How to identify cable strumming in underwater recordings?

Question

Cable strumming (Section 4.3.2 in Practical Guide for Underwater Noise Measurement):

Cable strum occurs when cables are pulled taut by the action of currents, and the cable is then caused to vibrate by the action of the water flow around it, producing parasitic low frequency signals.

Figure 7b in Dziak et al 2015 (see below) show a 'zoomed out' image of strumming along with other noise. For someone who has no experience with strumming, it can be difficult to identify and I have not been able to find a visual/descriptive explanation online to help someone identify this in their own recordings.

Question: How can you visually identify if you have cable strumming on your recording? Are there spectral/temporal characteristics that can help someone identify strumming and differentiate it from other biological or physical sounds?

Answer 1

If you follow the reference trail from the good practice guide above you will find a useful discussion on the characteristics and mitigation of cable strum in Urick classic 1983 textbook. This is also discussed in a 1984 US Navy report available here on ambient noise in the sea [AMBIENT NOISE IN THE SEA][1] and for even more detail from there it references if you can access papers from the Journal of the Acoustical Society of America.

Strasberg, M., Nonacoustic Noise Interference in Measurements of Infrasonic Noise, JASA 66, 1487, 1979.

McGrath, J. R., 0. M. Griffin and R. A. Finger, Infrasonic Flow-Noise Measurements Using an H-58 Omnidirectional Cylindrical Hydrophone, JASA 60, 390, 1977.

Hope that is some help.

Answer 2

An interesting problem. From the physics, I would assume that the cable length gives you the fundamental frequency (length = wavelength/2) and cable strumming would show up as harmonic lines. However, you may need to down-sample your data significantly. Taking you figure as an example, I would down-sample to 10 Hz bandwidth, and look for harmonic lines.

Caveat: I never have done it, but this is where I would start.

Edit: Taking Urick (1983) from the shelf one finds on Page 370 a better explanation:

This form of self-noise occurs in a current of water, and is the result of cable vibration induced by the eddies or vortices shed by the cable. This is the "aeolian harp" effect, or the singing of telephone wires in the wind, that has long been known in air acoustics. The frequency of vortex shedding is given by the simple expression f = S * v / d, where S is the dimensionless "Strouhal number," v is the water current speed, and d is the cable diameter in the units of length of v. The Strouhal number happens to be a constant equal to 0.18 over much of the range of current speeds and cable sizes occurring in water. This with a 1-cm-diameter cable in a 1-knot current (51.5 cm/s), the strumming frequency will be 9 Hz. Strumming noise can be readily alleviated by a number of means, including using a faired cable, keeping the natural frequency of the cable vibration well separated from the strumming frequency, isolating the hydrophone from the cable (as by such simple means as suspending it from the cable by rubber bands), and employing a hydrophone having an acceleration cancelling design.

I hope this quotation helps for those who have no access to Urick.