Yan, J. and Benson, T. (2025) Satellite enhanced computational modelling of underwater noise and ecological impacts in coastal areas. In: YCSEC 2025, 3-4 April 2025, Newcastle, UK.
Full text not available from this repository.Abstract
Underwater sound modelling has become an essential approach for understanding and mitigating noise pollution from human activities in the marine environment such as shipping, construction, dredging and mining. It is now an important aspect of Environmental Impact Assessments (EIAs), yet the lack of comprehensive baseline noise data significantly limits our ability to assess and understand the impacts of marine activities and proposed new developments on the underwater soundscape. Much of the long-term background (ambient) noise generated in the marine environment is generated by vessel activity in combination with natural sounds generated by weather (rain, wind and waves). We propose a novel computational model to incorporate Automated Information System (AIS) vessel data and weather data from the ERA5 global reanalysis dataset to provide accurate spatial predictions of the ambient noise sound spectrum. AIS data itself does not provide direct information on vessel noise emissions, posing challenges for accurate noise characterisation. To create a robust model for predicting underwater ambient sound maps for vessels in ports and harbours, a comprehensive evaluation of existing vessel noise source level spectrum models was first conducted, encompassing statistical approaches, and empirical models dependent of vessel characteristics (e.g. length, speed, type) and operational conditions. Notably, these source level models often yield different predictions for the same scenario, highlighting the need for a unified framework for underwater noise predictions. Due to its extensive field validation, the JOMOPANS-ECHO model was chosen as most suitable for modelling the sound source levels for moving vessels, which are dominated by propellor noise. A limitation of this model is that it does not account for engine noise. Hence, on its own, it is not suited to creating noise maps in coastal areas (particularly harbours) where many boats are moored or manoeuvring slowly. To overcome this limitation, we incorporated a second noise source model to account for the machinery noise, based on engine characteristics (e.g. engine mass and number). However, depending on the types of the AIS data available, and because of data gaps being present, these parameters are not always available. To fill these data gaps, an existing regression relationship was used to derive the most common engine characteristics using the available vessel length, speed and type information, allowing the model to be adapted to various types of AIS data. Using Falmouth Harbour (UK) as a test case, the modelled vessel source level spectra were then combined with local bathymetry data and fed into an underwater sound propagation calculation using an acoustic flux formula to generate maps of average vessel noise. The present study proposes a robust computational underwater noise model, aiming to provide reliable and quantitative prediction for anthropogenic underwater noise, to better support environmental impact assessments for both coastal areas and the open ocean. Further studies will build on the present work by incorporating more complex sound propagation methods to provide more accurate sound level predictions in shallow water environments. Additional regression analysis is also planned, which will improve the model parameter estimation and applicability of the model to the various types of AIS data.
| Item Type: | Conference or Workshop Item (Paper) |
|---|---|
| Additional Information: | https://conferences.ncl.ac.uk/ycsec2025/ |
| Subjects: | Maritime > General |
| Divisions: | Maritime |
| Depositing User: | Helen Stevenson |
| Date Deposited: | 14 Apr 2025 12:52 |
| Last Modified: | 14 Apr 2025 12:55 |
| URI: | http://eprints.hrwallingford.com/id/eprint/1681 |
Actions (for site administrators only - login required)
 |
View Item |
