Fig 1.
Example of speed profile for one vessel and fitting of a bi-modal distribution through the EM algorithm.
The confidence lines at 1.5 standard deviation around the first mode indicate the speed thresholds that were used to classify the AIS messages as fishing. The curve and speed profiles were analysed for each vessel resulting in specific classification threshold on the basis of individual fishing behaviour.
Fig 2.
Example of the validation of the method to detect fishing from non-fishing.
Blue points correspond to notifications of fishing activity from the logbook data. Points in red represent the position of AIS messages classified as non-fishing, and point in black are messages classified as fishing. The underlying density raster represents a kernel density estimate of the fishing grounds on the basis of the logbook data. Data refers to the fishing activity of a single vessel between January and August 2014.
Fig 3.
Cumulative level of uptake of the AIS in the EU fleet by three vessel length classes and month.
The percentage is calculated as number of vessel which activated the system against the total registered as active in the EU fleet register.
Fig 4.
Level of uptake of the AIS in the EU fleet by fishing port.
The size of the circles represents the number of vessels above 15 m of length registered in each port and the thickness of their contour line is proportional to the number of vessels for which no AIS message was recorded in the between January and August 2014.
Fig 5.
Results of the validation exercise.
The two distributions represent the values for messages classified as fishing and those classified as non-fishing in respect of fishing grounds defined for each vessel from the logbook data. Lower values indicate closeness to the historical areas of activity of each vessel.
Fig 6.
Map of fishing effort calculated from the AIS messages (density layer) and from the logbook data (ICES rectangles).
Fig 7.
Map of coverage of the AIS signal in the study area.