Underwater noise impact assessments are required in Environmental Impact Assessments to quantify the impacts of underwater noise on marine mammals. These impacts range from behavioural disturbance through to auditory injury and mortality. This process involves modelling species specific noise exposure thresholds with predicted noise levels at the site to create predicted areas in which different impacts are expected to occur. Unfortunately, there is no single standard way of doing this. There are many different threshold criteria that have been developed as well as many different models for underwater sound propagation. This new paper gives examples of how using different modelling procedures can affect the predictions created.
Why is this important?
Since different models produce different outputs, its very important to understand both what input data goes into the model and how the model works.
The authors of this paper state that:
“In practice, noise modelling for EIAs is often carried out using simplistic models, with limited environmental data, and without field measurements to ground-truth model predictions. In some cases, practitioners have developed proprietary models whose inner workings are not disclosed to regulators. This presents regulatory decision makers and their advisors with considerable uncertainty in the predictions of possible impacts”
Errors and uncertainties in the noise modelling process can lead to failings in the EIA process. For example, if the noise exposure is underestimated this can lead to an underestimation in the risk of injury and disturbance to marine mammals, potentially leading inadequate mitigation techniques and unforeseen impacts on marine mammals. Conversely, if the noise exposure is overestimated then this can lead to an overestimation in the risk of injury and disturbance to marine mammals and developments could be denied consent. Therefore it is vital that EIAs clearly state the input data used, the underlying assumptions and the scientific basis for their noise exposure predictions.