Page 64 - Summer2022

RAINFALL AT SEA Figure 2. The spectra of ambient sound during times of only rain or only wind. The data were collected from 14 open ocean locations over 90 months (Ma and Nystuen, 2005a), © American Meteorological Society, used with permission. present can be identified. For example, the noise generated by the wind is the major persistent noise component over the frequency range from 1 to 50 kHz, characterized by a simple spectrum that decreases with increasing frequency by about 16 dB per decade. When rain is present, however, the sound it generates usually dominates all other sources in that frequency range. When measuring rain, the dominant sound source is a signal that can be used to determine the properties of the rain. For example, the sound spectra of ambient sounds from wind only, rain only, and combined rain and wind are distinctly different (Figure 2) (Ma and Nystuen, 2005a). Importantly, the separate spectra of wind and rain in a combined spectrum can be distinguished by considering spectral slopes and relative spectral levels across different frequency bands. Effective Listening Area The underwater noise caused by wind and rain comes from the ocean surface. The sound intensity at a par- ticular depth below the surface is a summation of all the contributions of sound created at the surface. The rain- fall rate can therefore be obtained from a measurement of sound at depth because that sound is composed of all the sounds from the surface. In such a measurement, the sound sources are assumed to be uniformly distrib- uted over the surface. If the absorption and refraction of sound are neglected (in practice, just minor corrections), the rainfall measurement is independent of depth. Although the summation of surface sounds is theoreti- cally over the entire ocean sea surface, as a practical matter, the effective listening surface area has a radius only three times the hydrophone depth; sounds from beyond that radius make only minor contributions. Thus, an instrument located at a 100-m depth samples an area with a radius of 300 m or roughly 0.28 km2. Importantly, the measurement is inherently integrating over area, pro- ducing a spatially averaged rainfall statistic. In addition, because the total sampling period of a single measure- ment is about 1 min, the sound from many individual raindrop splashes is quickly accumulated, providing a robust measurement (Nystuen, 2001). Types of Rain Sounds: It’s All About the Bubbles Raindrops hitting the ocean surface generate acoustic sig- nals in two ways: the impact on the surface and the tiny bubble entrained by the drop and pulled below the surface. Surprisingly, the bubble is the loudest sound source for most raindrops, not the impact. When a bubble is cre- ated, the pressure inside is not in equilibrium with the surrounding water. During the impact process, a bubble is pushed and compressed. The pressure of the trapped air increases as the bubble shrinks by these forces and it becomes higher than that of the water. After shrinking, the bubble then expands, decreasing its pressure. In this way, a bubble oscillates between high and low pressure, a rapid process reaching an equilibrium at a high frequency and creating a unique sound at the bubble’s resonant frequency. The bubble is “ringing,” much like a bell rings. The sound radiates energy so the fate of the bubble is to lose its energy and reach equilibrium with the surrounding water. The next thing to keep in mind is that the resonance (ringing) frequency of a bubble depends on its radius and the local pressure and water density. The resonance frequency is inversely proportional to the size of the bubble, so the smaller the bubble, the higher the reso- nance frequency (the higher the pitch of the sound). This quite accurate relationship was defined nearly 90 years ago by Minnaert (1933). Of particular importance in the context of a gauge for ocean rain, bubble size is determined by drop size. If 64 Acoustics Today • Summer 2022