Automated Recovery of the 2012 Sumatera Aftershock Sequence Using Waveform Cross Correlation

We assess the performance of waveform cross correlation as applied to automatic and interactive processing of the aftershock sequence of the April 11, 2012 Sumatera Ms（IDC） = 8.2 earthquake. The XSEL （cross correlation standard event list） is compared to the REB （Reviewed Event Bulletin） issued by the IDC （International Data Centre）. The latter includes 1,181 aftershocks between April 11 and May 25. To automatically recover the sequence, sixteen aftershocks were selected from the IDC SEL3 （Standard Event List） available on April 13 and used as master events. Waveform templates from only seven IMS （International Monitoring System） array stations with the largest SNR （signal-to-noise ratio） for the main shock were used to calculate cross correlation coefficients. Approximately 620,000 arrivals were detected and then used to build events according to the IDC definition. After conflict resolution between event hypotheses with similar arrivals found by more than one master, all survived hypotheses populated the XSEL. The total number of distinct hypotheses was 4,924, with 2,763 hypotheses not matched by the REB, i.e. potentially new REB events. To evaluate the quality of the XSEL, we randomly selected a representative portion （-15%） from 2,763 hypotheses and reviewed them interactively according to standard IDC rules and guidelines. There were 119 new REB events built and 145 XSEL hypotheses rejected. Several frequency and probability density distributions have been constructed for all detections, the associated detections, the XSEL detections, the detections in the new REB, and rejected events.