Minutes of the 22nd Meeting of the Marine Observation and Data Expert Group (MODEG)

Date: Wednesday and Thursday 1-2 October 2014

Location: Belgian Federal Science Policy (Belspo) – Louizalaan / Avenue Louise 231, B-1050 Brussels

Minutes Prepared by the EMODnet Secretariat

Go to Meeting Agenda and Presentations

Meeting Participants

Day 1 - Wednesday 1 October 2014

Welcome and Introduction

Welcome and introduction to the meeting by the MODEG Chair and the EMODnet Secretariat, extending thanks to the Belgian Science Policy (BELSPO) for kindly hosting the MODEG meeting at their offices. The main purpose of the meeting was to (i) gather inputs and perspectives on the European marine data challenges from a range of stakeholders and invited technical experts; and (ii) provide an update and platform for discussing EMODnet progress and future developments.

Frank Monteny (Belspo) gave a brief overview of Belgian Science Policy (BELSPO) support to marine research activities in Belgium. Belspo funds scientific research but also supports the Belgian research vessel Belgica, which is planned in the coming years to be replaced by a new ship. Belspo hosts a number of institutes and organisations, namely the Secretariat of EuroGOOS and the newly established Secretariat of the Joint Programme Initiative (JPI) Climate.

Erik Buch (EuroGOOS Chair) gave a short overview of the EuroGOOS governance structure, its mission and five main objectives outlined in its strategic plan 2014-2020 to support European-scale operational oceanography in the context of the Global Ocean Observing System (GOOS). He drew attention to the upcoming 7^th EuroGOOS Conference in Lisbon where EuroGOOS will celebrate its 20^th anniversary. More information about European operational oceanography and EuroGOOS including the EuroGOOS Strategy 2014-2020 and information about the Conference can be found here: http://eurogoos.eu/.

Perspectives on European marine data management from UNESCO/IOC/IODE

Ward Appeltans (UNESCO/IODE) presented some of the major developments in history of UNESCO’s International Oceanographic Data and Exchange programme (IODE), and perspectives on the current and future European marine data landscape in context of global marine data management practices. In his presentation he clarified the key components of the marine data management process, highlighting the current shift in the model of data-flow from being based primarily on National Oceanographic Data Centers (NODCs) to a new model with associated data units where data is stored/handled and made available by many different types of organizations. He also observed that the current cost model of data management practices from data collection to services including operational/infrastructure is under pressure as it is not always clear who should cover the considerable investments involved, questioning whether this could mean the end of open access to data?

Questions, comments and discussions

A discussion followed on the costs of marine data management and trends in openness of access to data:

• In general MODEG members noted that marine data are becoming more openly available in most thematic areas. However, in some cases, even projects with an ‘open data’ policy were not necessarily able to make all the data freely available because of the high costs involved.

• While the cost of data management is considerable, it is very little compared with the cost of data acquisition/monitoring/observations. This point must be made very clear when advocating the added value of open access and to stimulate re-use of data to maximise returns on investment.

• Evidence shows there is a growing need for more information and more data for a range of applications and this is likely to further increase in the future. The need for a wide range of data and products has become so obvious that even politicians are becoming aware of it, which helps to secure financial support to make more data available without restrictions on use.

• DG MARE iterated that the European Commission is supporting an open access to data policy both in EMODnet and Copernicus. The EMODnet checkpoints should help to make the case for keeping the observations going as they are to reveal what we could do if we have more data and observations.

• To justify the investments in data management and open access, it will become ever more important to connect the services to users and document success stories. Cost-benefit analyses may be necessary to secure the level of investments needed to make data fully available for re-use.

• In many cases data outputs of scientific research are not made available because scientific research activities still often do not budget costs for data management. The Marine Biodiversity Network of Excellence (Marbef) showed a good example by promoting allocation of 10% of budgets to data management as a rule of thumb.

The distribution of marine OpenData via distributed data networks and Web APIs. The example of ERDDAP, the message broker and data mediator from NOAA

Conor Delaney provided an overview of some promising new approaches and technical solutions for providing data to users from a distributed network of data resources. In particular, the ERDDAPs (Environmental Research Division's Data Access Program) could present opportunities to circumvent some of the current bottlenecks experienced by EMODnet thematic lots. ERDDAP is a data server that provides a simple, consistent way to download subsets of gridded and tabular scientific datasets in common file formats and make graphs and maps (for more information see: http://coastwatch.pfeg.noaa.gov/erddap/index.html). While it is new, the technology is gaining traction and is now use by various organizations including NOAA and it is being adopted at JRC.

Questions, comments and discussions

As the technology is relatively new, most MODEG members were not familiar with ERDDAP Questions focused on its functions, advantages and disadvantages:

• ERDDAP was developed to provide a WEB API (Application Programming Interface) to an existing distributed data network. It acts as a message broker receiving data requests (via html) and converting these into formats used by other data servers; queries data at source archives and returns the result of the query; acts as a data mediator converting sophisticated marine science data formats and converting them to formats more familiar to web/software developers; and creates a catalogue of data being published at remote data centers.

• Among the advantages of ERDDAP is that it is free under the apache licence and it is useful to select subsets of data. It is not in competition with other systems but can complement other systems opening it up to specific users.

• While the potential is huge, ERDDAP sits at the end of the data chain; the main challenge is the actual harmonization at the basis which is still a pre-condition for the technology to be functional.

• If useful, it could be made mandatory in the next EMODnet tenders that data should be made available in a format usable by ERDDAP technology.

• Regarding compatibility with MyOcean and SeaDataNet, Conor Delany indicated that for the time being it may not be possible to connect to these systems because the data needs to be freely accessible without barriers. Ideally SeaDataNet and Myocean would have their own ERDDAP server to allow harvesting data in a meaningful way. Conor Delaney agreed to evaluate the compatibility by tapping into SeaDataNet and MyOcean systems and share findings.

Perspectives on data management from the Dutch Dredging industry

Mark van Koningsveld presented on data management practices by the dredging industry and lessons learned through various projects and business activities of Van Oord, a world leading dredging company based in the Netherlands. He recommended to: engage in an open dialogue with industry to use their experience of handling large and complex datasets; always put the user first when developing data systems and standards; build upon open source components; think long term and respect confidentially by maintain same workflow for open and close access.

Questions, comments and discussions

• On usefulness of EMODnet at the moment: physics and geology are most relevant themes but in practice most of the data is directly obtained from the geological surveys or from clients’ measuring equipment/systems.

• Regarding confidentiality: confidentiality period for data release varies widely depending on the activity/project, but the key is to deal with the data exactly the same way, only to adjust the ‘protection fence’ depending on the restrictions. For some project (e.g. 3TU), a system is put in place that automatically removes restrictions after a given time period.

• On use of biodiversity data: sometimes but mostly habitat (e.g. coral reefs), natura2000 or protected areas information/data are being used in this case.

• Commercialisation of the data: Van Oord never charges for making available data; the cost for data management is covered by the cost of dredging activity in general, unless in cases of scientific collaborations when it may be covered by science grants.

Data needs for Quaternary and Prehistoric Landscapes/Archaeology: Objectives and next steps

Nic Flemming presented the context and future requirements for data to support research into submerged prehistoric sites calling for support from EMODnet and MODEG. He highlighted the important preparatory work in several initiatives including the recent publication of the European Marine Board Position Paper “Land Beneath the Waves” which presents the state of the art, key opportunities and recommendations to advance the European Continental Shelf Prehistoric Research (see www.marineboard.eu/publications). EMODnet could play an active role in considering the best access to required data so that the paleolandscape could be reconstructed.

EMODnet Progress update phase II – Secretariat

Jan-Bart Calewaert gave a brief overview of EMODnet progress since the last MODEG Meeting. This included the feedback on EMODnet portals from the Secretariat user-survey and interviews which revealed several points of for improvement by all thematic lots as well as the central portal.

Questions, comments and discussions

A question was raised on how far down the chain EMODnet should go to develop data products without competing with private companies. DG MARE clarified that the aim is to avoid competition with consultancies or other commercial service providers by providing raw data in the first place and only develop basic data products (i) as examples to demonstrate the potential use of the data; and (ii) that are of common relevance for many users/applications but which no one else would make available otherwise.

Day 2 - Start at 9:00 on Thursday 2 October 2014

Marine Indicators and the Continues Plankton Recorder Survey (CPR)

Nick Owens presented some of the exciting findings of more than 1 million samples taken in CPR surveys since it was developed almost 100 years ago by Sir Alister Hardy.

Questions, comments and discussions

• The cost of CPR equipment is around 10-15k; the highest costs are linked to hiring specialist taxonomists.

• Measuring methane is technically possible but the problem is that the measurements are taken in the wake of the ship which interferes with measurements of air-sea fluxes for example; it may be useful for a rough indications rather than accuracy.

• The analysis of micro-plastic occurrence in current and past samples shows how interesting it is to have old samples to assess new trends. Results using molecular techniques based on DNA extracted from the samples are also very promising.

• Efforts to increase the number of CPR surveys, in particular in collaboration with private sector (e.g. merchant shipping companies, supermarkets and oil companies), are promising. One of the objectives is to increase the number of tows in particular arctic and subarctic areas.

• Up to recently CPR data has been rather restricted but this is about to change as SAHFOS (Sir Alister Hardy Foundation for Ocean Science) intends to make the data more available, among others by developing a webserver

Data needs to support Maritime Spatial Planning at sea-basin scale

Jens Perus presented his perspectives on marine data needs for Marine Spatial Planning (MSP) from a local/regional perspective based on his experience in various projects in the Baltic.

Making data products for maritime planners was recognized as a very demanding and complex endeavor. One of the main challenges is to avoid communication problems by raising awareness and ensuring that all stakeholders and contributors start from the same basic understanding. Another challenge relates to structuring information ensuring that both open data and restricted data is taking in account in the process of marine spatial planning.

Overview of central EMODnet portal development and difficulties of retrieving data from thematic portals

Francisco Souza Dias presented the development of the EMODnet central portal functionalities through a series of use cases. These use cases aim to allow users to retrieve data from several thematic portals which is also a good test to verify whether the portals are providing machine to machine services including their operational status

Updates from the EMODnet sea-basin checkpoints

Nadia Pinardi presented updates from the Mediterranean Sea-basin Checkpoint with an emphasis on the first deliverables: the literature study and the oil platform leak bulleting service. A first data adequacy report is expected for release by May 2015.

A discussion followed on the response to the oil platform accident service and simulation done by DG MARE and Secretariat with following main points:

• Currently the possible impacts of the spill, e.g. prediction of possible effects on the seabed and the amount of birds/mammals affected by the oil, is not yet taken into account but if the data sources or expertise is available it could be considered (e.g. in collaboration with the human activities, habitats and biology portals).

• The simulation is supposed to be provided within 2 days, but in reality the checkpoint could do it much faster (1 hour is possible) as this is also what is required and set-up for the coast guard services in the Mediterranean.

• In reality,e.g. in Italy, the coastguard has the system inserted into their own operational flow. There are also other elements to be considered in a serious spill such as airplane observations and satellite images which would provide continual feedback to the predictions.

Regarding the literature study following points were made:

• The literature exercise has been very beneficial to assess availability and appropriateness of the data. The Mediterranean checkpoint report and database holds a wealth of information, including on data holders.

• There appears to be large confusion about the different systems in place and who is holding what kind of data in marine data management landscape in Europe, in particular in view of the MSFD requirements. Data is gathered and required by such a wide range of organisations, public bodies and private entities that it is often difficult to maintain an oversight. While MSFD may provide a great impetus to gather more data, it is currently not clear who is doing what and how the different data systems (e.g. as under development by some of the Regional Sea Conventions) relate to each other with risks of overlaps and redundancy.

Updates from DG MARE

DG MARE highlighted the enormous progress that has been made in the last year with all EMODnet projects in phase II now fully up to speed. The Secretariat has helped to standardise practices, provided visibility and improved the transparency in monitoring the projects. The EMODnet central portal has been instrumental in maintaining an integrated picture of EMODnet activities online. Checkpoints are providing new insights at the regional sea-basin level. A range of new calls for tenders have been launched for new checkpoints and for a coastal mapping project which would include a range of new mapping tools to help member states to map the difficult areas near to the coast. Next year will see the launch of a set of new calls for tenders including one on bathymetry with an aim to make available data with a higher resolution where this data is available.

DG MARE will also provide a grant to Euro-Argo to pay about 80% of costs for certain floats in addition to Member State floats (grant would cover deployment and data processing). Another call for tender will be aimed at developing a data input/ingestion system/process for EMODnet. This is necessary to ensure more data is entering the system which currently sits outside of the standardised and EMODnet-connected repositories.

It was highlighted that MODEG has been very successful with more than 20 meetings since its start in 2008. Now it is timely to engage more people in an even more open process involving not only the project coordinators and selected experts but many representatives from the more than 100 EMODnet organisations. How exactly this can best be achieved still has to be worked out in the coming months.

Next MODEG Meeting (date and location)

It was proposed to hold the next meeting in March or April 2015, dates to be confirmed. The possibility to adjoin to other meeting such as DIKE or IHO or the EMODnet steering committee meetings will be investigated.