[jlopez2022 (OP)]

I need to use ultrasounds to communicate with sea robots and use them using sensors.
I am surprished because there is very few information and unuseful about that topic, I only need the maximum levels in the frequency spectrum to avoid disturb animals.
I supose that the way is to use high directionable waves, low levels and frequency over 100khz (as long as high frequency ones are attenuated at shorter distances and animals does not hear high frequencies)
Perhaps usingactual levels used by some animals lowered 6 decibels an be used taking in account if there are any dolphin in the area.
I do not have data about fishes and other animals.

[evan_au]

using actual levels used by some animals lowered 6 decibels

This sounds like a good approach.

Be careful not to equate peak power with average power.
- Echo-locating creatures like dolphins and whales emit high peak levels, and then listen for relatively quiet echoes
- The same goes for communication with other members of their species - the responses are much quieter than the call.

So you could use pulsed communication with a similar ON/OFF cycle, or use continuous transmission 6dB below the average power emitted by these creatures.

I understand that one of the noisiest ocean creatures for their size is shrimp which create ultrasonic shock waves with their snapping pincers.

[Colin2B]

You might find the following site useful. You might try contacting them for more info or advise.

http://www.dosits.org/animals/effectsofsound/howdoyoudetermineifasoundaffectsamarineanimal/

They used to have the following tables on their site but I can't find them anymore. Cut and paste from my own notes don't seem to work well.
Depending on the range you need and what noise rejection your modulation system can deliver, I suspect you can get away with much less than these levels.

Examples of manmade noise
Ships underway Broadband source level (dB re 1μPa at 1m)

Tug and barge (18 km/hour) 171

Supply ship 181 (example: Kigoriak)

Large tanker 186

Icebreaking 193


Marine-mammal soundlevels

Source.         Broadband source level (dB re 1 μPa at 1 m)

Sperm whale clicks.  163 – 223

Beluga whale echo-location click.    206–225 (peak to peak)

White-beaked dolphin echo-location clicks.     194–219 (peak to peak)

Spinner dolphin pulse bursts.   108 – 115

Bottlenose dolphin whistles.    125 – 173

Fin whale moans.  155 – 186

Blue whate moans.    155 – 188

Gray whale moans.    142 – 185

Bowhead whale tonals, moans and song.    128 – 189

Humpback whale song.  144 – 174

Humpback whale fluke and flipper slap.  183 – 192.   

Southern right whale pulsive call.   172 – 187

Snapping shrimp.     183–189 (peak to peak)

[jlopez2022 (OP)]

I will be more happy using >200khz ultrasounds, as long as attenuation is only double, so better also and animals can not hear that.  Also I would have more bandwidth to transmit. Perhaps if all sonars would work from that frequency and at a maximum levels by law the mammals would suffer nothing with that

[Colin2B]

Attenuation is frequency dependant - greater for higher frequencies. Dolphins, for example, need higher frequencies for greater target resolution but only at very short range. Military sonar on the other hand is seeking larger targets so can use lower frequencies to achieve greater range.
Attenuation will also depend on where transmitter and receiver are located. If both are relatively shallow you might be able to use the surface waveguide effect, the disadvantage of this is that propagation is then dispersive and that could cause problems for wideband transmission. Also in the surface layer bubbles cause additional attenuation for higher frequencies because of their resonance and scattering effects.
Your transducer supplier should be able to advise based on your exact application.

[evan_au]

The February 2016 edition of IEEE Communications magazine had a theme on underwater communications.
And it appears that there was a 2016 conference on underwater communications.

I don't think it directly answers your question about the best frequencies to minimize harm to wildlife, but it certainly addresses a number of issues you will strike, like attenuation, multi-path propagation, background noise, etc

http://ieeexplore.ieee.org/search/searchresult.jsp?newsearch=true&queryText=Underwater%20Communications
Note that some of these articles will be behind a paywall, but you can see the abstracts for free, and your organization may have an institutional membership.