Blue Earth Bathymetry 2.0
Custom terrain data for seafloor mapping<br>We developed Blue Earth Bathymetry for making attractive small-scale maps of the world ocean floor. The first version released in 2020 was a manually edited and highly smoothed version of General Bathymetric Chart of the Oceans (GEBCO), the go-to source for world seafloor data.<br>This second version of Blue Earth Bathymetry improves upon the first by merging GEBCO with BathDNN25, a new dataset based on SWOT satellite altimetry and processed with deep neural networking. SWOT has revealed 56,000 previously unknown seamounts, many in the remote southern hemisphere.<br>We produced Blue Earth Bathymetry 2.0 with the sometimes conflicting goals of maximizing ocean bottom detail and minimizing visual noise. Accuracy is greatly improved over version 1.0. These data are intended for general-purpose mapmaking, which we hope you will find useful.<br>Tom Patterson and Bernhard Jenny<br>July 6, 2026
Name
Message
0%
Some required fields are missing. Please review the form and submit again.
*er:
Submit
Blue Earth Bathymetry 2.0
GeoTIFF DEM, 21,600 x 10,800 pixels, Geographic projection, 960MB
Contact us<br>We would enjoy seeing how you have used Blue Earth Bathymetry 2.0, and hearing your ideas on how to improve it.<br>Tom Patterson<br>Bernhard Jenny
Download the data
Blue Earth Bathymetry 2.0 rendered as plan oblique relief
Blue Earth Bathymetry 2.0 is more detailed and accurate than version 1.0. Ireland is visible in the upper right.
About the data<br>Blue Earth Bathymetry 2.0 is a hybrid dataset that attempts to combine the best characteristics of GEBCO and BathDNN25.<br>GEBCO offers the highest resolution data that derive from multi-beam sonar surveys, accounting for 28% of the total coverage. It is mostly found offshore from countries with advanced economies or near features of special interest, such as the Mid-Atlantic Ridge. Most other areas derive from satellite altimetry data. Remarkably, the height of the ocean surface varies slightly depending on the bottom topography far below. Satellites can measure these minute height differences, compensating for waves and tides, to produce a coarse map of the ocean floor at ~5 km spatial resolution.<br>A problem with GEBCO dataset is the distracting linear artifacts that occur where narrow multi-beam sonar tracks cross flat areas comprised of satellite altimetry (see image below).
BathDNN25 bathymetry west of the Canary Islands. This shaded relief and the one above were rendered with 20x vertical exaggeration.
Data mosaic: Black indicates BathDNN25 and blue indicates GEBCO.
World: Ocean + Land
GeoTIFF DEM, 21,600 x 10,800 pixels, Geographic projection, 476MB
World: Ocean only
GeoTIFF DEM, 3,158 x 3,158 pixels, IBCAO Polar Stereographic projection, 40MB
Arctic: Ocean + Land
GeoTIFF DEM, 3,158 x 3,158 pixels, IBCAO Polar Stereographic projection, 15MB
Arctic: Ocean only
Terms of use<br>You may use Blue Earth Bathymetry 2.0 free of charge provided that you cite the sources from which it derives: GEBCO and BathDNN25.<br>Acknowledgment<br>We developed Blue Earth Bathymetry 2.0 with funding from the National Geographic Society. They generously granted us permission to share it with the wider cartographic community. Thank you.
Blue Earth Bathymetry 1.0 is available here.
BathDNN25 is new satellite altimetry dataset with a spatial resolution of ~1,800 meters. It derives from data collected by the Surface Water and Ocean Topography (SWOT) mission, a US/French collaboration. The developers of BathDNN25, based in Australia, processed SWOT using deep neural networking to map ocean bottom features with more detail and improved accuracy compared to previous satellite altimetry data. Flat areas are free of linear artifacts (see image below). BathDNN25 coverage extends from 81° S to 81° N. The northernmost Arctic Ocean is unfortunately absent from the dataset.
GEBCO bathymetry with multi-beam sonar tracks west of the Canary Islands.
Presumed artifacts deleted from IBCAO Arctic Ocean data.
Combining GEBCO and BathDNN25 was accomplished graphically. The first step using Eduard software involved downsampling and exporting GEBCO at 60-arc second resolution to match the resolution of BathDNN25. Both 32-bit GeoTIFF DEMs were then imported and mosaicked as layers in Photoshop using the Geographic Imager GIS plugin. GEBCO was on the bottom layer and BathDNN25 was on the top layer. The top layer received a layer mask. Placing modulated black and white tones—consisting of slope and depth renders with fuzzy edges—in the layer mask allowed the two datasets to be composited selectively. BathDNN25 predominates in deep ocean basins and GEBCO is found elsewhere (see image below). The layer mask also received manual edits in critical areas using Photoshop's brush tool.
The Arctic Ocean bathymetry also availabe here is an edited version of International Bathymetric Chart of the Arctic Ocean (IBCAO), a companion dataset to GEBCO. Only a few...