The total is 378 OBSs, of which 127 are wideband and 9 are broadband.  212 of the instruments can be deployed for up to 1 year, including 88 wideband and 9 broadband OBSs.

Experiments using a large number of OBSs for passive seismology include the 2007-2008 NEAREST experiment in the Gulf of Cadiz (24 German wideband OBSs and the Italian GEOSTAR Observatory for 11 months), the 2009 Thales Was Right experiment offshore Greece (37 French and German OBSs for 2x 6-month deployments), the 2012-2013 RHUM-RUM experiment around Reunion Island (57 French and German wideband and broadband OBSs for 1 year) and the AlpArray experiment (pre-proposal stage: up to 48 wideband and broadband OBSs for 1-2 years).  Smaller long-term experiments include the FOMAR network maintained by ROA (Gulf of Cadiz and Alboran sea, 3 LOBSTER OBS for at least one year, using 6-month deployments) and the Lucky Strike network on the Mid-Atlantic Ridge maintained by IPGP-EMSO-INSU (5 INSU-IPGP OBS for 5+ years starting in 2007, using yearly deployments).

The OBS parks also have a broad experience in active seismic experiments, having conducted numerous surveys in in different tectonic settings (passive and active margins and mid-ocean ridges) over the past 20 years.

Several of these parks use the same or similar technologies: The UK, CSIC and French INSU parks all use instruments derived from the Scripps Institution of Oceanography OBSs, and the two German parks and the Italian, Portuguese and ROA parks use KUM/SEND OBS acquisition systems.  Enhanced collaboration between their technical teams could help each park to progress faster and more efficiently.

2. Scientific and Technical Issues and Opportunities

These sessions focused on seismological techniques that are enhanced or possible only at sea and  on the difficulties associated with OBS data.  Because of time constraints, the techniques were not discussed in detail, but they include sea floor compliance (the sea floor deformation under long-period ocean waves, sensitive to the subsurface shear modulus), ambient noise tomography (the strong signal in the microseism band from 10s to 1-2 Hertz is very favourable for these measurements), marine mammal monitoring and studies of infragravity wave sources (important for the generation of the Earth’s seismic “hum” and for nearshore processes).

There was also discussion of some of the problems affecting data quality that are specific to OBS recordings: specifically noise, timing and orientation.  OBS timing, based on synchronization before and after the experiment and the use of a temperature-controlled or microcomputer compensated crystal oscillator in between, is generally better than 1 sample spacing, but it can be worse if the instrument clock stops before the end of the experiment or if the clock implementation is imperfect.  Luis Matias talked about how to confirm and, if necessary, correct instrument timing using noise correlation techniques.

Orientation is a problem because compasses placed next to a seismometer or in the recording package are generally dominated by the magnets in the seismometer, disk drives, or other instrument parts.  One solution that DEPAS is looking into is using gyrocompasses, although cost and power consumption are a problem.  Another, mechanical, solution is to have a compass on the outside of the instrument, with a clamp that is held off the needle until the instrument is at the bottom using a dissolvable (for example, sugar) block.  The compass is farther from the instrumentation, but must be aligned with the seismometer and therefore not too far.  The US group of Leroy Dorman tested this method, we don’t know how well it worked.

In the meantime, and for pre-existing data, reorientation can be performed using teleseismic data and possibly noise correlation.  One problem may be that short-period sensors are not sufficiently sensitive to teleseismic data, and local earthquakes are problematic because of 3D propagation effects.  In active seismic experiments, reorientation is usually performed using the relative amplitude of airgun recordings on the geophone horizontal components.  A small airgun array could be used to orient OBSs used in passive experiments as well.

Noise is a particular problem.  In particular, the microseism peak can heavily degrade the signal from 10s to 5 Hz, and currents have a strong effect on the longer period signal.  Also, the motion of the sea floor under long-period ocean waves (compliance) can limit the long-period noise floor.  Methods have been developed to remove long-period noise on the vertical components due to currents and ocean waves, but there is no solution for the horizontal noise except to reduce the instrument’s profile or bury it.  Broadband instruments may be more sensitive to this problem than wideband instruments because of their larger size.  Open access to data would help us to evaluate this.  Shorter-period noise (5-10 Hz) may also be related to ocean currents, but this is largely unknown and should be investigated.

Another problem with long-term OBS deployments is their relatively high cost: the instruments must be deployed from a ship and they must carry all of their energy: long deployments require thousands of dollars worth of lithium batteries per instrument.  The instruments must also be highly reliable as there is no way to verify their data until they are recovered (for this purpose, some groups are working on data shuttles or on the possibility to install acoustic modem aboard the OBS) and it is extremely expensive to recover an instrument if it does not release from the sea floor when it is supposed to.

Finally, much OBS data is archived locally within the parks, in non-standard formats.  This not only runs the risk of losing the data, it also doesn’t allow processing methods developed by one park to be easily be used by another.  Nor does it allow standard seismological quality control to be applied.  And the data may never become available to the general public.  There are some issues about data propriety that are outside of the scale of this workshop, but we feel that data should be put into standard databases (with fully restricted access if necessary) as soon as possible, even if this issue is not fully resolved.

3. What is needed to improve the parks and their offer

The European OBS White Paper

The pre-meeting White Paper specified a number of “good practices” that should improve the use of OBS data by the general scientific community.   These are:

•Maintenance of a common website outlining hardware and services of the different parks and the local instrument request procedures

•Organizing yearly organizational and technical meetings

•Helping the OBS parks to transform their data into a seismological standard format and adopting policies that enforce the data fully after a reasonable embargo period

•Evaluate the data success rate of each experiment on a European level

•Coordinate funding developments and test experiments

•Clarify rules for obtaining ship time and the costs of the ships and work towards a unified request for ship time and OBS pool access

•Establish a limited pool of equipment and ship-request mechanisms for “rapid response” experiments

We discussed  this White Paper in detail.  The participants agreed that most of the ideas were sensible, but that they were currently impractical given the small number of people and resources available to OBS parks.  A two-part action plan was therefore proposed: 1) focus on the most useful and achievable “good practices” and 2) find resources for implementing these practices.

The new list of good practices is:

•Organize yearly technical or organizational meetings

•Help the OBS parks to archive their data in a European (e.g. WebDC.eu) or national seismological data centre that can permanently provide data over the web in a seismological standard format.

•Develop and distribute tools for standard OBS data postprocessing to remove noise, orient components and verify clock corrections

•Encourage openness about success rates and problems.  Create tools to aid in evaluating this, using simple quantitative measures.

•Lobby for a single (or at least coherent) process for funding and ship time.

•Encourage rapid response/ship time mechanisms in countries with OBS parks

We removed the common website, because none of the current OBS parks can dedicate the resources to run it.  We combined the yearly meetings between organizational and technical, because most organizational issues will be more efficiently resolved by a focussed committee.  We changed “seismological standard data…” to “archive their data in a … data centre” because this will ensure standard data and, moreover, will allow uniform evaluation of data quality. We added “develop and distribute tools” because this will greatly increase the usefulness of OBS data to the greater seismological community, and these tools are needed to incite parks to put their data into the data archives.  We removed “evaluate the data success rate” because that is part of putting all of the data into a qualified data centre.  We removed “coordinate funding developments and test experiments” because it was unclear. Finally, we modified the “ship time” and “rapid response” sections to what is actually feasible. 

Although the White paper focuses on passive seismology, during the meeting it was noted the importance of active seismic experiments in different aspects of seismological research. Modern modelling techniques (joint travel time inversion of refracted and reflected waves, full waveform inversion, etc) can provide accurate information on the structure, fault geometry and properties of Earth’s sediments, crust and upper mantle in all oceanic settings (e.g. active/passive margins, ocean ridges), which have direct applications in many scientific domains (crust generation, fracturing and alteration, earthquake seismology, slope failure, gas hydrates, hydrocarbons, etc). Furthermore, data collected during active seismic experiments should also be archived as continuous as well as shot-based data whenever possible, as these experiments may have unique and densely sampled information about otherwise unexplored regions. Active seismics are therefore now mentioned in the White Paper in the context of integrated land-ocean data.

We also agreed to modify the following White Paper text:

•We will note that the OBS parks will provide RAW data (non re-oriented, clock verified, or noise removed) and that it us up to the user or data centres to apply these corrections, but that the we will try to provide the tools and information to do so

•The paper will state that rapid response requires confidence between the park and the state.  Rapid response mechanisms are installed in Germany and UK, and worked well in Germany after the Chile event.  On the other side, a rapid response of French OBSs off of Haiti after the disastrous event there was never reimbursed.

•In the technological innovation section:

oAdd “tilt” and other data that help seismology to the listed multi-parameter measurements

o Remove “new power sources” (we don’t have the resources or expertise to develop these).

oFor “data shuttles”, we add the possibility of the instrument spitting out data and being lost (no recovery).

Finally, we will add a section specifying the support needed to implement these good practices.  The individual parks do not have the resources necessary to implement these practices: an overriding structure is needed.  To this end, we have formed a committee whose goal is to educate the European community on the importance of OBS work and its specific needs, and to find funding support necessary to implement the good practice points. 

The Committee for the Harmonization of European OBS Parks (CHEOPs)

Implementing these good practices will require more resources than currently exist for OBS parks .  The OBS parks don’t have enough personnel to allocate to coordination and cooperation and some OBS parks are reluctant to share data that they worked very hard to obtain.  Until now, the parks have been working separately and have not taken much advantage of major national and European seismological initiatives.  We must make it as easy as possible (and in their interest) for parks to put data in data archives, and we must present a clearer vision of our instruments and the unique constraints/requirements of our work to the seismological community.

We have formed a committee consisting of members from many of the OBS Park countries.  The committee’s primary objective is to find European and national-level support to allow the implementation of the policies outlined in the White Paper. Specific tasks we identified are:

¥Publicize and find support for OBSs in applicable national/European structures: OBS seismology, in all its aspects, needs to be better known by land seismologists. Collaborations between land and marine seismologists should be encouraged.  This is especially relevant for the study of margins and islands.  The Committee will clarify the advantages and disadvantages of each type of instrument (broadband, wideband and short-period) and emphasize the specific needs and constraints of OBS seismology:

ÐOBS data need to be prepared before they can be placed in data archives: we cannot use “standard” seismological loggers because of their high power consumption, size and relatively unstable internal clocks, so the data is stored in custom loggers with proprietary formats.  These data must be converted afterwards into a standard format taking into account the clock drift and, ideally, verifying the clock corrections, reorienting the components in geographical coordinates and removing seafloor-specific noise.  This requires a much larger infrastructure than is needed for archiving land station data

ÐOBSs are not available off the shelf: Developments are still being made to cope with the harsh seafloor environment, need for minimal power consumption and unique seafloor noise sources.  A technical group and resources are needed to develop new prototypes and to transform these into tomorrow’s instruments.

ÐOBS deployments require significant resources and coordination: Ship time must be acquired, the capacity of the ship to perform the work verified and the work required fully explained to the ship’s crew well before the mission.  Funding for the ship time is expensive and sometimes must be requested separately from the ship time request. Ship schedules must be strictly adhered to, instruments fully verified before deployment, technical problems solved at sea and all deployment/recovery parameters recorded (including clock drift).  Finally, OBS consumables can be expensive because of the need for energy autonomy over long periods.

¥Link OBS parks with European structures: the national OBS parks should be informed of and collaborate with existing or upcoming European structures and they should be involved with new national and European seismological initiatives.

•Create an OBS-centric European structure.   A European-level framework should be created to 1) facilitate the transfer of OBS data into data archives and 2) provide a coordinated view of European OBS parks.  This framework should include engineers to develop and distribute tools for processing OBS data and help put the first data from each park into the archives.  It should also coordinate and distribute information and ideas for advances in processing methods and technological developments (for example, instrument orientation, noise reduction, clock corrections and new data loggers) and provide a unified entry/information point for seismologists interested in using OBSs.

¥Identify scientific targets and the type and number of OBSs that will be needed to study them.  Examples include seismic hazards, active margins (Greece, AlpArray: 45 wideband+, 1 year+), lithosphere-asthenosphere interactions (Atlantic Lithosphere + MAR plumbing: 100 OBSs (wideband+?)?), mantle convection and plumes (RHUM-RUM 50+ wideband instruments), passive margins (hydrocarbons, failures), mid-ocean ridges (fault and magma processes) and non-solid earth (marine mammals, physical oceanography).

¥Define which broad issues are currently the most pressing : for example, Techincal Developments, Scientific Projects or Increasing the Number of Instruments?  The emphasis could shift with time as the most pressing issues are resolved.

¥Find funding for technical meetings: working at sea with free fall instruments is a challenging task: the harsh environment where OBSs are deployed and the inability to confirm data before the end of a mission can result in a non-negligible number of sites which fail to recover good data. Yearly technical meetings would help engineers and technicians of the European OBS parks to share technological advances and problems and to collaborate to improve the reliability of the instruments and their data quality. Members of the committee will look for funding resources within their national structures, or European projects, to organize these recurrent meetings.

The committee members are: Wayne Crawford (France), Mechita Schmidt-Aursch (Germany), Tim Henstock (UK), Valenti Sallares and Antonio Pazos (Spain) and Giorgio Mangano (Italy)

Coordination with EPOS

EPOS appears to be the European infrastructure the most adapted to our project.  We need to reinforce ties with them and to see if they can help us to get started on the path to being the most visible and useful possible.  This includes obtaining funding for initial meetings (including the first technical meetings) and advice/support for starting our own framework/infrastructure.

Technical Meetings

The first technical meeting should be held as soon as possible and unite all of the OBS parks’ technical groups so that the engineers understand what is available at the other parks.  We need to find funding for this meeting (EPOS?).  Future  meetings could be specific to one issue or instrument type, with possibly multiple meetings attended by interested park subsets. For example, groups with « SEND » electronics could meet at one of their centres, while groups using « L-CHEAPO » electronics could meet at another centre.

EGU Presentation

Finally, we discussed our 2012 EGU presentation in the “Improving seismic networks performances: from site selection to data integration » sesssion.  We will present 1) the scientific interest of OBS seismology, the existing equipment and recent experiments ; 2) the need for improved collaboration, specific issues (noise, ships, etc) and the very small and dispersed OBS community working on these problems and 3) the framework for action, based on the new White Paper.

European OBS Parks Workshop Summary

Dates: February 23-24 2012

Place: Institut de Physique du Globe de Paris

Participants: Wayne Crawford (INSU OBS Park, France), Giuseppe Danna (INGV OBS Park, Italy), Davide Embriaco (INGV OBS Park, Italy), Ernst Flüh (GEOMAR OBS Park, Germany), Audrey Galvé (GeoAzur OBS Park, France), David Graindorge (IFREMER-UBO OBS Park, France), Ingo Grevemeyer (GEOMAR OBS Park, Germany), Tim Henstock (UK OBS Park), Frauke Klingelhöfer (IFREMER-UBO OBS Park, France), Sergei Lebedev (Ireland), Marcia Maia (INSU, France), Giorgio Mangano (INGV OBS Park, Italy), Luis Matias (IDL OBS Park, Portugal),  Antonio Pazos (ROA, Spain),  Milton Plasencia (Antarctic Argentine-Italy seismic network, Italy), Maria Sachpazi (Greece), Valenti Sallares (CSIC OBS Park, Spain), Mechita Schmidt-Aursch (AWI-DEPAS OBS Park, Germany), Satish Singh (France), Hans Tybo (Denmark), Frederik Tilmann (GFZ Potsdam)

The goal of the workshop was to find common ground, interests and methods for collaborating between European OBS Parks in order to improve our offer to European scientific users.

The first part of the workshop focused on the current state of OBS parks and their recent or planned experiments.  The second part focused on scientific and technical issues and opportunities.   Finally, we discussed what is needed to improve the parks and their offer.  The basis for this discussion was the preliminary White Paper written by Wayne Crawford, Frederick Tilmann and Alex Brisbourne.  A modified White Paper based on this discussion is proposed.

1: OBS parks and recent or planned experiments

8 OBS parks were presented: 3 in France, 2 in Germany and Spain, and one from each of UK, Portugal and Italy.  Hans Thybo was also present as a representative of the Danish seismology community, which has obtained funding for a seismological pool which should include at least 24 OBSs for the Iceland Plume Experiment .  Finally, there is an effort in Poland to acquire OBSs, but the responsible persons did not feel it was sufficiently advanced to participate in this meeting.

The presented OBSs were: