Measurement Issues: dB in Air and Water
From an AEI report on seismic airguns, which contains footnotes and bibliography, detailing all references .
[DOWNLOAD REPORT (rtf)]
First off, a primer on the decibel system: decibels do not measure an absolute amount of sound, but rather represent a proportional increase above an arbitrary reference level of sound intensity. Each increase of 10dB represents a 10-fold increase in the sound’s intensity; thus 140dB is ten times more intense than 130dB, and 150dB is one hundred times more intense than 130dB. However, 140dB it is not ten times louder than 130dB; our perception of relative loudness decreases as sound intensity levels go up. To some degree, this matches the logarithmic scale of decibel measurement, so that a sound of 150dB will tend to sound about half again as loud as a sound of 100db. However, there is much individual subjectivity in perceived loudness, as well as no real knowledge of how other species perceive loudness. Also, when considering a sound’s impact on sensitive acoustic perceptual systems (i.e., ears or other organs and systems), the relative intensity is perhaps more important than the perceived loudness.
Comparing dB measurements in air and water involves two different measurement corrections. One of them is purely a numerical shift, while the other is more complicated, involving both mathematical and physical differences. The result of both corrections can at times mislead laymen into overestimating the impact of sound in the water. Adding to the situation is the fact that our tools perceive specific physical qualities of sound waves, and our measuring systems then abstract this information into quantifiable values, while the experience of sound (by either humans or other animals) is far more complex, involving physical responses that are more diverse, subtle, and integrated than those captured by our sound-measuring tools, as well as subjective responses to sounds that are both individual and unquantifiable. Despite these uncertainties, there are some straightforward and important corrections that scientists use in order to make the measurement of sound in water more closely align with the physical experience of the listener.
The first adjustment is a simple 26dB difference, related to the arbitrary reference level from which the sound intensity is measured. The air reference level is "dB re 20uPa" (uPa is a unit of measuring sound pressure, the micro-Pascal), which is the limit of human audibility (that is, humans can only hear sounds that have a pressure of at least 20uPa). In water, where the human hearing system is inefficient (because our eardrum does not respond well to the added pressure), it doesn’t make sense to count dB in relation to human audibility threshold, so it is instead measured from an arbitrary level of 1uPa. Since micro-Pascals and dB are related logarithmically, this means that a sound of a given intensity will be measured as being 26dB higher in water than in air. So (considering only this first correction factor), a sound measured at 126dB in water will only be as loud as a sound that measures 100dB in air. The two sounds are experienced identically (at least theoretically: air-adapted animals will in fact hear water in sound less effectively, and vice versa; and, every animal is adapted to hear a different range of freqencies best).
The second adjustment is a bit less cut and dried. Because water is much more dense than air, water has higher impedance, so sounds of equal measured pressure will (because of the physics and math of impedance), be measured at 36dB higher in water. However, this time, the difference is not purely mathematical, as in the reference pressure correction. To some degree, the sound as experienced in water will feel louderif you were in a bathtub listening to sound at 90dB, and then moved the speaker underwater, it would be measured as 126dB; unlike the previous correction, though, the sound would indeed feel stronger when experienced underwater, as well as being measured louder. Since this is a subjective experience, it is very difficult to quantify to what degree the 36dB measurement difference is actually experienced physically as an increase in sound intensity.
Since subjectivity is such a slippery slope, and researchers are by nature comfortable with the abstractions of measurements and mathematics, it is generally accepted in ocean acoustics that sounds of equal pressure (in the respective reference units) can be considered 62dB higher in water than in air (26dB plus 36dB). This means that when a sound is reported as 206dB in water, it will correspond to a sound measuring 140dB in air. When trying to imagine the impacts of sounds reported in ocean acoustics studies, it is crucial to keep this correction in mind (while also remembering that the subjective experience of the sound may be a bit different than the correction implies).