Performance Review of the Global Seismographic Network for the Sumatra-Andaman Megathrust Earthquake

Abstract

On 26 December 2004, a 1,200-km length of seafloor boundary between the India Plate and Burma microplate ruptured in the Sumatra-Andaman earthquake. This earthquake was one of the five largest earthquakes of the past century and the largest in the past four decades. The Sumatra-Andaman earthquake is the first large tsunamigenic event and the first with an estimated Mw ≥ 9 to be recorded by the Global Seismographic Network (GSN; Figure 1; Butler et al., 2004), as well as the observatories of the broader Federation of Digital Seismographic Networks (FDSN; Dziewonski, 1994). Earthquakes with Mw ≥ 8 are commonly termed “great” earthquakes, but those with Mw ≥ 8.7, not experienced on Earth since the 1960's, present hazards to lives and property that are far more extensive than a typical “great” earthquake. We therefore adopt the term “megathrust earthquake” after the common usage among paleoseismologists for exceptionally destructive earthquakes in the past ( e.g., Priest et al., 2000; Cummins et al., 2001; Leonard et al., 2004). The performance of the GSN during the Sumatra-Andaman earthquake, in combination with permanent broadband seismic observatories in other global networks ( e.g., PACIFIC21, GEOSCOPE, GEOFON), has more than technical and scientific implications. The tsunami generated by this seismic event caught the coastline of the Bay of Bengal largely unaware, took more than 200,000 lives, and destroyed innumerable communities and livelihoods. The event was tragic in the sense that some of its consequences could have been avoided. 20th-century experience with long-range tsunami hazards motivated the establishment of a sophisticated tsunami warning system for the Pacific Ocean basin (Pararas-Carayannis, 1984; Uchiike and Hosono, 1995; Dudley and Lin, 1998). Although the potential for destructive tsunamis in the Indian Ocean basin was recognized (Rynn and Davidson, 1999; Cummins et al., 2004 …