Data e-Infrastructure Initiative for Fisheries Management and Conservation of Marine Living Resources iMarine

Infrastructure Development and Operation

In order to support and realise the Ecosystem Approach envisaged by the CoP, iMarine will deploy and operate its own e-Infrastructure. This e-Infrastructure will not work in isolation but it will benefit from and rely on resources and services operated by existing infrastructures and service providers of various types.

Many of these existing infrastructures are available under the D4Science Federation. This federation of infrastructures is providing access to a large set of multi-type heterogeneous resources coming from different research and scientific domains (humanities, high energy physics biodiversity, environmental science, and others). Relevant examples of such infrastructures are: GENESI-DR, AquaMaps, EGI, etc.

Figure – iMarine Data Infrastructure and Interoperable Infrastructures
The infrastructures interoperable with iMarine Data Infrastructure will provide resources of different types which can be clustered in: computing and storage resources (red in Figure ), community-specific resources (blue in Figure ), and other relevant resources not directly produced within the EA domain (green in Figure ). Also represented in Figure (solid dark blue circles-lines) are the infrastructures currently belonging to the D4Science Federation.

The iMarine Data Infrastructure will be a completely new infrastructure, deployed for the first time by the project Service Activity, to address the Ecosystem Approach vision and needs. It will collect and provide access to the resources identify by the EA-CoP. From the hardware point of view, the infrastructure will offer a minimum of 250 servers (17 TB storage space linked by 1 GB network connection) available to host and operate the communities’ resources which will be made available through a number of well identified Virtual research Environments. These nodes will run the gCube software and other community specific applications and tools to deliver all capacities required to store, transform, curate, process, and access the communities’ data resources.

The Federated Model adopted to manage the interconnection with many of the cited infrastructures and systems will rely on: (i) Utility Services: services and applications common to all entities participating to the federation, and managed by the iMarine Data e-Infrastructure for all agencies and infrastructures to improve service and/or reduce costs; (ii) Shared Services: services and applications required by more than one participating entity, and managed by the iMarine Data e-Infrastructure to improve service and efficiency; and (iii) Agency-Specific Services: applications and services of a highly specialized nature for which there are no opportunities to add value through central management but that can use the other services of the federation at no-cost for the agency.

As explained, besides the resources hosted in iMarine Data e-Infrastructure itself, iMarine will establish close interoperability links with other infrastructures, many of them already available in the D4Science Federation. This collaboration will be based on a Research Collaboration Model where resources of one infrastructure can be consumed or accessed by another infrastructure and vice-versa through an agreement community-based approach under the control of the infrastructure’s middleware.

Infrastructures providing Computing and Storage Resources

EGI (European Grid Infrastructure) is a project willing to create and maintain a pan-European Grid Infrastructure in collaboration with National Grid Initiatives (NGIs) in order to guarantee the long-term availability of a generic e-infrastructure for all European research communities and their international collaborators. The European Grid Infrastructure will (1) Operate a secure integrated production grid infrastructure that seamlessly federates resources from providers around Europe, (2) Coordinate the support of the research communities using the European infrastructure, and (3) Work with software providers within Europe and worldwide to provide high-quality innovative software solutions that deliver the capability required by user communities.

EGI provides storage and computing resources, distributed across hundreds of sites worldwide and is based on different interoperable grid middleware solutions such as Globus, gLite, Arc, and Unicore. The resources offered by the EGI infrastructure will significantly extend the storage and computing capacity available under the D4Science Federation. These resources will be accessible from the iMarine Data Infrastructure under a Research Collaboration Model.

VENUS-C (Virtual multidisciplinary EnviroNments USing Cloud infrastructure) is focused on developing and deploying a Cloud Computing service for research and industry communities in Europe with the aim of: (1) creating a platform that enables user applications to leverage cloud computing principles and benefits, (2) leveraging the state of the art to bring on board early adopters quickly, incrementally enable interoperability with existing Distributed Computing Infrastructures (DCIs), and push the state of the art where needed to satisfy on-boarding and interoperability, and (3) creating a sustainable infrastructure that enables cloud computing paradigms for the user communities inside the project and new communities.

VENUS-C offers an industrial-quality service-oriented platform based on virtualisation technologies by providing an open and generic Application Programming Interface (API) at platform level for scientific applications. The VENUS-C platform is based on both commercial and open source solutions underpinned by the Engineering data centre, Microsoft through the Windows Azure and its European data centres, and two European High Performance Computing centres: The Royal Institute of Technology (KTH, Sweden) and the Barcelona Supercomputing Center (BSC, Spain). The resources offered by VENUS-C will significantly extend the storage, computing, and service hosting capacity available under the D4Science Federation. The VENUS-C abstraction allows the immediate access to a concrete set of resources providers (Engineering Data Centre, Microsoft Windows Azure, KTH, BSC). These resources could be made accessible from the iMarine Data Infrastructure initially under a Research Collaboration Model and afterwards under a commercial Cloud Model.

Infrastructures providing Community Resources

OBIS allows users to search marine species datasets from all of the world's oceans. It allows to identify large-scale ecological patterns, analyze dispersions of species over time and space, and plot species' locations with temperature, salinity, and depth. The OBIS schema is a list of data fields with names, descriptions, and format notes. It is an extension to the Darwin Core Version 2 standard. Darwin Core is a Biodiversity informatics data standard that consists of a vocabulary of terms to facilitate the discovery, retrieval, and integration of information about organisms, their spatiotemporal occurrence, and the supporting evidence housed in biological collections. Often abbreviated as DwC, it was originally created as a Z39.50 profile by the ZBIG Z39.50 Biology Implementers Group. The Darwin Core profile is expressed as an XML Schema document for use by the DiGIR (Distributed Generic Information Retrieval) protocol. A TDWG task group was created to revise the Darwin Core and bring it to the state of a ratified metadata standard. Ratification of the standard occurred on 9 October 2009.

Darwin Core is based on the principles of Dublin Core and decouples the definition and semantics of individual terms from application of these terms in different technologies such as XML, RDF or simple CSV text files. It not only allows on sharing data through a dynamic DiGIR provider, but also offers RESTful, OAI and OGC services. It will be used to enrich the data sets accessible through the iMarine Data e-Infrastructure with biodiversity metadata and content from multiple research areas.

FAO-IOC/IODE The United Nations agencies of FAO and UNESCO-IOC/IODE collaborate to provide information on standards and controlled vocabularies in oceanography, marine science, food, agriculture, development, fisheries, forestry, natural resources and related sciences. They produced a customized version of DSpace called OceanDocs for the International Oceanographic Data and Information Exchange (IODE) of the Intergovernmental Oceanographic Commission of UNESCO (IOC) and adapted it to the standards of the Oceanographic community. The OceanDocs Network, created in 2004, now has some 50 members. The FAO customized DSpace using the AGRIS Application Profile (AP) and is developing a plug-in for the use of controlled vocabularies for communities in food, agriculture, development, fisheries, forestry, natural resources and related sciences such as AGROVOC.

The communities of FAO and UNESCO-IOC/IODE active in oceanography and food, agriculture, development, fisheries, forestry, natural resources and related sciences will provide open access to their literature stored in those repositories. They will use the same standards for metadata, thesauri and other ontologies ensuring advanced access to the scientific publications in the field and the possibility to create new services for their researchers. The resources will be accessible from the iMarine Data e-Infrastructure under a Federated Model

IRD focuses on data discovery by implementing different metadata standards like Dublin Core (RDF schema) and ISO 19115/39 for spatial data as well as several TDWG standards that implement the GBIF profile of EML (Ecological Metadata Language). The main datasets produced by IRD (Observatoire Thonier) are already described with RDF Dublin Core metadata and tagged with URI. In 2011 the IRD these will also be described with ISO 19115/39 into a catalog (CSW server) and through TDWG standards by using IPT software. Also IRD reuses or plans to develop ad hoc ontologies (OWL) and to use entities URIs in the ontologies to tag metadata. The GBIF IPT is an open source, Java (TM) based web application that connects and serves three types of biodiversity data: taxon primary occurrence data, taxon checklists and general resource metadata. The data registered in a GBIF IPT instance is connected to the GBIF distributed network and made available for public consultation and use. These resources will be accessible from the iMarine Data Infrastructure under a Federated Model.

FAO Corporate Document Repository maintains FAO documents and publications, as well as selected non-FAO publications. It offers a standardised interface to access this repository of objects produced by FAO directly on the Internet. The Aquaculture, Fisheries Management and Conservation, Fisheries Policy and Planning, Biodiversity, and Climate Change collections are all relevant to the Ecosystem Approach envisaged by the CoP. Those collections will become accessible from the iMarine Data e-Infrastructure under a Federated Model.

FAO Fisheries Global Information System (FIGIS) was established to offer reliable, high-quality and relevant information on the state of world fisheries and support the Code of Conduct For Responsible Fisheries that represents the basis for policies aimed at sustainable fisheries. FIGIS is an information management tool that interconnects groups of institutional partnerships to build up a network of subsystems. FIGIS, as part of the FAO Fisheries and Aquaculture Department's regular activities acts as a framework with reference to FAO information management policy. FIGIS delivers expert knowledge, a set of software tools, collaborative mechanisms, and interoperability solutions to a broad range of needs in fisheries information. With the adoption by the Committee on Fisheries of the Strategy for Improving Information on Status and Trends of Capture Fisheries (STF) on 28 February 2003, FIGIS becomes one of the privileged tools for its implementation. The services offered by FIGIS will be integrated and made accessible from the iMarine Data Infrastructure under a Federated Model to offer a web-integrated fisheries information system.

FAO GeoNetwork provides Internet access to interactive maps, satellite imagery and related spatial databases. Its purpose is to improve access to and integrated use of spatial data and information.
GeoNetwork opensource is a standardized and decentralized spatial information management environment, designed to enable access to geo-referenced databases, cartographic products and related metadata from a variety of sources, enhancing the spatial information exchange and sharing between organizations and their audience, using the capacities of the internet. These resources will be accessible from the iMarine data infrastructure under a Federated Model to provide an easy and timely access to available spatial data and to existing thematic maps that might support informed decision making.

FAO AgroTagger that is being developed by IIT Kanpur and ICRISAT in collaboration with FAO is an automatic web-accessible tagging module for agriculture-related contents whose salient features are: language detection; ability to process various document types (doc, docx, pdf, odp, html); ability to assign tags, named agrotags, with or without a thesaurus or a training data set; ease of integration into the Drupal Content Management System. Agrotags is a subset of keywords selected from Agrovoc, by IIT Kanpur and ICRISAT, and acts as the “core” keywords that describe the domain, an abridged version of Agrovoc. This service will be accessible from the iMarine Data Infrastructure under a Federated Model and it will be used to enrich textual and structured data with agrotags keywords.

PESI promotes correct use of names and their relationships as an essential step for biodiversity management while the availability of taxonomically validated and standardised nomenclatures is fundamental for biological e-infrastructures. PESI integrates and secures taxonomically authoritative species name registers. PESI is a joint initiative of two Networks of Excellence: EDIT (European Distributed Institute of Taxonomy, and MarBEF (Marine Biodiversity and Ecosystem Functioning), funded by the European Commission under the Seventh Framework Capacities Work Programme – Research Infrastructures – and is led by the University of Amsterdam.

The PESI data warehouse uses the Common Data Model (CDM) that follows the TDWG Ontology, but modelling was influenced by other models and standards, such as the Access to Biological Collections Data (ABCD) schema, the Taxonomic Concept Schema (TCS) and the Structure of Descriptive Data (SDD) schema. PESI is also building a common intelligent name-matching device in consultation with principal initiatives like GBIF and LifeWatch. This provides a unified cross reference system to all stakeholders optimising their taxonomic meta-data service functioning. The services offered by the PESI infrastructure will be accessible from the iMarine Data Infrastructure under a Federated Model to significantly improve the taxonomic disambiguation capacity and improve the quality of the data sets.

eDAMIS (electronic Data files Management and Information System) s a modern communications management system allowing easier and more accurate transfer of data between various national, EU institutions and Eurostat, the central statistical office of the EU. Democratic societies do not function properly without a solid basis of reliable and objective statistics. Most European Union (EU) statistics are aggregated from hundreds of competent national authorities among whom the main ones are national statistical offices. The purpose of eDAMIS is to provide a simple generic solution for a transparent and reliable exchange of data files using an advanced control system with acknowledgements, notifications, content validation, monitoring and dispatching of data files towards the relevant recipients (application or users).  eDAMIS could be considered as an advanced post service which uses the latest internet technologies to guarantee the easy, reliable and smooth transmission of all statistical data files that have to be sent to Eurostat. It also offers tracking and monitoring reports that can be used as input to check the respect of the legal obligations to transmit data. This service will be accessible from the iMarine Data Infrastructure under a Federated Model to guarantee a proper interconnection to Eurostat to all participating institutions.

FISHSTAT Plus is a software for fishery statistical time series providing users with access to fishery statistics of various sorts. Any data having time series structure can potentially be stored and processed by Fishstat Plus. It provides access to approximately 600 species items included in the FAO International Standard Statistical Classification of Fishery Commodities (ISSCFC) (commodity categories cover products derived from fish, crustaceans, molluscs and other aquatic animals and residues caught for commercial, industrial or subsistence uses, by all types of classes of fishing units operating in inland, fresh and brackish waters, in inshore, offshore or high seas fishing areas. This service will be accessible from the iMarine Data Infrastructure under a Federated Model to guarantee a proper exploitation of the information available through FishStat Plus.

AquaMaps is a data service oriented to the marine environment that makes available species range maps plus a variety of thematic maps (taxonomic, climatologic, invasiveness, etc.). Each map is a complex information object.

All the AquaMaps collections are open source. Currently, several different communities of scientists access these collections, either through the AquaMaps user interface or through the interface of other systems (e.g. FishBase, the most widely used biological information system with over one million visitors per month; other species information systems such as SeaLifeBase and GBIF).

Infrastructures providing Other Resources

GENESI-DR (Ground European Network for Earth Science Interoperations - Digital Repositories), was an ESA-led, European Commission (EC)-funded two years project, that together with several data provider institutions (e.g. DLR, CNES, ASI) demonstrated how a distributed infrastructure can provide a reliable, easy, long-term access to Earth Science data and associated on-demand processing resources via the Internet.

A total of 15 different Digital Repositories hosting more than 160 dataset series are integrated already in GENESI-DR. Dataset series refer to heterogeneous and include satellite data, in situ data, images acquired by airborne sensors, digital elevation models and model outputs. As an example, the SeaDataNet scenario includes discovery via Web portal of heterogeneous data (satellite, in situ, maps), download of data, discovery of and access to processing services (on-demand generation of Sea Surface temperature maps from satellite data).

With the follow-on project EC FP7 GENESI-DEC started in May 2010, the work developed during GENESI-DR will be pursued targeting the Digital Earth Communities with a strong accent on distributed infrastructures security, semantics, ontology and advanced workflow management. GENESI-DEC aims to address Digital Earth Communities; a first set of these have already been selected, while additional ones will be identified and addressed during the project life. The communities that are now the focus of integration are: The Seafloor and Ocean Observation Community; The Global Atmosphere Observation Community using Aircraft; The Global Change Earth Observation Community; The Territorial development and spatial planning Community; The Black Sea catchment observation community.

OpenAIRE (Open Access Infrastructure for Research in Europe) is responsible of supporting both the FP7 Open Access pilot and the ERC guidelines for Open Access. OpenAIRE provides guidance and assistance to researchers in publishing their research results in Open Access. The project sets up a helpdesk system in 27 European countries, consisting of a network of experts and a portal with online tools. It also helps to raise awareness of Open Access and the EC pilot in research institutions; it provides valuable information and statistics on the use of Open Access publications; and explores scientific data management services.

OpenAIRE provides access to a distributed infrastructure of Open Access repositories focused in research publications (peer-reviewed articles, pre-prints, conference publications, others) and associated tools and services to manipulate these research datasets. The resources offered by OpenAIRE will extend and enrich the data repositories available under the D4Science Federation by bringing publications content from multiple research areas. These resources will be accessible from the iMarine data infrastructure under a Research Collaboration Model.

Technology Development

The extent to which a data e-Infrastructure succeeds as an application platform is largely dependent on the tools and services that it makes available to develop and execute the broad class of applications for which it is intended. For the iMarine infrastructure, the target applications are those required to implement the EA and thus to execute to collate, access, integrate, annotate, transform, search, curate, and publish a variety of data types, including statistics, biological records, environmental observations, and semantic structures. This calls for the systematic integration of a large number of technologies across the services and tools offered by the infrastructure. Section 1.1 and Section 1.2 have, respectively, introduced and motivated the technological targets of the integration task. In doing so, they have effectively characterised iMarine’s as a gCube-based Data e-Infrastructure with standards-based interoperability bridges to a variety of information resources that are already available to the emerging EA-CoP, albeit in an uncoordinated manner. This Section overviews the main themes of the development and integration task, including:

• 
  the changes and extensions that will be applied to key gCube services and service subsystems;
  
• 
  the use of software components already in the target domain for the implementation of those changes.
  

In essence, development activities will concentrate in four main functional areas⁴:

• 
  Consolidation and enhancement of the Enabling Services and Tools of the data e-Infrastructure;
  
• 
  Consolidation and enhancement of the Data Management Services of the data e-Infrastructure;
  
• 
  Consolidation and enhancement of the Data Consumption Services of the data e-Infrastructure;
  
• 
  Development of a set of Application Programming Interfaces (API) to key application-level services of the data e-Infrastructure;
  

The Enabling Services will attend to the operation of the Data e-Infrastructure, encompassing functions that relate to resource management (dynamic deployment, monitoring, brokering and match-making, resource publication, discovery, and change notification), security and policy enforcement (authentication, authorization, accounting), and process and workflow execution. Orthogonal but conceptually related to the Enabling Services are the tools that will support the development and execution of infrastructural services, i.e. application frameworks and service containers. As enabling technologies are core gCube offerings, development activities in this area will concentrate on extending and enhancing the suite of services and tools already available within that system. The design and development themes here are varied, ranging from higher-level models of usage (e.g. for workflow specification, composition, and reuse) and policy-driven management schemes (e.g. for resource publication and discovery and for security enforcement), to systematic alignment with standards (across the full functional spectrum). The key theme, however, is rooted in the expansion and re-organisation of the current model of shareable resource, with the goal to extend resource management schemes, security enforcement regimes, execution capabilities, and development support to a broader range of data and computational resources than those currently manageable in gCube-based infrastructures. The main outcome will be a flexible integration model for services and other components that were not specifically developed for the gCube system.

The Data Management Services will encompass functionalities to do with storing, accessing, describing, annotating, moving, and curating data sources that will be published, directly or indirectly, in the data e-Infrastructure. Again, gCube offer services that cover large portions of the functional spectrum. It also defines extensibility mechanisms to adapt general-purpose functions to domain-specific requirements, both at architectural level – within subsystems designed as service frameworks (such as the Content Management subsystem) – as well as within individual services (dynamic plug-in components). Exploiting such mechanisms, however, is matter of intensive development. While an important theme in this area is to do with general-purpose functionality relative to data curation (such as harmonisation, quality assessment, supervision, certification, and verification), the unifying development theme is undoubtedly related to interoperability, a requirement that descends from the rich variety of domain-specific data types that are expected to flow through the data e-Infrastructure (statistics, environmental data of various types, data related to biodiversity, and data ready for ontological reasoning and other forms of “semantic” processing). This is where development will capitalize on the flexibility of integration models endorsed by the Enabling Services so as to reuse software components produced in autonomy from gCube, from simple format conversion algorithms and standard implementation frameworks (e.g. OpenSDMX), to client libraries for external services. The activities will result in new services (e.g. outer types for content dissemination, multi-protocol data transfer mechanisms, distributed storage interfaces) and new service plug-ins (e.g. content access plug-ins for a number of document-oriented and data-oriented models). At a more abstract level, development activities in this area will most visibly root the infrastructure in the domain for which it was intended.

The Data Consumption Services will be strictly concerned with the manipulation, analysis, and rendering of the datasets managed in the data e-Infrastructure. Functionalities in scope are those to transform data representations, index and retrieve parts of the data, detect structural and content-based patterns within the data, visualise the data in ways that are germane to its type, and process the data with statistical methods for simulation and forecasting purposes. Some of these functionalities, such as those related to data transformation, indexing, and retrieval, are already supported in gCube. Yet the wide range of data types available in the infrastructure – particularly ontologies and other forms of semantic data, as well as data with spatio-temporal features – requires a re-thinking of the query language and calls for novel filtering, query pre-processing, indexing, and data transformation algorithms. Equally, the target domain demands interoperability with new query distribution protocols and result formats, while the possibility of uncooperative search engines requires new approaches to data indexing (e.g. query-based sampling). Some of these changes will rely on extensibility mechanisms built-in the gCube framework for search, but support for data mining, visualisation, statistical analysis, and semantic analysis will require development and integration of services and algorithms that are lacking in gCube. Like for Data Management Services, development activities in this area will rely on the integration models supported by the Enabling Services. Conversely, the resulting services will yield powerful “primitives” for workflow and process execution services, thereby maximising their usefulness within the infrastructure. More than elsewhere, the development effort here will align the infrastructure to the processing requirements associated with the implementation of the EA.

The last area of development will focus on providing an integrated set of APIs to the services available in the infrastructure. The expected outcome is a toolkit of client libraries that offer programmatic abstractions to simplify and optimise interactions with remote services, including high-level bindings of service inputs and outputs; linguistic facilities and design patterns (e.g. type-safe structures and familiar producer/consumer patterns); sophisticated interaction strategies (e.g. best-effort strategies); remote failure handling; uniform approach to cross-cutting requirement (e.g. security and scope management); and a range of pre-processing and post-processing tasks aimed at reducing latencies and service load across multiple service invocations. Collectively, these transparencies address a fundamental concern in infrastructure building, namely the excessive complexity of using its services. Without adequate interfaces to its functionalities, the infrastructure fails altogether as an application platform or else puts excessive burden on its development team. Accordingly, development activities in this area are key to the adoption as well as the sustainability of the infrastructure as a whole.

In conclusion, it is worth noting that, orthogonally to the various activities, the methodology of choice is based well-proven Agile Development principles [4], Software components will be developed and released in short and continuous iterations, each release addressing the need of new functionality or the need to revise existing functionality.

Realising Community Business Cases

The EA-CoP aims to promote the Ecosystem Approach to Management of Marine Living Resources, a holistic approach to management and conservation that recognizes more explicitly the interdependence between human well-being and ecosystem health, and the need to maintain ecosystem productivity for present and future generations.

The iMarine project aims at serving actual needs of this community in implementing this approach. This section presents three challenging real life EA business cases that will be specifically addressed in the course of the project. These will serve to demonstrate to the EA-CoP the effectiveness of the data infrastructure in support to real situation.

It must be clarified that the project does not pretends to respond to the entire spectrum of needs of the underlying initiatives, but instead that it will provide key (and complementary) solutions to specific unresolved problems through the development of a number of Virtual Research Environments specifically conceived to serve these scenarios. However by responding to different facets of EA-CoP’s needs expressed under each business cases, the project will end-up covering a broad range of EA-CoP’s generic requirements as discussed in Section . The connection to live initiatives also intends to exploit external co-funding synergies in order to expand the scope of data and services delivered through the iMarine Data e-Infrastructure. From the EA-CoP’s view point, the project will provide resources to the “glue” necessary for the connection of the various initiatives, and by doing so it will really contribute to the EA-CoP’s realisation . It should finally be noted that the distributed nature of the proposed data infrastructure favours this approach. The following description presents for each business case: (i) some background illustrating the context where the business case takes place; (ii) who are the actors involved; (iii) the facilities that the iMarine is expected to offer to support the business case; (iv) the planned approach for providing these facilities, and finally (v) the synergies which the project will build on.

Support to the European Common Fishery Policy

Background: As a member of FAO, the EU has adopted the FAO Code of Conduct for Responsible Fisheries (CCRF) and one of its key instruments, the Ecosystem Approach to Fisheries, has become one of the drivers of the European Common Fishery Policy (CFP). The implementation of this EU CFP requires a complex flow of fishery data and information from fishing industry activity, through member states, to supporting EC agencies. As overall picture, the statistics which result of fishery data workflow are exploited together with other thematic information flow, in order to formulate scientific advice on the basis of the assessed biological state of resources and socio-economic performance of the fishing sector. In turn, DG-MARE uses this scientific advice to elaborate its management and policy advice. Recently, the overarching need to have more integrated and higher quality data including in coverage, timeliness, resolution and accuracy has been stressed by DG-MARE and a substantial revision of the entire information workflow and supporting tools has been engaged since 2009. Concerned work streams and priority levels are (i) Activities database: forwarding of daily fishing vessels activity data, including VMS, log books, landings; (ii) Fleet registry – Fishing licenses database: work flows from member states to EU; (iii) Legal data; (iv) Scientific data for fixing quotas.

The EA-CoP Group: The CoP is broad and includes many actors, but the EU has organized them in a range of EC agencies institutions which strongly interact: DG-MARE for control and surveillance data, and management recommendations; Eurostat for fishery statistics; JRC-ISPRA for Vessel Monitoring System (VMS) data; scientific working groups under ICES (for biological advice) and JRC-ISPRA (for socio-economic advice); Regional Advisory Council (RACs) for the contribution of fishermen communities in the development of knowledge and the formulation of management decisions. FAO interacts in particular with Eurostat and ICES in scientific and statistical data exchange as part of the Coordinating Working Party on fishery statistical standards (CWP)⁵, and the Fishery Resources Monitoring System partnership (FIRMS).

The Goal: The iMarine data infrastructure will acquire an advanced capacity to exchange, curate, harmonize, aggregate and manage structured socio-economic data including fishery time series, fishing activity observation records, and VMS data. For the latter, this capacity will address demanding confidentiality requirements. By integrating this capacity as part of its complementary analytical, mapping and reporting services, it will offer a unique set of features able to contribute to EC decision making workflow, as well as that of other national or regional institutions.

The Approach: Advanced implementation of the SDMX protocol has already been achieved by Eurostat, and FAO is building on these efforts to develop OpenSDMX. Leveraging on these efforts, iMarine will integrate this OpenSDMX capacity and complement it with semantic services for harmonization of various sources of statistics. The capacity thus developed will serve as foundation for application to other data exchange protocols supporting structured data, such as the COST model for fishery observation records. Finally, consideration will be given as to how best these different protocols can be combined to exchange and process VMS data with utmost care on confidentiality requirements.

The Synergies: The project will build on initiatives developed under regular programmes of a number of institutions associated in the project. These on-going initiatives (thereafter summarised) will be assembled and operationalized in the context of iMarine technology deployment.

The Eurostat SDMX initiative has been brought in March 2010 to the attention of the CWP. Under this cooperation channel, FAO and Eurostat have decided to strengthen their collaboration by sharing expertise, software and services. In 2011, FAO will publish its authoritative code lists according to the SDMX format as per Eurostat request. The Eurostat e-Damis SDMX validation engine will be assessed in 2011 regarding its possible reuse and adaptation as part of FAO’s Open SDMX initiative which will grant FAO with advanced SDMX facilities. This collaboration is supported (including financially) by FAO’s on-going projects aiming at modernizing its FAOSTAT system.

Regarding exchange and processing of fishery observation records and VMS data, various initiatives are worth mentioning on the EU and international side. These include: ICES publishing of the COST format for handling data exchange and quality improvement of sampling, landings, and effort data from Commercial fisheries, including for Observers on board; commitment by IRD to test COST on its Tuna Observatoire and ObsTuna data; an XML schema and data dictionary for control data developed by DG-MARE; recommendations by the ICES Study Group on VMS data (SGVMS, 2010) on approaches for the more systematic use of VMS data, encompassing the delivery of anonymized aggregated data with VMS linked to logbook information; NEAFC and ICES collaboration regarding VMS data processing algorithms.

Finally, it is anticipated that the “EAF dashboard” features which will be developed as part of the third business case identified (cf. Section ) will be of interest in the context of DG-MARE’s general objective of an integrated system including scientific data and knowledge from fishing industry.

Support to FAO’s deep seas fisheries programme: balancing use of marine resources and protection of vulnerable marine ecosystems in the high seas

Background: in 2009, FAO established international guidelines for the management of deep seas high seas fisheries with the aim to protect deep sea fragile ecosystems. In support to the implementation of these guidelines, FAO has formulated a programme of action entitled balancing use of marine resources and protection of vulnerable marine ecosystems (VMEs) in the high seas (DSF programme). One key component of this programme is the development of a VME Mapping Information System which will facilitate the registration, description, analysis and publications of all forms of information related to VMEs, in support to the management of High Seas Deep Seas Fisheries (DSF).

In parallel, the Global Ocean Biodiversity Initiative (GOBI), an international partnership advancing the scientific basis for conserving biological diversity in the deep seas and open oceans, was launched in late 2008 in support to CBD’s goals of identifying Ecologically and Biologically Sensitive Areas (EBSAs) in the Deep Seas. GOBI’s work is supported by various sources of scientific data including OBIS and various other scattered information bases.

The EA-CoP Group: in the context of High Seas Deep Seas, such call for action can only be successful if the various groups work as one EA-CoP. The group is relatively well circumvented: an industry limited to about 285 vessels worldwide and concerning 27 flag states, 6 established Regional Fisheries Management Organisations (RFMOs) concerned in the management of deep sea fisheries and in the protection of VMEs, and a few international organizations such as FAO promoting sustainable fisheries together with the protection of VMEs, or UNEP, IUCN, and CBD promoting the protection of Ecologically and Biologically Sensitive Areas (EBSAs). As follow-up to the Busan workshop (May 2010), a DSF Discussion Community⁶ has been launched, involving fishery managers, fishery biologists, lawyers, biodiversity scientists, data managers, fishermen representatives.

The Goal: To realise a first version of a VME mapping information system. This will support the registration, description and publication of global information on VMEs, including through documents, maps, and tabular data. It will also provide access to Emerging scientific information regarding biological and ecological criteria relevant to the definition of VMEs.

The Approach: the VME mapping information system will be realised by federating and leveraging within iMarine a number of applications and services developed respectively within the FIGIS, Aquamaps, OBIS, and CRIA infrastructures. FIGIS will develop the VME database including an inventory of legally defined VMEs, and this application will be part of iMarine building on already achieved interoperability between FIGIS and gCube, in particular regarding Fact sheet and Geospatial services. By generalizing the species distribution modelling capacities developed in the D4Science-II Aquamaps scenario, developing more interactive mapping and analytical services (“R”), and interoperability with OBIS and CRIA, iMarine will strongly contribute to the identification of key biodiversity criteria required for the identification of VMEs. Other sources of environmental information will be added as required (e.g. predictive map of coldwater coral distribution in the South Pacific) through interoperability services with other systems (eg UNEP-WCMC’s IMAP, EMODNET, or International Seabed Authority), or by using the iMarine Virtual Research Environment for loading relevant data sources and maps (e.g. Seamounts on line).

The Synergies: The VME database development will start in January 2011 funded by FAO project GCP/GLO/309/FRA, a funding component of the FAO DSF programme. Other funding synergies with Global Environmental Fund (GEF) in support to FAO DSF programme and CBD-IUCN programmes on deep seas are expected to support the development of interoperability with other relevant systems as required.

In the north east Atlantic area, the VME mapping information system will also benefit from availability of processed VMS data made available by NEAFC (cf. Section ) in order to assess fishing activity in relation with established VMEs and areas suggesting patterns of VMEs. In this regard, the iMarine project will further strengthen the ICIS module developed during the D4Science-II project.

As part of the FAO-European Space Agency (ESA) collaboration, the opportunity of an additional module for tracking through satellite imagery the location of fishing vessels in the high sea will be considered. Once the imagery products will be developed through ESA funding sources, such products will be made available to the iMarine data infrastructure by Genesi-DEC which already has developed such capacity in D4Science-II.

Support to regional (Africa) LME pelagic EAF community

Background: EAF sets as principle that fishery management objectives should strive to reach an optimum balance between the fishery resource productivity, the biodiversity and habitat preservation, and the social and economic goals. The EAF implementation methodology addresses these principles at the level of a selected fishery and strives to ensure that the above mentioned values are given key focus in a management plan. As described by the EAF planning and implementation steps [6], the objectives for the fishery under consideration have to be specified and prioritized for each of these values, following which a set of multi-disciplinary indicators and reference points can be defined. These indicators are monitored through various approaches including scientific assessments, catch or effort monitoring systems, rapid appraisal surveys with participatory methods. The scientific indicators build on spatially explicit models on the distribution of fisheries activity, the socio-economic structure of fishing communities, the concentration and abundance levels of commercially desirable fish stocks, the distribution of non-target species and vulnerable aquatic habitats. For an effective management response, such monitoring across disciplines and levels of data/information/knowledge elaboration requires a well organized (often institutionalized) workflow which orchestrates the inputs of the EA-CoP actors vertically (from the data manager, through the scientist and up to the fishery manager and policy maker) and horizontally (across concerned disciplines).

The EA-CoP group: consists of the various actors involved in support to the (yet to be selected) EAF regional project on pelagic ecosystems in the context of African tropical or sub-tropical Large Marine Ecosystems (LMEs). These are fishery policy makers and managers, fishery scientists, ecologists, data providers, and fishermen representatives, representing national or regional institutions, among which FAO, IRD, UNESCO, FIN, Terradue and the Regional Fishery Body which will be selected (IOTC, SWIOFC, CECAF, GFCM).

The Goal: The iMarine Data e-Infrastructure has the capacity to offer a shared virtual environment integrating multi-disciplinary data sources, and supporting cross-cutting scientific analyses, generation of live reports, handling of business workflow, and social communication among EA-CoP’s actors. Such capacity will be exploited as a live dashboard in support to the EAF implementation and planning steps. The “EAF dashboard” will offer an interface showing in organic way the complete data/knowledge/decision making workflow and its feedback loops, highlighting the actors involved at each step (from data managers through scientists, managers and decision makers), the implementation status of each step. It will offer a capacity to relate the actors vertically or horizontally with social tools, as well as links to and live summaries of the data and knowledge they are responsible for.

The Approach: The project will not pretend to comprehensively address the requirements of the selected CoP, but it will set the “EAF dashboard” framework for the selected context and expand some of its most promising branches. It is expected that additional funding will be articulated around such kernel.

As first development goal, the “EAF dashboard” capacity will be set-up to reflect the Organizational level of an EAF initiative, building on interoperability with IOC’s ocean expert system, and FAO’s fishery institutions database (FIGIS). The use of the FOAF semantic model such as implemented by IRD in its Ecoscope system will be assessed in support to interoperability among these resources and other dashboard agents such as the data collections. In order to enhance the cohesion of the EA community, the actors will be also connected through social media tools. In the organic “boxes” where they fulfil a role, the actors will be able to load documents, organize and archive indicators and reference points, write reports, etc.

As second development goal, and following the fractal’s metaphor, some of these organic “boxes” will exploit the full capacities of iMarine’s Virtual Research Environments where access to rich data collection sources will be granted. Indeed, iMarine’s technologies offer excellent services to continue the development of such an integrated environment and will contribute to leverage the capacities of the information systems currently supporting these EA-CoPs by making them part of the whole process and offering new perspectives to the analysts using them. As example, Virtual Research Environments will be set-up for scientific information on Tuna fisheries, including catch, effort and by-catch indicators, building on SDMX and/or COST interoperability protocols (developed under business case 1, cf. Section ) with IRD’s Observatoire Thonier Tropical (OTT) and ObsTuna. Other “boxes” will be activated as virtual laboratories according to the most relevant co-funding opportunities.

As third development goal, semantic based services will assist the work of EA-CoP users throughout the EAF dashboard and its virtual research environment by enabling the discovery of resources available in the iMarine data e-Infrastructure. This will be achieved by mapping linked open data published by the various infrastructures partaking in iMarine, in particular IRD’s Ecoscope (which supports discovery of and access to IRD’s impressive collections of scientific data sets on almost all the information scope of the EAF), and FAO’s FIGIS and TechCDR systems.

The Synergies: EAF CoPs are active and supported by a few key research programs and information systems in the following LME frameworks, and a final choice will be made upon further consultation: the Canary Current LME (CCLME), the Benguela Current LME (BCLME), and the Agulhas and Somali Current LME (ASCLME). As a matter of fact, FAO, UNESCO, and IRD are involved in various aspects of the implementation of the EAF in these LMEs through established Regional Fishery Commissions (IOTC, GFCM, CECAF, SWIOFC) and their supporting EU or GEF projects.

Depending on funding opportunities and needs expressed by the EA-CoP, other data sources already identified will be added: ESA’s established satellite imaging for SST, Salinity, Primary productivity, and new ESA products for monitoring the location of fishing vessels; StatBase [5] which provides the core statistical support tool on coastal artisanal fisheries to SWIOFC (Commission); Fritjoff Nansen database on species abundance and related mapping applications; FIGIS which holds the global inventory of resources and fisheries; GOOS Africa with many compiled environmental maps; and OBIS with distribution of vulnerable species.

Outreach

Outreach will play a key role in promoting the exploitation of the iMarine data e-Infrastructure by the members of the EA-CoP. To this end, the project will implement its strategy for Communication, Dissemination and eTraining, taking as a starting point dedicated value propositions and core messaging to EA-CoP stakeholders, policy makers and society at large. The project will thus aim to ensure that communication and promotional activities cater to the real-world needs and interests of each of the different target groups by defining iMarine “take-home” messages for each one as the project develops over time. The ultimate goal is to widely disseminate three main outcomes of the project, that is:

1. 
   The iMarine Guidelines and Best Practices.
   

3. 
   The iMarine data infrastructure and its Virtual Research Environments.
   
4. 
   The enhanced services developed.
   

As most of the actors targeted operate in a market with clear objectives and business plans, carefully defined strategies are key to capturing and retaining their attention throughout the life-time of the project and beyond. One of the main ways in which the project will seek to pursue this goal is through the design and delivery of training and live demos of outcomes, either face-to-face or remotely, in order to capitalize on opportunities for interaction and knowledge exchange between the major actors involved. The live demos will be filmed and hosted on the project’s website along with the remote training, so as to digitally capture content of relevance for re-use.

To underpin this approach, the website will be an integrated and interactive Knowledge Management Channel with tagged multimodal content, enhanced wiki services and discussion forums, as well as a dedicated area for the iMarine Board (document uploading and sharing facilities with the possibility of one-to-one interaction). The Channel eTraining Environment (Virtual Classroom, Meeting and Mini Conference facilities: one-to-one, one-to-many with Moodle Chat and Just-in-Time) will host eCourses chiefly targeting EA-CoP application end-users, application developers and infrastructure managers, with potential sustainability beyond the project.

To maximize outcomes and broaden the scope for knowledge exchange, iMarine will aim to forge alliances with relevant initiatives and associations for the co-organization and joint promotion of outreach activities. To this end, EU project meetings, scientific conferences, and especially EA-CoP meetings and data-Infrastructure events, will be monitored in order to establish appropriate synergies. Strategic alliances will also be sought to reinforce outreach across all key iMarine outcomes, offering a gateway across disciplines and geographical boundaries to support the expansion of the user-base and sustainability. Such alliances will leverage strong partner networks and target related ESFRI projects, particularly the Environmental cluster, as well as European and international initiatives related to fisheries and marine ecosystems, including those connected with the EC’s Innovation Union Flagship.

The Channel will serve as a key reference point for the EA CoP, as well as for policy makers and society at large, by offering access to iMarine Board interviews, insights and testimonies related to Guidelines and Best Practices, as well as a showcase for the major benefits delivered by the data e-Infrastructure, its services and role in underpinning Virtual Research Environments to broaden engagement, integration and interoperation of different communities and Research Infrastructures, primarily conveyed through the eyes of end-user communities.

A Final Event will serve to strategically underpin the specific policy and technology outputs of iMarine by bringing together all the major stakeholders and policy makers from Europe and globally. Plenary, parallel sessions, round-table debates and interviews will shine the spotlight on sustainable governance and recommendations by defining clear “take-home” messages for the diverse stakeholders, along with the distribution on an eBook on iMarine Board Insights (interviews and primers) for the wider community and an eBook on Policy, Resources and Standards for the EA CoP and policy makers. Key features of the event will be digitally captured for content re-use and media partnerships, press invites, packs and conference will seek to enhance the dissemination of outcomes.

Proposal Activities Overall Strategy

The objectives of the project will be achieved through the cooperation of Networking, Service and Joint Research Activities.

The overall objective of the iMarine Networking Activities is to (i) develop the governance mechanisms of the iMarine data infrastructure and formulate a set of organizational and policy recommendations regulating resource sharing and service provision; (ii) make concrete steps towards the harmonisation of data, metadata and aspects of application level semantics as a basis for achieving higher interoperability levels; (ii) define and implement an effective communication and training strategy within and outside the EA-CoP.  This objective will be reached by: (i) operating  a iMarine Board involving representatives from major organizations working in the Fisheries Management and Conservation of Marine Living Resources, some of which already taking part in relevant EA-CoP Committee; (ii) performing activities aimed at enhancing coherence at data, schema and application level, with a special focus on reference data management, geospatial data, statistical data exchange and semantic technologies; (iii)  disseminating project outcomes to major EA stakeholder groupings, policy makers and society at large as the project develops over time; (iv) designing and delivering training and live demos of outcomes, either face-to-face or remotely, in order to capitalize on opportunities for interaction and knowledge exchange between the major actors involved.

The Service Activities aims at deploying and operating the iMarine data e-Infrastructure. Activities, procedure, mechanisms and resources needed to guarantee a 24/7 service availability will be put in place in order to assure a good quality service. These will take into account  policies and guidelines established by the iMarine Board.  The iMarine data e-Infrastructure will not be a closed environment. Rather it  will be obtained by exploiting other existing infrastructures , other implemented services and other available resources. Many of these resources will be provided by organizations of the EA-CoP, some of which are represented in the iMarine Board. This will facilitate appropriate synergies with these organizations on provision of resources and on development of new services. Service activities will also be responsible for supporting a number of Virtual Research Environments serving the needs of the three business cases relevant to the implementation of the Ecosystem Approach which  have been described in Section .

Finally, the iMarine Joint Research Activities aim at delivering the system offering the rich set of functionality required to implement the EA. This system will be obtained starting from the gCube framework, integrating components provided by other open-software initiatives, like OpenSDMX, FishstatJ, and eDAMIS, and enhancing it with new facilities.

Guidelines and best practices discussed and agreed in the context of the iMarine Board will be taken into account in this process. Enhancement will be done at the level of: (i) gCube enabling services, by enhancing their capability of interfacing external infrastructures, enforcing resource usage policies and distributing process executions among several computational platforms; (ii) set of data management services,  by making them powerful enough to  manage the large variety of data the EA deals with; (iii)  set of services for data consumption, including data mining and extraction of knowledge from raw data, data provenance, and data transformation; (iv) facilities,  like programmatic APIs  and standard protocols, for enabling the consumption of the iMarine resources  by third-parties.

The three activities just described are organised into cooperating work packages whose interaction is depicted in Figure .

Figure . iMarine Overall Activities

The three activities will operate in synergy toward the achievement of the D4Science-II objectives. In particular, the following work packages hold a critical role in linking the activities assigned to each area of the project:

• 
  NA2 Project Management: this work package organises the overall scientific and technical activities, monitor the achievements and recommend appropriate actions enhancing the activities effectiveness toward the project objectives;
  
• 
  NA3 Governance and Policy Development: this work package manages the iMarine Board which develops the policies driving the iMarine data e-Infrastructure operation and the guidelines for its implementation;
  
• 
  SA1 iMarine Data e-Infrastructure Deployment and Operation: manages the iMarine data e-Infrastructure by providing hardware resources, deploying and maintaining the infrastructure core services, providing monitoring and accounting information, and defining procedures to manage the infrastructure;
  
• 
  JRA1 – iMarine Data e-Infrastructure Enabling-technology Development: enhances the gCube enabling services by interfacing external infrastructures, distributing process executions to several external computational platforms and enforcing resource usage policies defined by the EA-CoP;
  
• 
  JRA2 – Data Management Facilities Development: integrates, enhance, and develop a set of services for managing statistical data (including but not limited to time series), marine biology, environmental data such as satellite data and sensor data; taxonomies, ontologies, and code-lists; structured and semi-structured textual data; and binary data;
  
• 
  JRA3 – Data Consumption Facilities Development: develops a set of facilities for supporting the data processing tasks the EA-CoP faces, e.g. data discovery and retrieval, generation and manipulation of data, mining and extraction of knowledge from raw data, generation of data provenance information, data transformation, and visualization and simulation of scientific data.
  
• 
  SA2 – Virtual Research Environments Deployment and Operation: deploys and operates the Virtual Research Environments running in the iMarine data e-Infrastructure by developing vertical solutions, integrating community applications and services with gCube services, developing common interfaces and tools, and providing community data resources.