A very brief summary of what I've found in the literature with respect to liquid sounds in the air. A proper survey document is to come.
- Almost all the sound is generated by bubbles. There are 'slapping' noises related to impacts.
- While water/air transmission is very poor, bubbles generated near the surface enjoy better transmission. Bubbles on the surface should have, of course, even better transmission, but I haven't seen any analysis of this.
- Bubbles in liquid have a analytical solution for their vibrations which agrees with experiment fairly well. The solution treats the vibrations as a damped oscillator where the mass, stiffness and damping of the oscillator are related to the liquid properties and bubble size. A summary of the solution can be found on this page.
- Turbulence in the fluid can generate sound waves, such as in a jet engine, but more commonly it excites existing bubbles to generate sound. An example of this is the gurgling garden hose, apparantly.
- One case which has been looked at fairly closely is the impact of a water droplet onto a flat water surface. There have been a few numerical studies of the shape of the impact cavity and lots of attention on the entrainment of a bubble by the impact. You can generate convincing underwater spectra of rain from these studies.