This content was originally posted to the Trek View blog on 2021-12-10. Some of the information found in it may now be outdated.
I was recently snorkelling with the camera no more than a few centimetres below the ocean surface.
I wanted to see how deep the camera could be submerged before losing a stable GPS lock.
The science
When electromagnetic waves impinge upon a conductive material (water in this case), currents are induced on and inside of the material that work against the incident electromagnetic wave.
This causes an attenuation of the wave as it travels through the material. There is also a power loss going into the material due to reflection. Going from one medium to another of a different conductivity, permittivity, and/or permeability causes some of the wave to be reflected back.
The skin depth, which is the distance that the wave must travel to be reduced by e^-1 in amplitude, is dependent upon frequency and conductivity.
So a very low frequency signal can penetrate the earth several hundred metres while the high frequency GPS cannot penetrate a few feet. This is why submarines use extremely low frequency radio frequencies (tens to 100’s of khz) for communications.
Specifically, sea water has a conductivity of 5 S/m. GPS has a frequency of around 1.3 GHZ. So a GPS signal has a skin depth of 6 mm. That means that the signal loses 36.7% of it’s amplitude after only 6 mm. The signal is reduced to 1% of its amplitude after only 2.87 cm.
Let’s put the science to the test…
Equipment used
A GoPro MAX with a water-proof cover for the camera (here’s the one I bought via Amazon). The MAX is waterproof, but I’m always a little skeptical 
Test conditions
Some things to note in these results:
- They are far from scientific. All measurements are rough estimates
- I took more than ten photos at each depth to get a wider sample, only selected images are shown below
- The conditions were perfect on the day (clear skies)
- I don’t know where the GPS receiver is in the camera (which might alter measurements by up to 1cm, but this is negligible – see first point on accuracy)
A note on lat/lon/alt values
Altitude can often appear particularly inaccurate. New GPS buyers often suspect their equipment may even be defective when they see the altitude readout at a fixed point vary by many 10’s of metres. This is normal.
With most low cost GPS receivers, the horizontal error is specified to be within about +/- 15 metres (50 feet) 95% of the time... Generally, Altitude error is specified to be 1.5 x Horizontal error specification. This means that the user of standard consumer GPS receivers should consider +/-23meters (75ft) with a DOP of 1 for 95% confidence.
Test results
Surface
GPS Altitude : 44.4 m Above Sea Level
GPS Date/Time : 2021:12:04 16:07:32Z
GPS Latitude : 28 deg 42' 42.47" N
GPS Longitude : 14 deg 0' 40.97" W
Summary:
GPS positions reported were all stable across all photos at this depth.
Roughly 5 cm below surface
GPS Altitude : 43 m Above Sea Level
GPS Date/Time : 2021:12:04 16:07:55Z
GPS Latitude : 28 deg 42' 42.61" N
GPS Longitude : 14 deg 0' 40.78" W
Summary:
GPS positions reported were stable, although 25% of photos at this depth had no GPS position reported.
Roughly 10 cm below surface (Seawater test)
GPS Altitude : 40.5 m Above Sea Level
GPS Date/Time : 2021:12:04 16:08:51Z
GPS Latitude : 28 deg 42' 42.53" N
GPS Longitude : 14 deg 0' 41.55" W
Summary:
GPS positions reported were stable, although 70% of photos at this depth had no GPS position reported.
Roughly 30 cm below surface (Seawater test)
Summary:
No GPS reported in any photos at this depth.
Test summary
Depths of 5-10cm below the surface of the ocean causes the cameras GPS receiver to start failing significantly making it useable.
In summary, your MAX is great for snorkelling where you can keep your camera on, or slightly below the surface. However, you’re going to need another way of tracking GPS if you intend to go on deeper.
A final note on freshwater
Perfectly pure water has a very low conductivity, practically zero. However, almost all water has various impurities and ions that raise the conductivity.
So if we assume that fresh water has a conductivity of 0.005 S/m, then the skin depth of GPS is about 20 cm. So the amplitude is reduced to 1% after 90 cm.
It was previously noted, GPS signal is reduced to 1% of its amplitude after only 2.87 cm in saltwater. So a rough calculation would assume depths of 31 times (90 cm / 2.87 cm) that of seawater are possible in freshwater. Doing the maths, 31 x 5cm = a depth of 1.55 m (though I’d expect the GPS to be incredibly inaccurate at this depth).