Abstract

An analysis technique is described that permits the classification of ambient noise sources (wind generated and shipping) without the need for information concerning the environment or shipping in the area of interest. Wind-generated and ship-generated ambient noise, or a combination of the two, can be classified by finding the zero-axis crossing time of the autocorrelation function versus frequency of the long-term (greater than two weeks) ambient noise time series for a given geographic area. Results obtained from the zero-axis crossing times of the autocorrelation function of ambient noise level versus time (measured by omnidirectional hydrophones and processed through 1/3-octave filters) are bounded by two distinct zero-axis crossing times. Analysis of data from four widely separated geographic areas produces the same bounds, which suggests that they may be independent of area. In the region of the spectra where 1/3-octave bands were highly correlated with wind speed, the decorrelation time (i.e., zero-axis crossing) was found to range from 26 to 40 h. In the region of the spectra that was not wind correlated and, therefore, attributed to ship-traffic noise, the decorrelation time was found to range from 4 to 8 h. For spectra that were composites of both noise sources, the decorrelation time ranged between 8 and 26 h. The observed correlation functions were fitted to a two-term equation to aid in determining the decorrelation times for all values of the wind-generated to ship-generated noise-level ratios. Each term of the equation (one being wind related and the other ship related) is a sinusoid whose decay parameter is equal to its corresponding mean value decorrelation time and whose sinusoidal period is four times that value. Subject Classification: 30.70, 30.80.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call