Proposal for the Compilation of a CNSS Unified Earthquake Catalog David Oppenheimer, Lind Gee, Steve Malone, Doug Neuhauser Thanks to the efforts of the members of the CNSS, the "finger quake" utility now provides easy access to recent earthquake activity occurring within an individual network. However, the finger-quake utility is, by design, limited in its ability. Many networks report only their real-time locations via finger-quake and do not update the finger- quake locations when the event is reprocessed by analysts. The finger-quake utility also has a limited memory, such that there is no common method for obtaining older data from the same networks. Though most networks make their catalogs available to the public via anonymous ftp, World Wide Web (WWW) interfaces, guest logins, and dedicated accounts, it is still a challenge to assemble a uniform catalog for any region spanning more than one network. Hypocenter formats differ from one network to the next, the type and uniformity of the parametric data, and search tools are not always available. Moreover, it is difficult to identify earthquakes that are present in two catalogs where regional networks overlap, or recognize the absence of an event that occurs between networks. We propose a simple solution to this situation in which members of the CNSS participate in the creation of a joint worldwide catalog available to the public via a WWW browser. Here's the scheme: 1. Member networks except NEIC are assigned a polygonal region inside of which they believe they have authoritative coverage. The polygons available from the CNSS homepage offer a first cut at the authoritative regions (http://www.geophys.washington.edu/CNSS/polys.txt). 2. At whatever frequency member networks wish (hourly, daily, random), they transmit to some central site a monthly file with agreed-upon format (see Appendix) for all earthquakes recorded by their network. Except for NEIC (see #3 below), events within their polygon are designated authoritative, but networks can still report on earthquakes outside their network. The filename describes the time period and network using the FDSN network code (e.g., 1995.01.CI.catalog). Any month could be placed in the directory--not just current data. This way networks can update their real-time information at a later date. Only one magnitude is allowed, so networks must choose which magnitude they believe best represents the size of the earthquake. Networks also decide the quality and magnitude threshold of the data they wish to report. - 2 - 3. NEIC transmits its entire catalog, but their solutions prevail only for regions outside of the authoritative polygons where no other information exists. NEIC would also choose only one magnitude from Mw, Mb, and Ms. In addition, the NEIC data set also provides their global locations. By extension, regional networks around the world could also participate in the generation of this unified catalog, with NEIC filling in the regions with no coverage. 4. At the central site, a program wakes up periodically to see if any new data have arrived. If so, it replaces any existing network-monthly file with the new one and chronologically sorts all of the available network catalogs for that month to create the unified monthly catalog. The software attempts to reconize duplicate events and selects only the authoritative event based on the network/epicenter/polygon relationship. 5. When the unified monthly catalog is updated, it automatically could be copied to other data centers who would like to have it. We propose that this scheme be implemented at the Northern California Earthquake Data Center (NCEDC) at UC Berkeley because Doug Neuhauser has already written software that manages catalog updates as described above, and he has also written a WWW browser interface that can search an earthquake catalog (http://www.quake.geo.berkeley.edu/ncedc/anon/catalog- search.html). To help members of the CNSS implement this procedure, we could distribute software that reformats the data into the common format, selects the events according to the desired criteria, and moves the data to the NCEDC. The browser software could also be revised to support other output formats if the current output formats ("readable", hypo71, hypoinverse, UCB catalog format) do not meet the needs of the contributors. If you would like to comment on this proposal, please send your ideas to David Oppenheimer (oppen@alum.wr.usgs.gov) or Lind Gee (lind@seismo.berkeley.edu). Appendix: Strawman Format for Unified Catalog *4d Year of event *2d Month of event *2d Day of event *2d Hour of event *2d Minute of event *7.4f Second of event *8.4f Latitude of event (decimal degrees, N is positive) *8.4f Longitude of event (decimal degrees, E is positive) - 3 - *8.4f Depth of event (kilometers) 3d Number of observations used to calculate the hypocenter *4.2f Magnitude *c Magnitude type: S=surface, B=body, W=moment, C=coda, L=Richter, etc. 3d Number of observations used to calculate magnitude 3d Maximum azimuthal gap (degrees) 7.4f Distance to nearest station (kilometers) 7.4f Error in origin time (seconds) 7.4f Error in horizontal (kilometers) 7.4f Error in depth (kilometers) 7.4f RMS residual of phases used in location 7.4f Error in magnitude *c Quality of event hypocenter (uniformly defined) *c Event flag: blank=earthquake, Q=quarry/reflection/refraction blast, N=nuke *2c FDSN network code *Required field.