Modernizing the ISC location procedures: model evaluation (2007)
Comparison of the ISC Bulletin: AK135 and JB
We relocated the 10 most recent months of the Published ISC Bulletin, covering the period from January to October 2004, using the AK135 velocity model. Below we briefly describe the details necessary to understand the contents of this new bulletin and explain how to download the bulletin for further analysis.
Selection of events for relocation
There is no point comparing the results of ISC location of small events. Neither JB nor AK135 ISC solutions are likely to be as accurate as a solution of a well-established local network. There is also no point producing an AK135 location in the case where a JB solution is either not possible or where ISC editors have chosen not to publish an ISC solution. Based on this we have only considered the following events:
- Period of time: Jan-Oct 2004
- Strictly from Published (Reviewed) bulletin
- No data, received after the original analysis was completed, are to be used
- Only events with ISC JB estimates available
- Magnitude >= 3.5.
We used the ISC magnitude estimate if available. If not, we used an average of all reliable magnitude estimates reported for the event. If no magnitude was available, events with at least one station associated at 10 or more degrees were considered at least as large as 3.5.
Setting relocation options
At the ISC we try to get a free depth estimate where possible. Often, due to a lack of appropriate data, the ISC has to fix the depth to a particular value. This value is one of:
- depth reported by local network with close stations and good azimuthal coverage;
- depth based on pP-P time intervals;
- default depth for the area.
For the purpose of our comparison for each event we had to use the same way of dealing with depth for AK135 relocation as was used in JB location. Otherwise the same location procedures and the same set of determining phases (P, Pn, Pg, Pb, S[weight factor 0.5 beyond 20 degrees], Sn, Sg, Sb) were used for both JB and AK135 solutions.
A great deal of time was spent on tuning the relocation procedures, where an ISC seismologist was making iterative spot checks of the results of relocation. This was done only for two months of data in order to come up with the final relocation procedure. There are still issues with some secondary phase identification and network magnitude computation that are not related to relocation process and therefore were not yet addressed to our complete satisfaction.
It should be everyone's understanding that AK135 relocations were not examined by the ISC seismologists as thoroughly as JB ones. Nevertheless, we hope that the great effort originally made by the ISC seismologists in editing the JB bulletin benefited AK135 bulletin enormously.
The results of relocation are given in the ISF format. The agency code for the original JB solutions is ISC_JB. A corresponding code for AK135 solution is ISC_AK. Station travel time residuals are given with respect to the AK135 solution. Residuals for JB solutions are only available as part of the original bulletin. We offer both catalogue (no station arrivals) and bulletin (with station arrivals) versions.The new bulletin can be downloaded from the ISC FTP site in monthly files: