Marine Sound Propagation Calculator

Across this calculator's full -2°C to 35°C range, raising Water Temperature raises Speed of Sound at any fixed Salinity and Depth -- warmer water is more compressible in a way that lets pressure waves propagate faster, which the Mackenzie equation's temperature terms capture empirically. This is why tropical surface water carries sound noticeably faster than the same water mass would at polar temperatures, all else being equal.

Neither one dominates uniformly. Over this calculator's full ranges, Depth (0-11,000 m) and Water Temperature can each shift Speed of Sound by well over 100 m/s at the extremes, while Salinity's 0-42 PSU range contributes a smaller swing on its own -- so which input matters most depends on where you are in the range, not on a single fixed ranking.

That's intentional, not a bug: this output is a simplified heuristic that estimates the sound channel axis purely from how far Water Temperature departs from about 4°C, since that's the dominant driver of where the sound-speed minimum sits in most open-ocean profiles. It does not read Salinity or Depth at all, so adjusting either one leaves this particular output unchanged while Speed of Sound itself still responds normally.

No -- it's computed for a fixed 1 kHz reference frequency using a simplified Francois-Garrison-style formula that, in this implementation, depends only on that fixed frequency. The real underwater-acoustics literature ties absorption to temperature, depth, and pH as well through chemical relaxation effects, so treat this output as a rough reference value rather than a site-specific measurement.

Yes -- the nine-term formula and all published 1981 Mackenzie coefficients, from Kenneth V. Mackenzie's paper in the Journal of the Acoustical Society of America (vol. 70, pp. 807-812), are reproduced exactly. The original equation was fitted and validated over 2-30°C, 25-40 PSU, and 0-8,000 m, while this calculator accepts a wider range (down to -2°C and out to 11,000 m), so values near or beyond those calibration bounds should be read as a reasonable extrapolation rather than a lab-verified figure.