Epicenter estimation using erroneous crustal model(s) and skew regional networks

Abstract

A simple, hybrid procedure is proposed for seismic location at regional distances when (1) the crustal model is not known, (2) the seismic network is not calibrated, and (3) the azimuthal coverage of recording stations is poor. Among all parameters related to event location, the backazimuth estimated with a skew, large-aperture network appears to be the least sensitive to the model error—under these difficult yet realistic conditions. This observation leads to a hypothesis that epicenters could be better constrained through triangulation using backazimuths derived from two skew, large-aperture subnetworks or arrays. If two subnetworks can be found such that the corresponding backazimuths are sufficiently apart (say, they intersect at an angle of 20° or larger), the resulting absolute mislocation associated with the intersection point is often less than 18 km, which is the permissible location accuracy as quoted in the Comprehensive Nuclear Test Ban Treaty (CTBT). The origin time and source depth can then be obtained from a standard location procedure with the approximate epicenter held fixed. If the network cannot be partitioned to provide two `good' backazimuths, then the conventional location procedure (which feeds on the phase picks altogether) can still be used. The proposed algorithm is tested with three events: (1) U.S. explosion Salmon of Oct. 22, 1964, (2) Kola explosion of Sept. 29, 1996, and (3) Pakistani event of April 28, 1991. The absolute location error associated with this hybrid method is typically comparable to or, in many cases, better than that of the conventional method. Furthermore, this new approach also offers a measurable quantity which positively flags when the location is more likely to be good—even for regions that are not calibrated. These features could have important implications for the CTBT implementation as the proposed algorithm can readily be used in areas previously not well studied as an ad hoc procedure on problem events while the International Monitoring System (IMS) stations are being deployed and calibrated gradually. For local seismic network operators, this algorithm could be an attractive alternative or supplementary location procedure.