Passive Acoustic Recognition of Fishing Vessel Activity in the Shallow Waters of the Arabian Sea: A Statistical Approach

摘要

The ambient noise system consisting of a vertical linear hydrophone array (VLA) with 12 elements was deployed in the waters of the Arabian Sea at a depth of 16 m, off Goa, India, for extracting the ambient noise during the fishing season (March, April and May, 2013 before the onset of the south west monsoon). This study focuses on fishing vessel activity by finding the domination of the vessel and wind noise at two 12 hourly periodic cycles that start at midnight and noon, using the statistical analysis. It is performed using statistical parameters, like the mean, median, and standard deviation, skewness, percentile and spread of data. It is observed that the vessel noise dominates during the 12 h period that starts at midnight and is an indication of the activity of fishing vessels while the wind generated noise is more during the 12 h period that starts at noon, which is a sign of the domination of the sea breeze effects. This is the first time that a statistical analysis has been carried out to study the ambient noise data collected off Goa, in order to find the fishing vessel activity during the pre-monsoon season. The results are verified with the fishing information from the Directorate of Fisheries, Goa, ship traffic data from the Mormugao Port Trust and wind speed measurements.