Empirical modelling for derived metrics as function of sound exposure level in marine pile driving

Abstract

To protect marine life, piling activities at sea are subject to regulations on sound emission. Different regulatory authorities base different action criteria based on different statistics of the emitted sound pressure. Modelling efforts to predict these metrics have achieved mixed success. While the sound exposure level (SEL) can be predicted relatively well, it has proven harder to model the peak sound pressure level (Lpk) accurately. It would therefore be valuable to have a reliable way of estimating Lpk, based on a prediction of SEL. Correlations between SEL and Lpk and between SEL and rms sound pressure level (Lrms) are obtained from measurements during piling of the Luchterduinen wind farm off the Dutch west coast, and are used to assess the applicability of correlations found at other wind farms. A metastudy using data from Luchterduinen as well as the German Bight gives a more robust trend line for use at other North Sea sites. Lrms is computed in two different ways, involving the 90 % energy signal duration on the one hand and the effective signal time duration (Teff) on the other. The value of Lrms based on Teff is found to be more robust. [Work sponsored by BOEM.]