Progress Report: March 18, 2002
- Working Group on Magnitudes
- Working Group on Standard Phase Names
- Working Group on Reference Events
- Working Group on Event Location Using Multiple Events
Members: Peter Bormann, Jim Dewey, Soren Gregersen, Won-Young Kim, Klaus Klinge, Bruce Presgrave, Bob Uhrhammer, Karl Veith.
Chairman: Jim Dewey
I invited several people who were unable to participate. The group as it stands may lack some expertise or geographic representation that will eventually be critical to our success. In particular, I am bothered that we have, I believe, only one member (Karl) who has extensive knowledge of the perspectives of the nuclear-test-ban community, and no one from the IDC. I still have two invitations outstanding that, if accepted, would remedy these deficiencies. The WG also does not contain people working outside of the U.S. and Europe. I believe we can add members as we identify needed additional perspectives.
On the basis of comments I have received thus far, I would like to propose the following goals and timetable for the group. I have adopted the point of view suggested by several that the "standard" procedures for measuring amplitudes and periods should be those that replicate as much as possible the amplitudes and periods produced by widely-used mid-twentieth century seismographs.
A. Overall purposes of magnitude WG
The main purpose is to provide standards for making measurements from digital data to be used in calculating earthquake magnitudes. We will give highest priority to magnitude-types that appear frequently in current international bulletins produced by the NEIC and ISC or that are being produced by regional networks of digital seismographs: these magnitudes are those commonly denoted mb, MS, ML, MD, mbLg, and MW (from local and regional data). We may include more than one "standard" procedure for some general kinds of magnitude. We will document widely-used measurement procedures that we do not label as "standard."
B. Goals for late spring
- Agree on how we will abbreviate the different magnitude types. I am sure I saw a list of abbreviations proposed by Peter that would be the logical starting point, but I haven't been able to find it again in looking through my files (I was trying to find his rationale for naming the phase that I am calling "mb").
- Obtain a preliminary sense of the percentages of amplitudes contributed by different types of seismographs to the NEIC and ISC at different times in the past.
- Agree on the qualitative nature of the filters, time-windows of measurement, etc. for the "standard" processing of digital data, so that these can be accounted for in a qualitative way in the Manual of Seismological Observatory Practice (MSOP)
- Review the magnitude section of the MSOP.
- Identify critical perspectives that are missing from the WG and expand the WG to include those perspectives.
C. Goals for the Sapporo IUGG/IASPEI meeting (June 2003)
- Develop recipes for each of the standards -- poles and zeroes of filters, length of time-windows of measurement,
- Agree on recommendations to be made to the IASPEI community-at-large on how the standards should be applied.
- As we encounter "non-standard" procedures that are not already covered in the MSOP, add these to the MSOP.
D. Desired seismological outcomes
—that depend in part on how measurements are made from digital data (abstracted from comments submitted by WG members). Most of these outcomes would not be the direct responsibility of the magnitude WG.
- Estimation of internally consistent magnitudes from data collected with seismic instruments and recording methods that have been used over the past century.
- Estimation of moment magnitudes from regional networks, in the midst of the meizoseismal regions of strong earthquakes, that are consistent with globally determined moment magnitudes.
- Accurate estimation of the yield of well-coupled underground explosions.
- Rapid estimation of the destructive potential of a recently-occurred earthquake
- Estimates of earthquake energy, source-dimensions, and stress-drop whose bias does not vary with the decade of the earthquake's occurrence.
- Estimates of earthquake energy, source-dimensions, and stress-drop whose bias does not vary with the geographic location of the regional network that assigns the magnitude.
- Magnitude-frequency relations that can be extrapolated beyond the magnitude range of the observed earthquakes.
E. Inelegant realities
—seen from the perspective of the NEIC, that may affect how we go about our task
- There has never been a time when NEIC and ISC mb magnitudes were computed only with WWSSN SP data using an agreed-upon time-window.
- Consider the filter that is used on digital data at the NEIC to synthesize short-period records from which P-wave amplitudes and periods are measured. Ray Buland, who developed the filter, has told me that he developed the filter with an intent to produce short-period mb's that were consistent with the mb's being produced from other NEIC data. The filter was not explicitly constructed to be identical with the WWSSN-SP response. I will get from Ray the poles and zeroes of both SP and LP NEIC filters -- I don't have them now.
- Given the choice of "catalogued earthquake with no associated magnitude", and "catalogued earthquake with an associated magnitude that is biased", many of those who use the catalogs of the NEIC and the ISC would prefer "catalogued earthquake with an associated magnitude that is biased."
F. My guess as to the biggest source of internal WG controversy
—based on comments already circulated by early WG members.
Agreeing on how the standards should be applied by the NEIC and ISC. There are "hard" approaches: e.g., the NEIC/ISC should use only certified standard measurements. There are "soft" approaches: e.g., the NEIC/ISC will encourage measurements to be made in the standard ways but will publish all contributed data and document the measurement practices of the stations for which they can obtain this information. Both hard and soft approaches can be defended, depending on the use that is to be made of the magnitude.
G. Next steps
(can we aim to have these by mid-March?)
- Peter—Provide proposed nomenclature for magnitude types (I am making this request of him because I am quite sure that I saw such a nomemclature prepared already, and I am hoping that this won't be too much of a burden in a busy time for him).
- Peter—Provide electronic copy of current draft of MSOP (Peter tells me that the draft that is currently on the IASPEI is out-of-date, and he would like WG members to review the current version)
- Jim and Bruce—Provide response characteristics of filters currently used at the NEIC to produce short-period and long-period records from digital data.
- Jim and Bruce—Provide an approximate break-down of the sources of magnitudes and amplitude/period data used in the NEIC.
- All—Look through past e-mails sent to only a few of the WG and send to other members of the WG those e-mails that you think everyone should have.
- All—Provide feedback on items A., B., and C., above.
Members: B.Kennet, B.Engdahl, J.Havskov, J.Schweitzer, P.Bormann, D.Storchak
Chairman: Dmitry Storchak
The Working Group was set up 6 months ago in September 2001.
We have since:
- Discussed the general principles and a structure of the list and produced a First draft Version of the phase list, which was put together by the chairman with a large input from Johannes Schweitzer and Robin Adams. - On the November, 12 2001 the First Version was distributed among WG members. The corresponding comments, corrections, suggestions were eventually received from all members of the WG. This was a success bearing in mind numerous other responsibilities of the WG members.
- Based on these comments, the Second Version was produced and again discussed in details in a very small group.
- On the February, 8 2002 the Second Version was distributed among the WG members as well as selected colleagues in large data centres and other colleagues whose response we thought could be useful for the project.
- We have received several replies so far, but a lot more is expected. Corresponding reminders are being sent round.
We have originally planned to finish the job by 1 April 2000. At this point it becomes apparent, that the list is not going to be produced as planned. We believe it is not worth producing the list, which was not generally approved by a reasonable number of colleagues. We shall nevertheless still aim for it's completion two months after we got the desired response on the Second Draft Version.
Based on a considerable input to the work of the Group, I have invited Robin Adams to become a member of the WG. He has agreed
Members: E.A. Bergman, Bob Engdahl, Bondar, P. Firbas
Chairman: Bob Engdahl
The IASPEI reference event list has been carefully vetted and a new release is in preparation.
There has been some preliminary work on GT selection criteria for events with reporting stations at regional and teleseismic distances. We have assembled and vetted a database of about 1800 explosions and earthquakes globally whose locations are known to 5 km or better (GT5). These events have been relocated at fixed depth using stations at all distances with the EHB algorithm and the mislocations in km determined. These mislocations were compared to the secondary azimuth gap (the largest azimuth gap filled by a single station) both for explosions whose locations are known to 2 km and for explosions and earthquakes. In both cases, there are marked improvements in the percentage of events mislocated by 15 km or less (GT15) at secondary azimuth gaps of 180 and 130 degrees. More about this in the next report after we have performed further testing.
The IDC has informed us that reference events submitted through their calibration program will be released after certification.
Members: E. R. Engdahl, Paul G. Richards, Ray Willemann
Chairman: Paul G. Richards
Our work since the previous report of September 19, 2001, has been to acquire datasets of two types: first, sets of ground truth events (origin time, epicentre, depth, magnitude); second, sets of phase picks (regional, teleseismic). The purpose of such data acquisition, is to permit later evaluations of the performance of different algorithms for locating multiple events.
Thus, with help from Istvan Bondar we have acquired ground truth information for a set of mine-induced earthquakes at Lubin, Poland. For these events, we have also obtained the associated phase picks at IMS stations. Most of the picks are for regional waves recorded in Europe, but there are also some teleseismic picks.
Felix Waldhauser and Paul Richards at Lamont have begun joint efforts with two seismologists in China to study clusters of seismicity in three different regions of China (located with and without methods of multiple event location), and to relate the relocated events to fault structures.
Eric Bergman and Bob Engdahl at the University of Colorado have begun a similar effort with other seismologists in China to study clusters of seismicity in China using multiple event location techniques (HDC).
I anticipate that we shall see further progress in meetings held during a workshop being organized by NORSAR April 22-26 2002, which will attract several researchers working on multiple event location.