Why the ISC continues to search for `unreported' events

Alison L. Bird, D.A.Storchak, R.J.Willemann, R.D.Adams
(presented at ST5, AGM of IUGG, Birmingham, 1999)

Since 1974 the International Seismological Centre has searched its unassociated phase readings to find events not reported by other agencies. About 100 to 200 such events are found each month, and while the number of earthquakes published by the ISC has risen dramatically over the decades, the number of "new" events has remained steady. Many are genuine new events located in remote areas, such as oceanic ridges, where instrumental coverage by national agencies is poor, or situated near national boundaries, where no single country can obtain a solution with its own stations alone. One example of a new event was the first earthquake to be confirmed as occurring in continental Antarctica, in 1982. Many new events are already known to local agencies, but their origins are not reported to ISC because they are too small or are known to be of non-seismic origin, such as mining explosions. Also, the new events seldom have teleseismic amplitude and period readings for ISC to determine a magnitude, and locally determined values of ML are rarely available. Reviewing these events takes a considerable proportion of ISC's analysis effort. ISC plans to review the procedure to ensure that genuine new events, particularly in remote areas, continue to be discovered without devoting undue effort to rediscovering events already known. We report on regional variation in the proportion of new events not in local catalogues, particularly those requiring data from two or more networks.

Patterns of Search Events

(1974-1994)

In early years the distribution of Search events was widespread, with many seismic networks submitting only phase data and with the global location agency, Prototype International Data Centre, not yet reporting data to ISC. [N.B.: In all maps, symbol size is an indication of magnitude.]

Pacific Rim Events

(1995-1996)


The recently improved global location capability of agencies such as the Prototype International Data Center has greatly reduced the number of search events in remote areas. ISC is also more informed regarding controlled activity in mining areas. ISC continues to locate unreported earthquakes, in Hindu Kush, Myanmar, and around the Pacific Rim. Many more events are now being found in areas with good local coverage, such as Alaska, Mexico, Andean South America and New Zealand, where for various reasons not all locally determined origins are reported to ISC.

Eastern Indonesia


It should be noted that while Indonesia ceased to report hypocentres for a number of years, they have recently reinstated this contribution and it will be shown to be a valuable aid in constraining ISC earthquake locations, starting data-month February 1997.

New Zealand and Kermadec


Due to the limited azimuthal distribution of New Zealand stations with respect to local activity, many locations in the northern half of this area - in the region of the Kermadec Islands - would be unreliable using these stations alone. With the regional assistance of stations in Australia and Asia, it is often possible to obtain reasonably constrained locations.

Those earthquakes located under the islands of New Zealand are below the network's strict magnitude cut-off (MR=4).

Alaska


Athough station readings from the Northwest Territories and northern British Columbia may contribute, the majority of earthquakes in this region are located using NEIS data alone. That no epicentre was reported by them is an indication the events are of magnitude below the network's threshold. The few magnitudes obtained during ISC analysis support this theory.

As there were only 137 Alaskan events in the past two data years, their inclusion has a negligible affect on overall statistics.

Mexico


Until recently, only phase data were sent from Mexico's national seismic network. For completeness, it was necessary to locate earthquakes in this region during the Search analysis.

Mexican hypocentres will begin appearing in the January/February 1997 edition of the ISC Bulletin.

Andean South America


Earthquakes located in Search for Andean South America are generally done so with phase data from more than one local network, and often with contributions from the Brazilia Array, plus Central and North American stations. Occasionally, Australian and Asian stations are able to provide teleseismic phases.

While hypocentres in this region are not currently reported to ISC, it is hoped that this situation might change in the near future.

Conclusions

Although the practice of locating earthquakes using non-associated phase readings can be argued to be redundant in that many such events could be re-located events from below network magnitude cut-offs, the reasons for continuing this analysis are strong:

  • the primary contribution from Search is the location of earthquakes not reported by local networks
  • the accumulation of data from more that one local network (most likely from neighbouring countries) can result in a hypocentre estimate not attainable from a single network's data alone
  • as significant numbers of earthquakes are located in locally monitored regions, those networks are encouraged to include hypocentre estimates in the data they report, as has recently been the case for Mexico and Indonesia
  • ISC is able to provide a more complete picture of global seismicity