Modeling source levels of marine shipping

Noise from shipping traffic is often the dominant anthropogenic contributor to the underwater soundscape. Therefore, modeling underwater ambient sound requires accurate methods for estimating the source levels of marine vessels. This paper reviews recent developments in modeling source levels for ships and other marine vessels within the context of underwater ambient sound prediction. Source levels for ships are typically estimated using incomplete descriptions of vessel characteristics from data sources such as the Automatic Identification System (AIS). Large datasets on vessel source levels, such as the ECHO database, have enabled the development of improved models for marine shipping source levels. Source levels are often estimated statistically, based on information about ship class, operating conditions, and design characteristics. While source levels for individual vessels may have high uncertainties, source levels for populations of vessels follow predictable trends. This paper reviews ship noise models used in large-scale soundscape mapping projects, such as JOMOPANS and NAVISON. The different approaches are compared, and limitations and data gaps are discussed.