A study on how environmental conditions affect shipping noise propagation in the north Arabian Sea

The anthropogenic underwater noise has a negative effect on the marine fauna in the form of the damage on communication and echolocation of marine mammals, body malformations of fish and invertebrates, higher mortality of eggs and zooplankton. Anthropogenic underwater noise is recognised as a form of pollution by the Convention on Migratory Species, the International Maritime Organization and the European Commission. The noise generated by ships is considered to be the main contributor to the underwater noise with the potential to impact marine ecosystems on a global scale, and this is why the shipping noise is the focus of this study. In the upper ocean the oceanographic parameters are subject to the seasonal changes, formation of the seasonal and daily thermoclines, fronts, filaments and eddies which may influence the underwater sound propagation.

One of the most popular sound propagation models is the Range-dependent Acoustic Model (RAM) [Collins, 1995] which is based on the solution of the parabolic equation. However, RAM requires a significant number of environmental data such as bathymetry, seabed characteristics and sound speed distribution in 3D, which is often difficult to obtain. Additionally, the model is relatively slow and computationally expensive in particular in the case of multiple sound sources (ships). To overcome this limitation, simplified acoustic propagation models have been developed. One of these is the energy flux model (EFM) which assumes the homogenous sound speed distribution in all 3 directions. The advantage to use the EFM is to have a fast and efficient model, which produces results in minutes on a typical desktop PC. However, the EFM may produce results which are not very accurate in the areas with significant variability of sound speed.

The purpose of this study is to compare the EFM against RAM in an area of significant spatial and temporal variability. We set up the EFM that computes shipping noise for a frequency of 50Hz in the northern Arabian Sea including the Gulf of Oman and the Persian/Arabian Gulf. The noise at source is calculated using the modified Ross formula [Erbe et al., 2012] and ships locations and velocity, are obtained from the Marine Traffic project. We also set up RAM for the same input parameters as the EFM plus the 3D sound speed distribution calculated form our operational ocean model for the Arabian Sea.

In order to assess the effect of 3D variability of sound speed in the ocean, we have compared the EFM against the reference RAM in winter and summer season, and calculated the statistics showing the differences between the models under the same input conditions. The analysis shows the seasonal and inter-annual dependence of the differences in the received level of noise at different depths, as well the role of the thermocline, eddies and fronts in modifying propagation of acoustic energy from surface ships.