Mapping Underwater Sound Noise and Assessing its Characteristic Based on AIS

Underwater noise from shipping is a growing presence throughout the world's oceans. Exposure to loud noise can have various effects on marine animals. This paper presents a method which can characterize sound exposure from shipping based on Automatic Identification System (AIS) shipping data. Firstly, the tonnage, speed, draft, latitude and longitude of each ship can be obtained from real-time AIS data. Then we calculated the source level of each ship by the D.ROSS formula. Secondly, a large scale area was derived over a regular grid having a horizontal resolution of 0.5nmx0.5nm in order to facilitate the engineering application. The sound band pressure level is calculated using the source level of a ship after making a numerical calculation of 1/3oct transmission loss at each grid point for every ship by using the real-time sound velocity profile and bottom parameters based on a fast field program. The distribution map of ambient noise was plotted from the result of incoherent superposition of the ship's noise energy. Finally ambient noise at different times, <1kHz and wideband sound, was introduced to an unsupervised neural network: Self-Organizing Map(SOM), which plots high-dimensional input data onto a low dimensional array whilst preserving the topological relationship between the data. We obtained sound exposure characteristics of ambient noise in this area. The method is demonstrated to be an efficient tool in predicting sea ambient noise levels based on vessel distribution and also proved to be effective in monitoring the implication of ambient noise for marine organisms.