Aquarium Acoustics

Sometimes having real bathymetry data complicates things more than it helps. Say you want to have a look at your noises in a more controlled environment. This could be to teach students about sound propagation, or for finding out if building an underwater noise barrier around a shipping channel would actually make a difference. I'm going ahead with my traffic noise barrier thought.
Figure 1: A traffic barrier saving stickman's hearing. Picture from 
One could imagine a barrier like this (Figure 1) around shipping channels, but as underwater acoustics are not entirely the same as air acoustics, we run into some problems.
Because the speed of sound in water is about 4.5 times faster than in air, all wavelengths are longer by the same factor. This means that for similar frequencies to be blocked, the walls need to be 4.5 times thicker than their airy counterparts.
Another big difference relates to the physics of reflecting sound. Relatively thin, but hard and heavy barriers work well in air because the high impedance mismatch leads to reflection of the sound. This reflection is harder to obtain under water, as water has a high impedance already. One can use bubble-nets to get that impedance mismatch, but as they are active devises, they can be hard to maintain.
Lastly, the fact that the surface reflects sound means that sound can go "over" a wall much easier.

So how does one go about modelling this?

Because bathymetry data is often just a raster file, it is easy to make your own personal bathtub. I've used a spreadsheet editor to make the following:
Figure 2: My very own channel scenario. I just put some numbers in a spreadsheet and saved as .csv
This can then be imported directly into dBSea (yes you can also make a smiley).
So how much does a barrier actually affect the sound propagation? Note here that I've built the barrier from water saturated gravel, and this is taken into account when calculating the sound levels. If you change the material the attributes of the barrier changes as well.
Figure 3: Neither the wall or the poles/spikes seem to do much dampening of the sound.
In the picture above the barrier does not seem to make much of a difference, but as usual there's more to it than meets the eye. In the figure below you can see how a wall might completely eliminate higher frequency noise escaping a shipping channel. Notice also how the higher frequencies are dampened by the spikes/poles in the right side of the plot.
Figure 4. All noise under 250 Hz propagates unhindered through the barrier, but the more biologically relevant noise over 250 Hz is eliminated completely. Notice how the spikes diffuse and dampen high frequencies as well. Red dots are probe positions for the spectrograms in the top figure.
I'm not currently aware of anywhere where this tactic for noise mitigation is being used, but maybe it would be a viable idea for some of the areas where busy harbours and noise sensitive wildlife coincide.

So to round up: If you have a problem that's more theoretical or too complicated to run in a full scale environment, get your inner bathymetry architect out and make a simplified model.

Bonus: Using the probe menu can give you valuable information on the frequency dependent spreading of noise.

Thanks for reading