[kajecounterhack]

Radio is part of the EM spectrum. EM waves (including light) decay under water at an exponential rate due to absorption by the medium. So data transmission underwater must rely instead on mechanical waves (sound).

This is why underwater robots mostly use tethers, otherwise you couldn't control them very well (RC control would stop working at a very shallow depth).

Note: why are folks downvoting an honest question?

[marche]

> (including light)

I remember being told in some physics class that visible light is actually the radiation which penetrates the most, and that indeed this may be the very reason why it's visible: eyes were developed when life was still aquatic, therefore they evolved to be sensitive to the range of frequencies that could reach them.

[kajecounterhack]

Yeah it does penetrate the most, so you do see some communications utilizing visible light comms. But the bitrate is bad and gets worse very quickly as you go deeper. Like, sunlight will penetrate the water a lot especially since there's so much of it coming down broad spectrum, but shoot a laser into seawater and you'll find the beam dissipates rather quickly.

It's mostly water-to-surface comms that can tolerate being in the very top of the water column that utilize this tech. Unfortunately aligning the transmitter/receivers is a bit tricky -- a lot of research has gone into better ways to send and receive information from air into water.

I guess it makes sense if eyes developed aquatically first, but it doesn't seem that interesting in the sense that most things developed aquatically first. Like even if it didn't develop aquatically, seems like evolution would have found its way on land, where sight is more useful, especially since you can't use things like lateral lines to detect pressure fluctuations (I mean there's terrestrial hearing, but seems like being immersed in a dense medium makes pressure sensing that much more important).

[potatoman22]

Don't EM waves decay at an exponential rate in all mediums?

[Filligree]

No, usually it’s polynomial. Radius squared, that is.

[msandford]

Sure but what's the exponent?

[hutzlibu]

"Note: why are folks downvoting an honest question?"

I suppose because of Friday.