July - September 2003

New Member Elected to Executive Committee

Akio Yoshida of JMA completed his six-year term on the Executive Committee with his participation in the meeting in Sapporo in June. The Governing Council elected Akio when it met in Thessaloniki, Greece during the 1997 IASPEI Assembly, and brought the perspective of a large and very busy national monitoring agency.

To succeed Akio, the Governing Council elected Jaroslava Plomerova of the Czech Academy of Sciences. The Czech Academy includes both contributors of data to the ISC, primarily from the Czech National Network, and users of ISC data, often for lithospheric structure studies. Jarka has represented the Czech Academy on the Governing Council since 1996, where she gently yet persuasively advocates changes that improve ISC operations. Partly in response to new services from the ISC, in 2001 Jarka helped to arrange for the Czech Academy to double the annual funding that it contributes to the ISC.

As an expert in deep lithospheric structure, Jarka has worked with the University of Leeds in its development of a system to automatically make SKS splitting measurements that will be included in the Bulletin. Collaboration has extended to hosting Leeds Ph.D. student Matt Evans for an extended visit in Prague and resulted in a joint poster at the IUGG 2003 Assembly.

Jarka Plomerova, at the Czech Academy of Sciences, Prague.

Jarka will be involved in telephone and e-mail exchanges on ISC developments over the coming months. For the Committee's next meeting at the ISC in June 2004, she plans join the other Committee members elected in previous years by the Council, Chris Browitt (Royal Society, 1999) and Gary Gibson (Seismic Research Corp., 2001), and the ex officio members, Adam Dziewonski (Governing Council chairman), John Woodhouse (Oxford University) and Oleg Starovoit (IASPEI).

Director Speaks in Vienna

Director Ray Willemann gave an invited talk in September at a seminar associated with a conference in Vienna, Austria on the Comprehensive Nuclear Test Ban Treaty (CTBT). The seminar was organized by Vertic, a non-governmental organisation that monitors verification issues related to several treaties, including the CTBT.

Speaking on "The role of non-IMS networks in explosion monitoring", Ray pointed out that seismic stations installed to monitor earthquakes greatly outnumber seismic stations of the CTBT system in almost all land areas. With current techniques, he said, data from the CTBT monitoring network and these other stations together can be used to obtain results that are more accurate or reliable than results with data of the CTBT monitoring system alone.

But Ray went further, and suggested that continuing, rapid growth in the number of broadband stations deployed for earthquake monitoring means that some advantages from using non-IMS are likely to grow, despite improvements in methods for using the data. Figures and text from Ray's talk are posted on the ISC web site.

German Task
Force Catalogue

Dr Helmut Grosser of the German Task Force for Earthquakes at GFZ, Potsdam, has contributed the Task Force's catalogue of aftershocks of the 1992 Erzincan, Turkey earthquake to the ISC.

The Task Force co-ordinates rapid responses to earthquake disasters around the world, despatching complementary teams of geoscientists, engineers, and relief workers. The teams deliver prompt results to local authorities to help minimise further damage, and carry out more detailed analyses later to improve responses to future disasters. Erzincan is one of eleven Task Force deployments since 1992, and further aftershock catalogues may be contributed to the ISC in the future.

The Erzincan mainshock was a strike-slip rupture of the North Anatolian fault at the northern margin of the Erzincan basin, causing considerable damage in Erzincan and the surrounding villages. In the Erzincan basin and the nearby mountains the Task Force operated a network of ten seismic stations for 3 months. They recorded more than 3,000 aftershocks, of which they located 505. The catalogue that the Dr Grosser contributed to the ISC includes scalar seismic moment, source radius, stress drop and average slip of 394 aftershocks and the fault plane solutions of 53 aftershocks.

Some of the aftershocks are concentrated near the mainshock epicentre along the North Anatolian Fault and have strike-slip mechanisms, but others are in a tight cluster far from the North Anatolian fault and have predominantly normal-faulting mechanisms. Complete details about he project were published in Grosser et al., Pure Appl. Geophys., 152, 456-505 (1998). The aftershocks can be retrieved as part of the ISC on-line Bulletin, or the complete catalogue can be retrieved on its own from the ISC web site starting from http://www.gxjydl.com/cgi-bin/special_list.

Oxford Summer Student

Oxford University student Katherine Evans has concluded a summer appointment at the ISC. Thanks to her work, the ISC database now includes all of the moment tensors that have been published by the NEIC. NEIC has been contributing all newly-computed parameters for several years, but some moment tensors from past years had been missing at the ISC.

At the ISC, Kat learned to use Unix scripting tools to retrieve files automatically from web sites, parse data from the files, and insert the data in a database. After retrieving the NEIC moment tensors, she also updated the ISC's links to final moment tensor solutions on the "BATS" web site of Academia Sinica in Taipei. At her farewell party, Ray Willemann said that Kat learned on her own so quickly that it seemed as if the only thing she ever asked for was the password that she needed to insert her results in the ISC's main database.

Oxford University student Katherine Evans added to the ISC's data collection.

With moment tensors in hand from several sources for the same events, Kat prepared a summary of the differences between them. She found, for example, that NEIC and Harvard moment tensors are very similar to each other for intermediate depth and deep-focus earthquakes, but differ significantly for shallow earthquakes. What's more, Kat was able to show that the systematic change of the differences with depth are due entirely to the innate uncertainty for shallow earthquakes in moment tensor components related to shear traction parallel to the earth's free surface.

Kat has now returned to Oxford to study for an M.Sc. in seismology with Dr. Shamita Das, where she plans to put her new knowledge of both moment tensors and scripting to good use.

Project to Measure SKS
Splitting Nears Fruition

For the last year, Leeds University student Matt Evans has worked at the ISC implementing a system to automatically measure "splitting" of SKS and SKKS arrivals. Measurements made with the Leeds software are expected to be included in the on-line Bulletin in the near future.

As shear waves propagate up from the core they may be split into vertically and horizontally polarised arrivals with slightly different onset times. This splitting has been interpreted as evidence of anisotropy in some parts of the mantle by many seismologists, including Matt's Ph.D. advisor, Prof. Mike Kendall.

To measure the delay of one polarisation after the other, seismologists use a widely accepted numerical algorithm, so automating the calculation might seem to be "a doddle". But almost all splitting measurements to date have involved manually adjusting control parameters and subjectively selecting the "best" results, or even discarding measurements that seem unreliable.

Matt has shown that he can automatically select parameter values similar to those used by seismologists experienced in measuring splitting. For example, he picks onset time near the peak of the ratio of the short-term to long-term average (STA/LTA) of the waveform envelope. For SKS duration, Matt uses the longest window length that gives a numerically stable result.

Some of Matt's criteria for identifying noisy arrivals are based on his parameter adjustment algorithms. If the STA/LTA peak isn't large enough or occurs too far from the expected arrival time, then no result is recorded. In addition, Matt computes statistical uncertainty of the results using several methods that make different assumptions about noise that might be interfering with the signal, and flags results that are questionable according to any of the methods.

Splitting is measured in a window from before the picked onset to an end time when small variations to not effect results. The selected window automatically excludes other arrivals.

Mis-measurement can occur even in the absence of noise if the signal has unanticipated features. To address this possibility, Matt's software computes the linearity of particle motion in both the original data and the data corrected for splitting using his best estimate. A measurement is rejected if either the original particle motion is nearly linear, or if the particle motion remains non-linear after the correction for splitting.

Matt plans to talk about his project at the AGU meeting in San Francisco in December.

lSC Seismologist Publishes in PAGEOPH

ISC seismologist Nurcan Ozel is the first author of "Focal mechanisms of intermediate-depth earthquakes beneath southeastern Hokkaido, Japan: Implications of the double seismic zone" in Pure and Applied Geophysics. Her paper with Takeo Moriya is based partly on 100 new focal mechanisms that they computed from P first motions.

In a double seisimic zone extending from 60 km to 120 km deep, Nurcan and Takeo find that earthquakes in the lower plane have predominantly down-dip extensional mechanisms. But earthquakes in the upper plane have more variable mechanisms, including strike-slip and down-dip extensional mechanisms in addition to the expected down-dip compression. They conclude that unbending of the slab, the weight of the deep slab, and lateral variation in slab geometry all contribute to stresses.

Nurcan and Takeo have agreed to contribute their focal mechanisms to the ISC. Data from this and other research projects can be retrieved from links in the ISC's "Special Projects" web page (http://www.gxjydl.com/cgi-bin/special_list). Also, data from research projects are included in selections from the "Comprehensive" version of the ISC on-line Bulletin (http://www.gxjydl.com/search/bulletin/).

Figure 4 from Ozel & Moriya (2003, Pure Appl. Geophys., 160, 2279-2299). Focal mechanism solutions of the events benetth southeast Hokkaido. The hypocenters are projected onto a vertical plane striking N30°W-S30°E. The focal mechanism solutions are plotted on an equal-area lower hemisphere projection. The hatched and open areas refer to dilatational and compressional quadrants, respectively.