Periodic Reporting for period 2 - EO4wildlife (Platform for wildlife monitoring integrating Copernicus and ARGOS data)
Reporting period: 2017-01-01 to 2018-12-31
EO4wildlife main objectives at the beginning of the action were:
- Architecture a system to access Sentinel Earth Observation data, ARGOS databases of archive and real time, thematic wildlife databank portals and other Earth Observation and MetOcean databases.
- Design and develop an operational and easy-to-use platform to query, search, mine and extract information from these different databases. End users should be able to cross correlate heterogeneous data in order to discover patterns and detect potential similar or reproducible behaviour or favourable conditions associated to the movement of the animals.
- Provide additional functionalities via a toolbox: connections to other external databases facilities to develop and run algorithms for dedicated data analytics, models and data visualization functionalities.
As conclusions of the action, overall assessment received was that the project has fully achieved its objectives and milestones, and delivered exceptional results with significant immediate or potential impact.
The different scenarios have been fully addressed, even if fish and marine mammals have not been fully exploited as turtles and birds, mainly by the rather and intrinsic difference in the data-set available. Both birds and turtles have very structured online archives which have facilitated the implementation of more tools for the data analysis.
The final platform looks very complete and well structured, and even if it does not seem fully user-friendly, detailed how-to documents have been prepared to facilitate approaching the platform and using its facilities.
Considering that the platform will be operational for a minimum of six months after the end of the project, it was strongly recommended that as many users as possible use it, in order to facilitate its future sustainability.
The wildlife scenarios work package was in charge of designing, implementing and validating different scenarios based on the use case requirements in the field of wildlife movements, habitats and behaviour. EO4wildlife project considered several use cases: management tools for authorities to help them make real-time decisions to protect selected seabird species; enhancing knowledge of scientists about the pelagic fish migrations routes, reproduction and feeding behaviours in order to better manage species; and setting up tools for marine protected areas managers.
During 2018, the team was heavily involved in finalizing the implementation of the four scenarios in the platform and their validation. The first complete scenario for sea turtles was implemented at the end of 2017 and required some adjustments and validation at the beginning of 2018.
A second release of the Implementation Scenarios was prepared and released before the annual review of the project with the European Commission in January 2018. This document reflected the implementation of the scenarios in the platform version V3.0 which was released at the end of the year 2017.
In order to open to external users, a list of additional users were selected from partners’ networks and listed the minimum requirements in terms of platform services and functionalities to validate in priority. These inputs were expected in order to evaluate the business opportunity and sustainability of the platform. Some services required more validation rounds than expected, and some other external constraints coming from CMEMS interfaces update induced that the opening to external users was finally delayed towards November for the Final event.
The EO4wildlife platform is composed of various functional components: 1) An internal data catalogue for aggregating geo-referenced products from external heterogeneous sources; 2) An ingestion module that allows the retrieval of data for exploitation by the platform services and; 3) A service Manager with which developers and/or data scientists manage the life cycle and execution of deployed services. Moreover, the platform has built-in visualization features for the resulting geographic data from the processing services. The service management mechanism in the Big Data infrastructure is built on the containerization concept (i.e. Docker) which allows encapsulating each service into an independent component that can be easily deployed on the cloud. An orchestration technology (i.e. Kubernetes) is used to manage container life cycle so that the underlying infrastructure becomes totally transparent.
In order to promote the projects’ results and give a wide visibility of these results towards the scientific and professional communities, the project maintains the website eo4wildlife.eu and accounts in social media (Twitter profile @EO4wildlife and the Facebook website https://es-la.facebook.com/eo4wildlife).
The EO4wildlife platform provides connectors to access data from various data sources, such as the CMEMS (Copernicus Marine Environment monitoring service) catalogue, indexing hundreds of ocean-related EO products, or the AVISO catalogue for the altimetry domain. An internal EO4wildlife data catalogue is maintained to aggregate products from all these external sources, and a specific task was carried out to develop the data connector to the CLS Datastore catalogue of product.
The Consortium has also worked on a generic EO4wildlife XML format to standardize not only the Argos location data coming from the users’ platforms (Seabirdtracking.org seaturtles.org) but also any animal track from any other scientific community. A special emphasis was done on the metadata description of the Argos and EO data and EO4wildlife services to follow the ISO/OGC/European standards.
As a conclusion, EO4wildlife contributes to the “efficient and widespread exploitation of the existing and planned European space infrastructure (especially Copernicus with its sentinel satellites)” by delivering a platform (1) enabling the access to Sentinel Earth Observation data, (2) ARGOS databases of archive and real time, (3) additional thematic databank portals (such Seabirdtracking.org) and (4) other Earth Observation and MetOcean databases; with (5) additional facilities to develop and run algorithms for dedicated data analytics and models.