WP 3: Data product creation | Emodnet Biology

WP 3: Data product creation

Lead: U Sheffield

+  VLIZ, MBA, SMHI, ICES, IH Cantabria, University of Liège, CEFAS, Deltares, NIOZ

Objectives

Data products take the data supplied to WP2 (Access to Marine Biological Data) and turn them into the outputs which address the needs and questions of end users identified in WP4 (Uptake, Outreach and Communication). This work is at the interface of academic and government science and research software engineering.

In previous phases of EMODnet Biology, a range of products have been developed which adhere to FAIR principles, and which illustrate both the potential and limitations of EMODnet data. These products serve scientific user communities and are also clearly linked to the needs of other users, for instance their relevance to EOVs (Essential Ocean Variables) has been documented as part of the Product Stories. In addition, in the previous phases, libraries of code and instructions have been provided along with the products, facilitating their reuse. However, the technical bar to reusing and repurposing the products remains high.

The primary objective of WP3 in this next phase of EMODnet Biology will be to improve the software engineering of the product ecosystem to provide more complete and integrated tools to address the specific questions of the user community, for example in the form of self-contained and documented packages or browser-based apps. Alongside this, addressing scientific objectives will continue to be an important component of WP3, for instance developing methodologies to robustly visualise and assess changes in species abundance and extent over time and space.

We also recognise that answering user needs frequently requires integrating data from across EMODnet lots (e.g. EMODNet Chemistry, EMODnet Human Activities, EMODnet Physics and EMODnet Seabed Habitats), as well as from other relevant sources (human activities layers and datasets, helping to showcase localised effects).

In phase III the interoperability of different data types has been addressed by matching species occurrence data to environmental data from EMODnet and elsewhere, and by integrating data across EMODnet Biology and Seabed Habitats lots. A major focus of the new phase will be to increase the linkages between different data sources, providing products which fetch and process data required to address a set of specific user questions.

 

Methodology & activities

This work package will build on the methodology developed during the Phase III project. Noteworthy advances here included the application of Machine Learning methods to create gridded maps of species abundance in space and through time.

The zooplankton product has been adopted as the plankton Operational Oceanographic Products and Services (OOPS) by ICES as part of their Ecosystem Overviews which describe the trends in pressures and state of regional ecosystems. Considerable progress has also been made in creating robust maps of both species’ presence and absence, using EurOBIS data for a range of functional groups, including >1300 benthic taxa, produced and published following the Phase III workshop. Quantifying absence as well as presence is a crucial step in deriving indices of change which are robust to variable sampling effort.

Other products have acted as proof of concept for linking EMODnet biology data to data from elsewhere in EMODnet (e.g. Seabed Habitats) and from external sources (e.g. environmental data such as sea temperature or pH, and species traits such as fish living modes). The linkage of data from different thematic EMODnet portals reveals new insights and creates added-value products. Some of this work has used High Performance Computing facilities, e.g. by developing parallelised workflows to efficiently run products for very large numbers of species (for example, matching millions of occurrence records for thousands of species to gridded sea temperature products).

In Phase IV we will expand the taxonomic, geographic and temporal extent of presence-absence maps, refining Machine Learning interpolation and distribution modelling approaches, and extending and improving linkages to other EMODnet Lots and to other relevant external data sources (e.g. Copernicus environmental data) and products.

A key development will be to structure products around specific questions and evidence needs of the user community, collaborating closely with WP4, in particular using the outputs of the proposed WP4 questionnaire to capture specific user stories and requirements, as well as targeting priorities identified in the Phase III London Workshop from groups including the Regional Sea Conventions, the EEA (European Environment Agency), ICES (International Council for the Exploration of the Sea), MARS (Marine Research Institutes and Stations), MBON (Marine Biodiversity Observation Network) and the MSP (Marine Spatial Planning) community. Aligning products with relevant EBVs (Essential Biodiversity Variables) and EOVs (Essential Ocean Variables) will also be done in close cooperation with WP4.

Activities will include annual intensive workshops, using the productive model adopted in Phase III to progress product development on targeted themes (e.g. temporal trends, climate drivers, migration routes). Online collaboration will be facilitated with quarterly WP3 community focused calls including training and instruction (e.g. on effective use of GitHub and version control which will facilitate ongoing collaborative work on individual products) as well as discussions of product development.

 

Deliverables

D3.1 Quarterly WP3 community calls; call leader or other nominated team member to produce summary report of each call for publication on EMODnet website (M3-M24)

D3.2 Annual intensive workshops, in person with online participation options. Workshops involve collaborative product development on one or more targeted themes derived from WP4 user needs questionnaire (M12-?)

D3.3 Publish R package to link EMODnet Biology data with data from other EMODnet sources (M24)

D3.4 Develop method to use Phase III presence-absence maps to display time series of distribution change (M12)

D3.5 Produce position paper outlining questions that can be addressed using EMODnet data, together with remaining gaps, and strategies for filling these (M24)

D3.6 Add/update data product metadata in the EMODnet Biology catalogue (M24)

 

  •  
 

FOLLOW US

Website hosted and developed by VLIZ

Copyright 2017 - EMODnet - THE EUROPEAN MARINE OBSERVATION AND DATA NETWORK