After centuries of ocean exploration, it may surprise many that only a fraction of the ocean seafloor has been mapped in detail. Since the early 1930s, ship-collected echo-soundings have been used to map the seafloor. Ship-collected bathymetric information can be divided in two categories: soundings obtained during dedicated bathymetric surveys of specific areas, and “en-route” soundings, concentrated along shipping lanes. As a result, large areas of the world ocean remain mostly void of any ship soundings (Figure 4.1).
Since the advent of satellite altimetry, satellites have also been used to map the seafloor topography, since they present the huge advantage of covering the whole globe (sometimes with the exception of polar regions). Although satellites only measure sea surface levels, it is possible to relate changes in sea level to small variations in the Earth’s gravitational field, and therefore to topographic features. This conversion to bathymetry relies on appropriate geologic assumptions locally calibrated by ship echo-soundings.
The combination of ship sounding and remote sensing from satellite has finally allowed a complete coverage of the world oceans in terms of bathymetry. However, it should be pointed out that not all areas are equally well documented.
The bathymetric shown in the Ocean Potential tool is based on bathymetric data from the GEBCO_08 and ETOPO1 datasets. For the coast of the United States of America, the tool uses bathymetric information from the U.S. Coastal Relief Model.
The GEBCO_08 (General Bathymetric Charts of the Oceans) dataset is a 30-second gridded bathymetry dataset maintained by the British Oceanographic Data Centre (BODC). The ETOPO1 dataset is also a gridded product of global bathymetric data with a 1-minute horizontal resolution maintained by the National Geographic Data Center (NGDC) from the National Oceanic and Atmospheric Administration (NOAA). It replaces earlier gridded products with coarser horizontal resolutions (ETOPO2v2 and ETOPO5). Both products are built from the combination of ship track soundings and satellite altimetry. In both cases, a so-called bathymetric model interpolates the satellite-derived gravity data between soundings. In addition, additional data sources are also used in certain areas where they improve on the existing grid.
The U.S. Coastal Relief Model is also provided by the NGDC and only focuses on the coast of the U.S. East and West coasts, the northern coast of the Gulf of Mexico, Puerto Rico, and Hawaii. The dataset aims at integrating offshore bathymetry with land topography into a seamless representation of the coast with a horizontal resolution of 3 arc-seconds. Depending on the areas, the dataset extends up the continental slope and in places even beyond. It is built using data from satellite altimetry and ship soundings.
Data and data description are available at:
- Amante, C. and B. W. Eakins, 2009. ETOPO1 1 Arc-Minute Global Relief Model: Procedures, Data Sources and Analysis. NOAA Technical Memorandum NESDIS NGDC-24, 19 pp, March 2009
- British Oceanographic Data Centre (BDOC), 2008. The GEBCO_08 Grid, General Bathymetric Chart of the Oceans (GEBCO). http://www.gebco.net/
- NOAA National Geophysical Data Center, U.S. Coastal Relief Model, 2013-10-15, http://www.ngdc.noaa.gov/mgg/coastal/crm.html
Performance and limitations
Ship soundings (especially recent ones using multibeam sonars) are the absolute reference in terms of reliability of the bathymetry. However, such high-resolution and high-quality data are of course not available for the entire world. Because it relies so heavily on such ship soundings, bathymetric data from the U.S. Coastal Relief Model is very reliable. In addition, the requested accuracy of ship soundings carried out from 1957 onwards is known and documented.
The ship tracks used in GEBCO_08 and ETOPO1 are not necessarily the same, and the interpolation techniques between ship tracks also vary, so that both products are difficult to compare accurately on a global scale. In some areas, one product may perform slightly better on certain aspects than the other. However, specific detail comparisons for selected areas have been carried out.
The overall quality of both products relates directly to the present-day challenges with respect to employing satellite altimetry for measuring bathymetry. At this moment, the conversion of gravity to topography works best in the deep ocean and for relatively smooth bathymetric features. In addition, it requires soundings for calibration of the conversion because of its relation to the local geology. Figure 4.2 presents an example (located to the southeast of Iceland) of the result of the combined use of detailed ship soundings and interpolation in less well documented areas for the GEBCO_08 bathymetry. The gridded dataset shows areas that seem to present a much higher degree of detail, and other areas that are relatively smooth. The detailed area corresponds to the area for which multi-beam bathymetry data is available, whereas the smoother area has been interpolated using remote-sensing data. In this smoother area a few thin lines may be visible, representing single-beam echo-sounding lines used to constrain the satellite-predicted bathymetry (GEBCO, 2013).
As a result, specific comparisons of gridded bathymetric products such as GEBCO_08 and ETOPO1 to actual bathymetries conclude that gridded products generally perform quite well in deep areas, but that their overall quality rapidly degrades in areas of high rugosity and in shallow areas (Debese and Lequentre-Lalancette 2009). In general, it typically translates to islands having incorrect shapes and sizes in these gridded datasets, or to the bathymetry lacking features or on the contrary exhibiting artificial features. A more specific comparison of (among others) GEBCO (the 1-minute resolution version) and ETOPO1 was carried out by Mleczko et al. (2010) for the Australian island of Lord Howe, located between Australia and New-Zealand. Both the ETOPO1 and GEBCO dataset missed one seamount present in the area. The exact position and size of the islands on the shelf was incorrect in both datasets, and the depth according to ETOPO1 was 150 m too shallow (see Figure 4.3). Unfortunately, no specific comparison of ETOPO1 and GEBCO_08 could be found in literature, but considering the way these gridded products are built, it is expected that the main conclusions would not differ much.
However, it should be kept in mind that the islands chosen for this particular example are located in a more remote area of the global ocean that has not been surveyed very extensively. In addition, efforts are constantly being undertaken in order to improve the overall quality of these products (GEBCO for instance) in coastal areas using nautical charts. The map in Figure 4.4 shows an overview of the areas subjected to such improvements as of September 2013 for the GEBCO_08 dataset.
The bathymetric information shown in the pre-assessment tool is presented for general information purposes. Although the presented depths are generally quite reliable, inaccuracies may still be present, especially in the shallower parts and in the details of the shelf slope. For specific design aspects for which a detailed knowledge of the seafloor is required, it is essential that a dedicated, specific bathymetry survey be undertaken.
Debese N., Lequentre-Lalancette M.F., 2009, Analysis of bathymetric datasets quality, Gebco science day, sept 2009, Brest.
Mleczko R., S. Sagar, M. Spinoccia and B. Brooke, 2010. The Creation of High Resolution Bathymetry Grids for the Lord Howe Island Region. Geoscience Australia report. Record 2010/36. GeoCat #70649.
U.S. CRM. http://www.ngdc.noaa.gov/mgg/coastal/model.html
Sandwell, D.T., W.H.F. Smith, S. Gille, S. Jayne, K. Soofi and B. Coakley, 2001. Bathymetry from Space: White paper in support of a high-resolution, ocean altimeter mission. Available for download at http://topex.ucsd.edu/marine_grav/white_paper.pdf