Methods to get more information from sparse vessel monitoring systems data

Abstract

Read online

Vessel Monitoring Systems (VMS) and other vessel tracking data have been used for many years to map the distribution of fishing activities. Mapping areas with low levels of fishing activity can be of particular interest; for example to avoid conflicts between fishing and other ocean uses like offshore renewable energy or to protect relatively pristine ecosystems from increasing fishing pressure. A particular problem when trying to delineate areas that are lightly fished, is the relative sparsity of vessel monitoring data in these areas. This paper explores three novel methods for estimating the distribution of fishing activity from VMS data, with particular focus on lightly impacted areas. The first new method divides the area of interest into a nested grid with varying cell sizes (depending on the density of data at each location); the second new method uses Voronoi diagrams to define polygons around observations and the third method applies a local regression to generate a smooth map of fishing intensity. The new methods are compared with two established methods: applying spatial grids and interpolating fishing tracks. The track interpolation method generally performs better than any of the new methods, however it is not always possible or appropriate to apply track interpolation; in those cases the local regression method is the best alternative.

Keywords

• vessel monitoring systems
• automatic identification systems
• fisheries
• spatial distribution
• mapping
• local regression