International Federation of Digital Seismograph Networks

Thread: Re: FDSN Backbone

None
Started: 2008-07-07 20:01:04
Last activity: 2008-07-16 01:08:40
Rhett Butler
2008-07-07 20:01:04
Dear Colleagues,

In anticipation of discussions at Working Group 1 at our upcoming
meeting in South Africa (6 months away), I am proposing to expand the
size of the FDSN Backbone, with the addition of 35 globally
distributed real-time GSN stations (list attached). These stations are
currently available at the FDSN Data Center with documented metadata
and open real-time access.

I welcome discussion and other recommendations of stations from FDSN
Networks.

Best regards,

Rhett


On Aug 13, 2007, at 8:07 PM, Pawel Wiejacz wrote:

Dear Rhett,

I haven't been to Perugia nor Santiago but to my opinion:
"• Should the FDSN backbone of high quality stations be increased
from 200 to perhaps 300?"
Definitely. I would increase it from 200 to 400 so 300 looks a good
compromise with those who do not wish increase.

"On the topic of the FDSN backbone the issues to be considered were
the expansion of the backbone in terms of the number of stations, the
goal of making this a real time network, maintain the high quality of
the network and to insure that the FDSN backbone was properly
documented with metadata in FDSN SEED format. This issue was referred
to the WG I for consideration.
Maintain it a high quality network with documented metadata. Most of
these
stations are real time within some network. Why should FDSN get into
the business that someone else is already doing? As of new stations,
non-real time stations should not get accepted into the increased
pool,
as of existing non-real-time stations: if technical or financial
barriers have
prevented a local/national operator from making the station real-
time, how
FDSN expects to do that? Buy equipment or pay for data transmission on
behalf of some foreign operator? Unrealistic.

"Working Group I issues discussed lead the Excom to recommend that
the expansion of the FDSN backbone is a desirable thing especially
when one considers all of the new members. Each member should
designate at least one new station to be become part of the FDSN
network. The focus of the backbone is still to be broadband when
possible, geographically appropriate and if possible a station that
delivers data in real time."
I agree with this.

Pawel Wiejacz
Inst of Geophysics Polish Academy of Sciences.


On Aug 13, 2007, at 2:05 PM, Rhett Butler wrote:

Dear Colleagues,

I enjoyed our meetings in Perugia and fellowship after hours.

There has not been a review/update of the FDSN Backbone in many
years, and WG1 discussed the consideration of additional sites.

At the minutes of our meeting in Santiago, there was discussion of
increasing the size of the Backbone:

"• Should the FDSN backbone of high quality stations be increased
from 200 to perhaps 300?"

"On the topic of the FDSN backbone the issues to be considered were
the expansion of the backbone in terms of the number of stations,
the goal of making this a real time network, maintain the high
quality of the network and to insure that the FDSN backbone was
properly documented with metadata in FDSN SEED format. This issue
was referred to the WG I for consideration.

"Working Group I issues discussed lead the Excom to recommend that
the expansion of the FDSN backbone is a desirable thing especially
when one considers all of the new members. Each member should
designate at least one new station to be become part of the FDSN
network. The focus of the backbone is still to be broadband when
possible, geographically appropriate and if possible a station that
delivers data in real time."


In the spirit of the discussions and with an interest in soliciting
ideas from other Networks, I have reviewed the current contribution
of the GSN to the FDSN Backbone. In the attached spreadsheet
(information is from the FDSN station list), I submit 27 stations
for consideration within an expanded FDSN Backbone. All of the sites
have or will have open real-time data.

The USGS has 5 of 9 new GSN sites in the Caribbean are completed,
all will have open, real-time data. It may be of interest to
consider these sites for the Backbone.

There are a number of additional GSN sites in the US which are not
proposed.

I am receptive to your perspective and comment.

Best regards,

Rhett

<for FDSN Backbone.xls>_______________________________________________
fdsn-wg1-stations mailing list
fdsn-wg1-stations<at>iris.washington.edu
http://www.iris.washington.edu/mailman/listinfo/fdsn-wg1-stations




  • Seiji Tsuboi
    2008-07-09 02:19:34
    Dear Rhett and members of FDSN WG-1

    Thank you very much for your email regarded with the FDSN backbone discussion.
    I was asked by Torild last June to start discussion on the definition of FDSN
    backbone network, which was originated through IRIS-DMC - ORFEUS coordination
    meeting. I am sorry that I forgot to send email on this issue to the WG-1
    mailing list. I would be grateful if those who have ideas for updated
    definitions of FDSN backbone network would send their ideas to this mailing list
    now. To increase the number of stations should be one of the options that we
    should consider. I will update the mailing list as soon as possible, since the
    member of the mailing list is somewhat out of date.

    Best regards,
    Seiji Tsuboi

    -----Original Message-----
    From: fdsn-wg1-stations-bounces<at>iris.washington.edu
    [fdsn-wg1-stations-bounces<at>iris.washington.edu] On Behalf Of Rhett
    Butler
    Sent: Tuesday, July 08, 2008 8:01 AM
    To: fdsn-wg1-stations<at>iris.washington.edu
    Subject: Re: [fdsn-wg1-stations] FDSN Backbone

    Dear Colleagues,

    In anticipation of discussions at Working Group 1 at our upcoming
    meeting in South Africa (6 months away), I am proposing to expand the
    size of the FDSN Backbone, with the addition of 35 globally
    distributed real-time GSN stations (list attached). These stations are
    currently available at the FDSN Data Center with documented metadata
    and open real-time access.

    I welcome discussion and other recommendations of stations from FDSN
    Networks.

    Best regards,

    Rhett


    On Aug 13, 2007, at 8:07 PM, Pawel Wiejacz wrote:

    Dear Rhett,

    I haven't been to Perugia nor Santiago but to my opinion:
    ". Should the FDSN backbone of high quality stations be increased
    from 200 to perhaps 300?"
    Definitely. I would increase it from 200 to 400 so 300 looks a good
    compromise with those who do not wish increase.

    "On the topic of the FDSN backbone the issues to be considered were
    the expansion of the backbone in terms of the number of stations, the
    goal of making this a real time network, maintain the high quality of
    the network and to insure that the FDSN backbone was properly
    documented with metadata in FDSN SEED format. This issue was referred
    to the WG I for consideration.
    Maintain it a high quality network with documented metadata. Most of
    these
    stations are real time within some network. Why should FDSN get into
    the business that someone else is already doing? As of new stations,
    non-real time stations should not get accepted into the increased
    pool,
    as of existing non-real-time stations: if technical or financial
    barriers have
    prevented a local/national operator from making the station real-
    time, how
    FDSN expects to do that? Buy equipment or pay for data transmission on
    behalf of some foreign operator? Unrealistic.

    "Working Group I issues discussed lead the Excom to recommend that
    the expansion of the FDSN backbone is a desirable thing especially
    when one considers all of the new members. Each member should
    designate at least one new station to be become part of the FDSN
    network. The focus of the backbone is still to be broadband when
    possible, geographically appropriate and if possible a station that
    delivers data in real time."
    I agree with this.

    Pawel Wiejacz
    Inst of Geophysics Polish Academy of Sciences.


    On Aug 13, 2007, at 2:05 PM, Rhett Butler wrote:

    Dear Colleagues,

    I enjoyed our meetings in Perugia and fellowship after hours.

    There has not been a review/update of the FDSN Backbone in many
    years, and WG1 discussed the consideration of additional sites.

    At the minutes of our meeting in Santiago, there was discussion of
    increasing the size of the Backbone:

    ". Should the FDSN backbone of high quality stations be increased
    from 200 to perhaps 300?"

    "On the topic of the FDSN backbone the issues to be considered were
    the expansion of the backbone in terms of the number of stations,
    the goal of making this a real time network, maintain the high
    quality of the network and to insure that the FDSN backbone was
    properly documented with metadata in FDSN SEED format. This issue
    was referred to the WG I for consideration.

    "Working Group I issues discussed lead the Excom to recommend that
    the expansion of the FDSN backbone is a desirable thing especially
    when one considers all of the new members. Each member should
    designate at least one new station to be become part of the FDSN
    network. The focus of the backbone is still to be broadband when
    possible, geographically appropriate and if possible a station that
    delivers data in real time."


    In the spirit of the discussions and with an interest in soliciting
    ideas from other Networks, I have reviewed the current contribution
    of the GSN to the FDSN Backbone. In the attached spreadsheet
    (information is from the FDSN station list), I submit 27 stations
    for consideration within an expanded FDSN Backbone. All of the sites
    have or will have open real-time data.

    The USGS has 5 of 9 new GSN sites in the Caribbean are completed,
    all will have open, real-time data. It may be of interest to
    consider these sites for the Backbone.

    There are a number of additional GSN sites in the US which are not
    proposed.

    I am receptive to your perspective and comment.

    Best regards,

    Rhett

    <for FDSN Backbone.xls>_______________________________________________
    fdsn-wg1-stations mailing list
    fdsn-wg1-stations<at>iris.washington.edu
    http://www.iris.washington.edu/mailman/listinfo/fdsn-wg1-stations





    • Winfried Hanka
      2008-07-09 16:43:31
      Dear Seiji, Rhett and all,

      before submitting a concrete proposal for new backbone stations from our
      side I would propose to better define the criteria for the selection. We
      are installing about 35 new stations around the Indian Ocean (half of
      them existing, the rest to be installed until 2010) and I could surely
      contribute some of those for the backbone network. But how many more
      stations we need e.g. in Indonesia, Sri Lanka, Madagascar and Kenya? We
      (will) have two stations each in Sri Lanka, the Maldives, Yemen (a 3rd
      one on Socotra), Tanzania, Kenya and Madagascar. But how many to select
      for the backbone? On Rhett's list most cases are obvious ones. But e.g.
      for another station in Germany there no justification from poor
      coverage. So to me the criteria are not totally clear yet and before
      submitting a proposal I would like to have this clarified. Also if we
      select now only among existing stations or also future ones, and if yes,
      how far in future?

      Regards,

      Winfried

      • Lyons, Jim
        2008-07-09 16:02:41
        Dear All,

        I agree with Winfried that the best place to start is to clearly
        enunciate the overall purpose of the FDSN backbone network. Is it to be
        a one-stop-shop for global seismology? All of these FDSN stations are
        freebies, in the sense that none are installed or maintained at FDSN
        expense for pure FDSN purposes. They are typically used for
        national/regional earthquake monitoring +/- CTBT verification.

        A similar question arises with the insistence on real-time access. In
        the body of Rhett's e-mail of July 7, I believe Paul Wiejacz observed
        that most FDSN stations are already real time within some network. Why
        should FDSN get into the business that someone else is already doing?

        I am not trying to be unduly negative here. I understand the credo of
        network operators everywhere that more and faster is better. But I
        would like to work towards an explicit understanding of what it is we
        are trying to achieve with the FDSN backbone.

        Best regards,
        Jim


        -----Original Message-----
        From: fdsn-wg1-stations-bounces<at>iris.washington.edu
        [fdsn-wg1-stations-bounces<at>iris.washington.edu] On Behalf Of
        Winfried Hanka
        Sent: July 9, 2008 03:44
        To: Seiji Tsuboi
        Cc: fdsn-wg1-stations<at>iris.washington.edu
        Subject: Re: [fdsn-wg1-stations] FDSN Backbone

        Dear Seiji, Rhett and all,

        before submitting a concrete proposal for new backbone stations from our
        side I would propose to better define the criteria for the selection. We
        are installing about 35 new stations around the Indian Ocean (half of
        them existing, the rest to be installed until 2010) and I could surely
        contribute some of those for the backbone network. But how many more
        stations we need e.g. in Indonesia, Sri Lanka, Madagascar and Kenya? We
        (will) have two stations each in Sri Lanka, the Maldives, Yemen (a 3rd
        one on Socotra), Tanzania, Kenya and Madagascar. But how many to select
        for the backbone? On Rhett's list most cases are obvious ones. But e.g.
        for another station in Germany there no justification from poor
        coverage. So to me the criteria are not totally clear yet and before
        submitting a proposal I would like to have this clarified. Also if we
        select now only among existing stations or also future ones, and if yes,
        how far in future?

        Regards,

        Winfried
        _______________________________________________
        fdsn-wg1-stations mailing list
        fdsn-wg1-stations<at>iris.washington.edu
        http://www.iris.washington.edu/mailman/listinfo/fdsn-wg1-stations


      • Rhett Butler
        2008-07-16 01:08:40
        Dear Seiji, Winfried, Jim, Pawel, and others in WG1

        Here are some of my thoughts on the discussions to date. I will follow
        with some suggestions for increasing the FDSN Backbone.

        Rhett


        On Jul 8, 2008, at 9:43 PM, Winfried Hanka wrote:

        Dear Seiji, Rhett and all,

        before submitting a concrete proposal for new backbone stations from
        our side I would propose to better define the criteria for the
        selection. We are installing about 35 new stations around the Indian
        Ocean (half of them existing, the rest to be installed until 2010)
        and I could surely contribute some of those for the backbone
        network. But how many more stations we need e.g. in Indonesia, Sri
        Lanka, Madagascar and Kenya? We (will) have two stations each in Sri
        Lanka, the Maldives, Yemen (a 3rd one on Socotra), Tanzania, Kenya
        and Madagascar. But how many to select for the backbone? On Rhett's
        list most cases are obvious ones. But e.g. for another station in
        Germany there no justification from poor coverage.

        Agreed. On the basis of coverage, it is redundant. Nonetheless, BFO is
        an exceptionally quiet site.

        As a sidebar, BFO is an Affiliate GSN station. Although data are
        distributed by GSN, the station is independent.


        So to me the criteria are not totally clear yet and before
        submitting a proposal I would like to have this clarified. Also if
        we select now only among existing stations or also future ones, and
        if yes, how far in future?

        Originally when the "FDSN network" was initiated, most of the planned
        stations were relatively far in the future. The criteria seemed to be
        when there was actual funding. By this old criteria, potentially all
        of the new trans-Indian-Ocean Geofon stations should be considered.



        Regards,

        Winfried


        On Jul 9, 2008, at 3:02 AM, Lyons, Jim wrote:
        Dear All,

        I agree with Winfried that the best place to start is to clearly
        enunciate the overall purpose of the FDSN backbone network. Is it
        to be
        a one-stop-shop for global seismology? All of these FDSN stations are
        freebies, in the sense that none are installed or maintained at FDSN
        expense for pure FDSN purposes. They are typically used for
        national/regional earthquake monitoring +/- CTBT verification.

        For the GSN, the primary funding purpose was/is science (at least by
        US National Science Foundation's measure). Yes, there are now many
        other useful purposes for the data. And, our USGS partners are funded
        for earthquake hazard response.

        It is a great success of FDSN that its members are willing to sharing
        their data freely.




        A similar question arises with the insistence on real-time access.

        I do not see an insistence upon real-time. There are still existing
        Backbone stations without real-time access. Nonetheless, real-time is
        a stated FDSN goal, i.e, "Improving Access to data in real time." If
        an FDSN network wishes to make its data available to all in real-time,
        why not embrace it?


        In
        the body of Rhett's e-mail of July 7, I believe Paul Wiejacz observed
        that most FDSN stations are already real time within some network.
        Why
        should FDSN get into the business that someone else is already doing?

        I agree that the "real-time" component is up to the FDSN member
        network. FDSN simply encourages the sharing of data, whether real-time
        or not.



        I am not trying to be unduly negative here. I understand the credo of
        network operators everywhere that more and faster is better. But I
        would like to work towards an explicit understanding of what it is we
        are trying to achieve with the FDSN backbone.

        Best regards,
        Jim




        On Jul 9, 2008, at 11:34 PM, Pawel Wiejacz wrote:
        Dear All,

        I also agree with Winfried and Jim.

        The issue seems to me more conceptual than purely scientific. FDSN
        as it is consists of a growing number of stations and with so
        many stations it is no longer as manageable as it was some time ago.
        E.g. if data stops flowing in from some particular station, it is no
        longer feasible to look after this and inquire why - especially in
        places like Europe where FDSN stations are numerous and one
        can find another station from an alternate location fairly nearby.

        In this sense, redundancy is a good thing.



        So - it appears to me - there is this desire to have some stations
        designated FDSN-backbone, so that these stations make a
        more-or-less uniform global seismic network with some assured
        level of data availibility (open to discussion if this availibility
        level
        is to be 95, 98 or 99%).

        There has never been a data-availability standard for FDSN. Perhaps
        the best FDSN can do is to have sufficient redundancy to have good
        "effective" coverage.



        This matter is not easy. On one hand I do not see a reason why FDSN
        should not accept another station that is broadband, low
        noise and offering free data.

        In principle, many networks offer up all of their stations, and some
        share only a few (or perhaps none at the moment). Every network is
        encouraged to nominate/offer at least one station to the FDSN backbone.

        On the other hand I can imagine someone putting up a dense broadband
        network on a small area
        and offering all its data (e.g. 100 stations) to the FDSN.

        This already occurs. The USArray transportable array of 400+ stations
        is entirely open to FDSN in real-time. However, there is no intent to
        make this a part of the Backbone.

        It is not the point so that FDSN becomes a free data repository to
        just anyone (who e.g. might wish to save money on disk array - keep
        the data just on his PC and have a backup copy at the FDSN
        for the just-in-case). The line must be drawn somewhere.

        Agreed. The FDSN Backbone data are actually archived and open to the
        FDSN. Currently the FDSN archive for the Backbone resides at the IRIS
        Data Management System. However, there are a number of other FDSN data
        centers, which offer additional/parallel/complementary FDSN data.

        A number of FDSN members make their own bilateral arrangements to
        archive data at various FDSN archives.

        Given our current discussion toward increasing the size of the
        Backbone (by a some small factor), the impact of additional stations
        is still small.



        It should be the FDSN leaders to decide on this, but to my opinion
        the FDSN-backbone network should be much denser
        than the CTBTOs. Or the interest in the FDSN-gathered data
        will be limited.

        Currently, CTBTO is not open, and is available only to States Parties.
        The CTBTO IMS is composed of about 170 primary (50) + auxiliary
        stations (120) + a few "T-phase" stations on islands. The FDSN
        backbone is already more dense at 200+ stations. Since the CTBTO IMS
        is fixed by the Treaty, our discussion to increase the size of the
        FDSN Backbone will make it relative more dense than CTBTO.

        In principle, it would be nice someday to incorporate the CTBTO into
        the FDSN.


        Of course one must keep in mind that the CTBTO network is
        cunning - it has been constructed with maximum possible station
        density that is still more-or-less uniform throughout the world.

        CTBTO uses a lot of arrays for its primary network, and relies heavily
        upon FDSN stations for its auxiliary network.


        So should FDSN go to a greater station density, then naturally
        there would be regions where getting a greater station density
        than CTBTO shall not be possible. Moreover, FDSN is unlikely
        to surpass CTBTOs resolution in areas like Russia or China where
        much of the data is restricted; according to the CTBTO treaty the
        data is ported to the ISC, but digital data gets then restricted
        there at Vienna.

        Although there are some increasing data restrictions, data in Russia
        and China are still available through respective joint efforts in
        collaboration with GSN.



        It is a separate question if FDSN should archive all the data or
        perhaps use some distributed-archive scheme. But with distributed
        archive, the number of data contributors, and lack of funding for
        specifically running the distributed-archive at the contributors'
        sites - managing a distributed archive may be a problem.

        The FDSN does have utilize distributed data centers with growing
        success. However, each of these data centers are funded more-or-less
        for their own purposes, for which we are fortunate that these purposes
        include international data sharing.



        Regards,
        Pawel