entire deep sea floor by 2030 at a resolution higher than that of GEBCO_2014. The project has established target resolutions ranging from 100 × 100 m grid cells in depths up to 1,500 m to 800 × 800 m resolution for water depths of 5,750–11,000 m (Mayer et al. 2018).

The Seabed 2030 effort has set up regional data assembly and coordination centers whose short-term goal is to locate regional data not yet made available to global compilations and whose long-term goal is to coordinate and facilitate regional mapping efforts that will “fill the gaps.” A global data center will handle central coordination and production, establish data standards, and distribute global gridded products.

Using the new variable resolution grids, only 6 percent of the seafloor has been mapped by multibeam sonar. What would it take to map the entire seafloor with multibeam sonar? Excluding depths of less than 200 m (which represent only 7 percent of the global seafloor, are typically within the territorial waters and exclusive economic zones of coastal states, and are -inefficient to map because of the depth-dependent coverage of multibeam sonar), it would take approximately 350 ship-years to map the 93 percent of the world’s seafloor deeper than 200 m using current multibeam sonar and surface ship technology (Mayer et al. 2018), at an estimated cost of $3–6 billion.

For context, what is the likelihood of spending billions of dollars to produce high-resolution maps of a planet? This has actually been done. The Moon and Mars have been mapped to much higher resolution (on the order of meters, in many cases) than the Earth’s seafloor. While the cost of individual planetary mapping missions has varied, there is no question that many billions of dollars have been spent over the years. This has been well-spent money—the information returned from these missions has greatly increased knowledge and understanding of these celestial bodies. Why, then, the unwillingness to devote a similar level of expenditure to understand this planet?

New Technologies

New approaches and technologies may reduce the estimated cost of high-resolution mapping of the seafloor. With the widespread use of echo sounders on commercial and recreational vessels, the concept of “crowd sourcing” of bathymetric data is gaining momentum. Echo sounders do not have the resolution of multibeam sonars, but the large number of soundings by “the crowd” can increase the percentage of the seafloor mapped. Quality control is always an issue with crowd-sourced data, but efforts are under way to explore the feasibility of low-cost, “authoritative” echo sounders that are self-calibrating and offer data of known uncertainty. The jury is still out concerning the value of crowd-sourced data for global bathymetry, but in the simplest terms, something is probably better than nothing.

Given the vast areas of unmapped seafloor and the tradeoffs among resolution, propagation, and coverage, the biggest constraint in collecting global multibeam sonar mapping data is the cost associated with the relatively slow-moving surface survey vessels. There are limits to the speed at which multibeam sonar data can be collected (without degrading quality and/or leaving mapping gaps), and a vessel’s fuel costs go up significantly as its transit speed increases.

Figure 7 

One approach to reduce costs is to use autonomous vessels for survey operations (Manda et al. 2015). While early efforts have focused on small vessels with shallow water mapping systems, the growing use of autonomy for larger vessels offers the possibility of developing an “autonomous mapping barge” (figure 7, upper). Such a vessel is ideally suited for a large multibeam sonar array and can carry huge supplies of fuel, be piloted remotely, and transmit data to a base station in real time. Such “telepresence” technology is already well established (Marlow et al. 2017) and will advance rapidly in the years ahead. The multibeam barge can also be a platform for many other autonomous measurements (oceanographic, atmospheric, etc.). The estimated cost of operating such a platform would be one-half to one-third that of a staffed research vessel, and with a significantly lower cost of construction.

In addition, efforts are under way to mount a multibeam sonar on an autonomous sailing vessel; the viability of this method for the collection of oceanographic data over long periods (e.g., 6 months) has been demonstrated (Voosen 2018). The developers of these small “Saildrones” are working with a sonar manufacturer to install a deepwater multibeam system on a large (22 m) Saildrone (figure 7, lower). Remotely piloted and with a satellite link to transmit data to a control center, fleets of these vessels may be able to collect high-resolution multibeam sonar data in remote areas of the ocean in a very cost-effective way.

While the approaches described above focus on reducing platform costs, innovations in the distribution of acoustic sensors (e.g., sparse arrays) may also be important in future mapping efforts.

Conclusions

Only 6–9 percent of the seafloor has had a direct measurement of depth at a scale needed to understand ocean processes, protect or exploit seafloor resources, or support engineering operations. With growing awareness of the ocean’s fundamental role in sustaining life, mediating climate, and supporting commerce, there is increasing interest in seeing the entire seafloor mapped at the scales needed to appropriately understand and manage ocean resources. Efforts involve both the search for existing data that are not yet publicly available and, most importantly, the organization of global and -regional efforts to collect new high--resolution mapping data.

The costs of ocean mapping can be reduced through the use of “crowd-sourced” data and autonomous surface vessels that can be deployed for long periods in remote areas at a fraction of the cost of staffed research vessels. The goal of completely mapping the oceans’ seafloor and providing the critical geospatial context needed to address a plethora of ocean-related issues is achievable, but will require international collaboration, a long-term commitment, and the recognition that understanding Earth is worth the effort.

References

D’Amico A, Pittinger R. 2009. A brief history of active sonar. Aquatic Mammals 35(4):426–434.

Defence Council. 2011. Admiralty Manual of Navigation (BR45), vol 2: Astro Navigation. London: Ministry of Defence.

Fonseca L. 2001. A model for backscattering angular response of gassy sediments: Applications to petroleum exploration and development programs. PhD thesis, University of New Hampshire.

Hall J. 2006. GEBCO centennial special issue—Charting the secret world of the ocean floor: The GEBCO project, 1903–2003. Marine Geophysical Research 27(1):1–5.

Manda D, Thein M-W, D’Amore A, Armstrong A. 2015. A low cost system for autonomous surface vehicle based hydrographic survey. Proceedings of the US Hydrographic Conference, March 16–19, National Harbor, MD.

Marlow J, Borelli C, Jungbluth SP, Hoffman C, Marlow J, -Girguis PR, AT-36 Team. 2017. Telepresence is a potentially transformative tool for field science. Proceedings of the National Academy of Sciences 114(19):4841–4844.

Maury MR. 1855. The Physical Geography of the Sea. New York: Harper and Brothers. Courtesy of NOAA Photo Library. Online at https://oceanexplorer.noaa.gov/history/quotes/early/media/ seafloor.html.

Mayer L, Jakobsson M, Allen G, Dorschel B, Falconer R, -Ferrini V, Lamarche G, Snaith H, Weatherall P. 2018. The Nippon Foundation–GEBCO Seabed 2030 Project: The quest to see the world’s oceans completely mapped by 2030. Geosciences 8(2):63.

Smith WHF, Sandwell DT. 1994. Bathymetric prediction from dense satellite altimetry and sparse shipboard bathymetry. Journal of Geophysical Research 99(B11):21803–21824.

Smith WHF, Sandwell DT. 1997. Global sea floor topography from satellite altimetry and ship depth soundings. Science 277:1956–1961.

Stranne C, Mayer L, Weber TC, Ruddick BR, Jakobsson M, Jerram K, Weidner E, Nilsson J, Gårdfeldt K. 2017. Acoustic mapping of thermohaline staircases in the Arctic Ocean. Nature Scientific Reports 7:15192.

Voosen P. 2018. Saildrone fleet could help replace aging buoys. Science 359(6380):1082–1083.

Weatherall P, Marks KM, Jakobsson M, Schmitt T, Tani S, Arndt JE, Rovere M, Chayes D, Ferrini V, Wigley R. 2015. A new digital bathymetric model of the world’s oceans. Earth and Space Science 2(8):331–345.

Weber TC, Mayer LA, Jerram K, Beaudoin J, Rzhanov Y, Lavalvo D. 2014. Acoustic estimates of methane gas flux from the seabed in a 6000 km2 region of the Northern Gulf of Mexico. Geochemistry, Geophysics, Geosystems 15(5):1911–1925.

 


[1]  Brian Meyer, associate scientist, Cooperative Institute for Research in Environmental Science (CIRES), NCEI affiliate, personal communication by email, July 16, 2018.

 

[2]  GEBCO operates under the International Hydrographic Organization (IHO) and the Intergovernmental Oceanographic Commission (IOC) of the United Nations Educational, Scientific, and Cultural Organization (UNESCO).

 

[3]  The GEBCO_2014 Grid, the latest release of the GEBCO 30 arc-second global grid of elevations, provides an update to the previous release, GEBCO_08 Grid. It is available at https://www.gebco.net/data_and_products/gridded_bathymetry_ data/gebco_30_second_grid/.

 

About the Author:Larry Mayer (NAE) is a professor in and director of the School of Marine Science and Ocean Engineering and the Center for Coastal and Ocean Mapping at the University of New Hampshire.