Development of the modified time-to-failure method and main shock location search technique for intermediate-term earthquake prediction

Daniel Jirush Brehm, Purdue University

Abstract

The purpose of the study was to evaluate the potential of the time-to-failure method as a tool for intermediate-term earthquake prediction. The investigation in this thesis consisted of validating the accelerated energy release phenomenon, which is utilized in the time-to-failure method, and providing empirical and theoretical bases for the determination of two of the four unknown parameters. The empirical relationships constraining the two parameters, the coefficient k/m and exponent m in the modified time-to-failure equation, were identified by using the actual values of the modeled main shocks in a retrospective prediction of the main shock. The modeled main shocks that were used in this investigation are from three different catalogs, the New Madrid Seismic Zone, western United States, and southern California. The remaining two unknown parameters are the time-of-failure and the magnitude of the main shock. The predicted values are determined by constructing a prediction contour diagram. The modified time-to-failure method does not work when two main shocks occur near each other in both space and time (interfering events), or if the earthquake catalog, from which the precursory events are selected, is not complete for the minimum size precursory event (magnitudes larger than about 2.5 units smaller than the main shock). The modified time-to-failure equation is utilized in a main shock search location technique which is a method for identifying the location of an impending main shock. A systematic search of a particular region for a future main shock is conducted at defined grid points. The technique identifies and records acceleration sequence and search radius information associated with a given search magnitude for each grid point. The acceleration search radius data are used to calculate a Normalized Search Radius (NSR) range for each grid point. The NSR data are contoured, and local maximum of the NSR anomaly corresponds to the impending main shock location. The investigation of the time-to-failure method resulted in the development of a modified intermediate-term earthquake prediction method that provides the predicted time, size, and location of a future main shock using precursory earthquakes available from present earthquake catalogs. The error estimates associated with each predicted value allow for an objective determination of the success or failure of the prediction. The entire research project (using all three earthquake catalogs) consisted of evaluating 60 separate main events, 26 of the events could be modeled by the modified time-to-failure method. The 34 main shocks that could not be modeled either were located near interfering events, or the earthquake catalog was not complete for the minimum size precursory event. In addition, an analysis of the practical application of the method indicates that up to 33% of the acceleration sequences may end with several small events instead of one main shock (a false-positive sequence). However, all main shocks that met the acceleration sequence criteria were observed to have accelerated energy releases that could be modeled by the modified time-to-failure method (no false-negatives were identified for main shocks, in any of the three earthquake catalogs, that met both acceleration sequence criteria). In retrospective predictions, errors in main shock magnitude are usually less than about 0.5 magnitude units. Errors in the time-of-failure are usually less than about one year. (Abstract shortened by UMI.)

Degree

Ph.D.

Advisors

Braile, Purdue University.

Subject Area

Geophysics

COinS