Satellite-derived bathymetry is a way of estimating seabed depth from space instead of with ships. In the open ocean it is worked out from tiny bumps in the sea surface caused by gravity. In clear shallow water it can come from how much light reaches the bottom.
Key takeaways
- Satellites cover the whole seafloor, but give only a general picture, so detail is limited and some features stay unseen.
- In deep water the depth is predicted from satellite measurements of gravity, which pulls the sea surface into slight bumps over seabed relief.
- The GEBCO_2026 grid behind this map uses such predictions as its base and adds ship-survey data on top.
- A satellite estimate is blurrier than a sonar survey, so treat it as a guide to the big picture.
How satellite-derived bathymetry works
Ships can only map the sea under and beside them, so for most of the ocean the first map comes from space. The whole seafloor has been mapped using data collected from satellites, but those data give only a general picture of what is there [2]. Detail is limited, so some important features, such as seamounts, and objects, such as shipwrecks, remain unseen [2].
There are two separate methods that share the name.
Gravity and sea-surface height. The sea is not perfectly level. Rock is heavy, so a large undersea mountain pulls a little extra water towards it, and the sea surface rises into a very slight bulge above it. Over a trench the surface dips. Satellites measure these tiny bumps and a computer turns them into depths. The GEBCO_2026 Grid is built on this approach for much of the world: its base dataset is a fusion of land topography with measured and estimated seafloor topography, with the estimates being predictions based on satellite-derived gravity data [1]. The latest release uses a new gravity field from the Surface Water and Ocean Topography (SWOT) satellite and machine learning methods [1].
Light through clear water. Close to the coast, satellite images can show the bottom. When water is clear and the seafloor is sufficiently reflective, depth can be estimated from how much light the satellite sees and how far light should penetrate [3]. This works in shallow, clear water only.
Because the first method is so indirect, it is blurry. It picks out the big shapes, such as ridges and large seamounts, but it cannot resolve small ones. On the homepage's how the seabed is mapped section you can see how the satellite estimate is combined with sonar, and the accuracy section says what that means for the detail you can trust.
An example from our data
The Mid-Atlantic Ridge is a long, broad feature, exactly the kind a satellite picks up easily. The deepest cell the grid finds near it reads 1,477 m. Large features like this stand out in satellite-based maps, which is why the ridge and the East Pacific Rise look so clear even where ships rarely go.
Compare an area of the open Pacific Ocean, mean depth 4,315 m. Away from ship tracks, the map there is mostly a satellite prediction. The grid still has a value every 15 arc-seconds, but the real information behind each value is much coarser.
The thing people get wrong
A common assumption is that a satellite sees the seabed the way a camera sees the ground. In deep water it does not. It measures the sea surface and works the depth out, which is why it fails to spot narrow or small features. The idea that a map with a value every few hundred metres must be detailed is also wrong, as the spacing of the grid is not the same as the sharpness of the data.
Satellite estimates vs sonar surveys
| Satellite-derived | Sonar survey | |
|---|---|---|
| Coverage | Whole ocean | Only where a ship has been |
| Detail | General picture, small features missed | High detail |
| How it works | Predicts depth from the sea surface | Measures depth by sound |
Frequently asked questions
Can satellites see the seabed?
Not directly in deep water. They measure the height and gravity of the sea surface, and a computer turns that into an estimate of the seabed below. In clear shallow water, some satellites can see the bottom as light.
Is satellite-derived bathymetry accurate?
It shows large features well but not small ones. In the deep ocean it cannot show anything much smaller than a few kilometres across.
Why do we need ships if we have satellites?
Because ship sonar gives far more detail. Satellites give complete coverage, ships give sharpness.
Sources
- GEBCO_2026 Grid, GEBCO.
- How much of the ocean has been explored?, NOAA Ocean Service.
- Bathymetry, Wikipedia.
- Every figure from our own map is worked out from the datasets on the data sources page (checked against GEBCO_2026 grid).