PILOT : PELAGIC FISH STOCK ASSESSMENTS

The general approach for this pilot is to acquire data assets and utilize them together with other sources of insight. Some of the data assets can be supplied by remote sensing, including ocean surface currents and temperatures, and meteorological information. Other information must be measured locally, such as estimates of fish species and densities. These data assets will be provided by fishing vessels equipped with the necessary sensors and communications abilities. Other assets include catch reports, oceanographic simulations, and stock simulations.

Considerable effort in refining the data assets is needed to create value from them. The data will need to pass through several levels of processing and analyses to cultivate information relevant to fish stock assessments. It is expected that value creation is achieved by fusion of curated big data sets with pre-existing knowledge. The knowledge includes simulation models of both physical and biological processes. The hybrid analytics to be used is expected to entail several analytic types, including descriptive and predictive analyses, with activities within model creation, (re)training, and data assimilation.

This pilot will use hybrid analytics and the availability of big data is essential. Big data is needed for training fish stock models, and for assimilating/validating combined biological and oceanographic models. The data needs include satellite data (meteorological and oceanographic), model data (descriptive and predictive), local measurements (buoys and shipborne instruments), and reports on fish catches.

The fishing fleet will benefit from better stock estimates. This may lead to less variable fishing quota from year to year, and it may also lead to more sustainable fish stocks. Reliable stock estimates will provide the governing institutes a better basis for determining fishing quota, both in relation to stock migration and distribution of the stocks outside of conventional research cruises. In total, this pilot will give the means to accomplish better management of the fish resources.

The main motivation for this pilot is:

• Ensure a crowdsourced data collection initiative from fishing vessels.
• Perform collation of many data assets relevant for enabling big data analytics.
• Create a purposeful data pipeline that improves the data readiness level of assets.
• Explore big data analytics methods for fish stock assessments.

The main KPIs of the pilot are:

• Create a data pipeline that takes a variety of data assets and makes them ready for hybrid analytics in fish stock assessments – improving data readiness level.
• Investigate whether and how we can leverage big data technologies in making better fish stock assessments.