Introduction

Much of the history of forensic seismology has seen identification criteria introduced with claims that verification is solved, only to be followed by evidence that the claims were over optimistic and that exceptions have been found. Before AWE Blacknest discovered the advantages of deconvolving SP P seismograms, including correcting for attenuation, an analysis procedure was devised to collect statistics on the effectiveness of what then seemed to be the most promising criteria – first motion, spectral ratios and complexity – using initially only the array seismograms. Later the measurement of Ms at stations of the WWSSN was included so that the effectiveness of the mb:Ms criterion could be evaluated. This chapter describes the procedure used for the collection of identification statistics, and then goes on to look in detail at the identification criteria that have proved to be the most useful.

Routine processing and analysis

By 1964 AWE Blacknest had settled on a standard procedure for the analysis of the array seismograms. For each P signal an eight-channel output, such as that shown in Figure 9.1, is produced; the channels are listed in Table 9.1. Channel 2, the DS, is the most important channel, from which onsets and amplitudes are ideally read.