Executive Committee Approves Change

Dmitry Storchak (at left) and Ray Willemann (at right) met with the Executive Committee members (left to right) Oleg Starovoit, Chris Browitt, Gary Gibson, Adam Dziewonski, John Woodhouse and Akio Yoshida at the ISC on 6 - 7 June.

The ISC Executive Committee has given the go ahead for the most important change in ISC location procedures in more than 30 years.

Last year the Governing Council saw results from tests using S arrival times, which improved many ISC hypocentres. But depths of lower crustal events were too deep, so the Council required that the Executive Committee review further tests before ISC procedures were changed. Richard Luckett has since corrected an S phase identification error in ISC's software, and this year he demonstrated to the Committee that ISC depths using S no longer differ systematically from reported depths. The ISC will use S times to compute hypocentres for events from January 2001.

The Committee is pleased that the CTBT Preparatory Commission decided to exchange seismic data with the ISC, but disappointed that no provision was made for events after 2001. The Commission will meet only a few times before the ISC begins its final analysis of 2002. Data exchange must be given priority if consensus is required on a model agreement for exchanging all types of treaty monitoring data before exchanging further data with the ISC.

Committee members urged ISC staff to seek agreements with programme operators so that measurements from processing of even temporarily deployed networks are included in the ISC database. They endorsed discussions with the OBS-IP programme and with several PASSCAL investigators. But they were concerned that standardised automatic scanning for arrivals in data from PASSCAL and other programmes may be implemented without arranging to contribute results to the ISC routinely.

Staff members at the meeting described plans for ISC to co-ordinate collection and distribution of measurements from additional analysis of selected events. The Committee discussed the need to minimise additional costs, partly by starting with a small number of contributors. They also noted that the ISC database schema might need to be amended to accommodate new data types that could be contributed, such as relative arrival times from waveform cross-correlation.

The Committee decided that a gradual reduction in the time interval between the occurrence of events and release of the ISC Bulletin should continue. The Governing Council decided last year to start cutting the interval from the two-year limit in the ISC Working Statues, and it is being reduced by about one month per year. The Committee heard that no critical data have been missed out even though some agencies barely meet the advancing schedule.

The Committee endorsed a plan to improve the next set of ISC CDs using software contributed by the British Geological Survey (BGS). The BGS is developing a Graphical User Interface to select events from data files on the CD and display them in maps and other figures. The program runs only under Microsoft Windows, but in a survey of ISC users last year most respondents who sought a graphical interface used Windows. Nevertheless, the Committee encourages other institutional members to contact the ISC staff about developing a Unix X-Windows GUI or making other in-kind contributions.

The Committee was pleased the ISC has started work on a stand-alone location program and encouraged the staff to be sure that the program can be adapted readily to using data in different formats and that it is straightforward to link in an alternative Fortran location routines. The program should be developed consistently with the objectives in ISC's proposal last year to NSF regarding services for users and eventually to support on-line editing by seismologists at the ISC.

The Committee noted development of new techniques for identifying depth phases by combining waveform data from different stations. Sums of functions computed from automated triggers may be especially amenable to use at the ISC, since individual network operators could be asked to run detectors on data from their own network and report coda triggers to the ISC. ISC staff members plan to start work with several networks, and hope to present test results to Governing Council next year.

Beach Balls in the Bulletin

Thanks to a program contributed by Governing Council member Cliff Frohlich of the University of Texas, the ISC On-Line Bulletin now includes the familiar "beach ball" to represent each focal mechanism posted to users. The ISC collects focal mechanisms from more sources each year, now including regional catalogues such as those of JMA, MedNet and PNSN, as well global catalogues such as those of Harvard and NEIC.

The ISC already made it easier to use mechanisms from disparate sources together by grouping those that represent the same earthquake and rewriting from each native format to IASPEI Seismic Format (ISF). As seismologists know well, however, it is still difficult to visualise focal mechanisms by scanning lists of numbers, and comparison is even more challenging if some mechanisms are expressed as moment tensor components while others are expressed as principal axes or nodal planes.

Beach balls, which are included in the On-Line Bulletin if the "include links" option is checked, aid visualisation by graphically representing nodal planes and principal axes, regardless of how each focal mechanism is expressed. If sufficient data have been contributed to the ISC then P and T axes are marked and zones of compressional first motion are shaded. The most refined version of the figures, with non-planar nodal surfaces, is plotted if the contributed data are sufficient to determine the CLVD component of a moment tensor.

More information about focal mechanisms and beach ball figures is available from the ISC web site at http://www.gxjydl.com/doc/misc/b-balls.html

Amplitudes Corrected

Thanks are due to Tamara Jesenko of the Seismological Observatory of Ljubljana, Slovenia for spotting erroneous amplitudes in the ISC's On-line Bulletin.

For phase arrivals during January 1999 to August 2000, the ISC mistakenly interpreted amplitudes reported by the Ljubljana Observatory on broadband channels as micrometers rather than nanometers. The misinterpretation affected a total of 4422 amplitudes at stations BISS, CESS, CEY, DOBS, LJU and VBY. A further 1333 amplitudes from short period long period, and unspecified channels at the same stations during those months were not affected.

The broadband amplitudes were not used by the ISC in computing a magnitude, and so were not included in the printed Bulletin. But the incorrect broadband amplitudes for these stations are in data files on the CD of 1999 earthquakes, and were on the web site and in AutoDRM e-mails from the ISC until 28 May 2002.

By default the ISC web site and AutoDRM e-mail provides users with the "comprehensive bulletin", which includes changes after publication such as this correction of amplitudes. Users who wish to can override the default to request the "published bulletin", which provides data from the bulletin without corrections and without data added after publication. FDSN channel codes distinguish between broadband and other amplitudes. Data with channel codes can be retrieved from the ISC web site by selecting phases at particular stations in user-specified time intervals from www.gxjydl.com/search/arrivals/.

EHB Bulletin Posted

The Engdahl, van der Hilst and Buland (EHB) bulletin has been completely integrated into the ISC database. With the completion of this work, the results from EHB's re-processing of data from the ISC Bulletin are now available from the ISC web and AutoDRM servers.

EHB re-determined the identities of over 5 million arrivals, including many previously un-recognised depth phases. The new phase identifications are based on ak135 travel times, geographic variations in ocean depth, and procedures to reduce biases. Partly thanks to these identifications, EHB's hypocentres for the largest events from the ISC Bulletin may be better for tomography and some other purposes than those originally computed by the ISC.

Users interested exclusively in events re-processed by EHB can use the ISC's standard web page for earthquake data, http://www.gxjydl.com/search/bulletin/, and select "EHB" in the "agency" pull-down menu. To retrieve EHB's identifications of phases associated with these events, users should additionally select "collected" in the "bulletin type" pull-down menu. Users with general interest in special sets of data collected by the ISC, such as the EHB bulletin, may find it more convenient to us links in http://www.gxjydl.com/doc/misc/special.html rather than starting from the blank bulletin selection page.

Standardising Phase Names

Accurate identification of phase arrivals is one of the most important features of both the ISC Bulletin and regional bulletins on which the ISC's work is based. To ensure that identifications are as reliable as possible, the ISC keeps three identifications for each phase: the name just as reported, an interpretation of the reported name in ISC's nomenclature, and ISC's own identification based on the depth of the associated origin and its distance from the station.

Standardising phase names among seismologists would reduce the chance of mis-interpretation and ease reliable exchange of bulletins. ISC's senior seismologist Dmitry Storchak took a big step towards this goal by leading an IASPEI working group on phase names. Following a public comment period, the working group finalised its consensus list of names, which is now posted on the ISC website.

Most likely, this list will be endorsed by IASPEI's Commission on Seismic Observation and Interpretation at its meeting next year. Then the real work will begin as the group works to persuade agencies around the world to adopt the standardised names in their own bulletins. This will be easier, of course, if the group can convince authors of widely used programs such as SeisAn, Antelope and SAC to modify the programs to use names from their list.

Threshold Adjusted

ISC seismologists did not review of most 3.0-3.4 JMA events during June, July and August 2000. This is a departure from standard procedures at the ISC, which include reviewing all events with a reported magnitude of 3 or more, with a reported arrival at a station more than 5º from the epicentre, or with reported data from two or more agencies. Each event that met any other criterion was reviewed no matter how small its magnitude.

The reason for temporarily adjusting the threshold with regard to JMA magnitudes was a prolific swarm of activity near Honshu, which included large earthquakes on 6 June and 30 July 2002. Analysis of the smaller events by ISC seismologists would have delayed production of the Bulletin and added little to JMA's analysis of events that were not recorded beyond Japan.