GPS-Shield: Towards Robust Characterization of Tsunami Source in Few Minutes after an Earthquake

The catastrophic consequences of 2004 Indian-Ocean and 2006 Java tsunamis demand for developing modern and robust tsunami early warning systems. The greatest challenge of the GITEWS (German Indonesian Tsunami Early Warning System), led by the GFZ, is to provide early tsunami warning for the Indian-Ocean coast of Indonesia located close to the potential tsunami source and where tsunamis are expected to arrive only 20-40 minutes after an earthquake. Usually the base for the tsunami early warning systems is the estimation of the earthquake magnitude using seismic methods. However, modeling of the tsunami source and wave propagation shows that teleseismic-magnitude methods may fail in predicting the tsunami amplitude in the city of Padang (west Sumatra, more than 750,000 inhabitants). This is because of the presence of the massive Mentawai islands between Padang and the subduction trench. Depending on the slip distribution at the fault, the major co-seismic uplift may happen either on the ocean-side or the Sumatra-side of the islands, resulting in a 5 (!) times different tsunami height in Padang for earthquakes of the same magnitude. This result is an extreme manifestation of the previously known fact that for reliable prediction of tsunami wave heights in the near-field (like at the Sumatran or Java coasts) even precise knowledge of earthquake magnitude is not sufficient in general. To solve this problem we suggested a novel real-time warning technique based on special types of near-field GPS arrays ('GPS Shield') that are appropriate for Sumatra and elsewhere around the globe (Sobolev et al, 2006).

Key elements of the “GPS Shield” concept for Sumatra are arrays of several real-time high-rate (1-10 Hz) GPS stations located on the islands west of Sumatra and at the Sumatran coast (left-side inset in Fig.1). Stations on the islands are closely spaced (10-20 km) and aligned perpendicular to the trench, i.e., parallel to the expected gradient of surface coseismic displacement. The arrays measure surface displacements using instantaneous GPS positioning techniques that, for single-epoch 1-Hz displacement measurements, achieve an accuracy of better than several centimeters for absolute displacements and a few centimeters for relative displacements between the closely-spaced stations. The GPS-Shield arrays are set along the trench with a spacing of 100-200 km, a spacing that is appropriate to resolve major trench-parallel slip heterogeneity. In the case where there are no islands between the trench and Sumatra, we suggest the use of ocean buoys equipped with GPS devices and bottom pressure sensors.

References

 

• Sobolev, S.V., Babeyko, A.Y., Rongjiang Wang, R., Galas, R., Rothacher, M., Sein, D.V., Schröter, J., Lauterjung, J. and Subarya, C. (2006): Towards real-time tsunami amplitude prediction, EOS, 87, No 37, p. 374, 378
• Sobolev, S.V., Babeyko, A.Y., Wang, R., Galas, R., Rothacher, M., Sein, D.V., Schröter, J., Lauterjung, J. and Subarya, C. (2007) Concept for fast and reliable tsunami early warning using “GPS-Shield” arrays, J. Geophys. Res., 112, B08415

Global application of the GPS-Shield concept. Solid lines show subduction zones where the land is located closer than 100 km to the seismogenic zone. Dashed lines are the zones where the GPS-Shield arrays can at least resolve the seismic moment of the closest large earthquakes, also within less than 10 minutes of an earthquake. Left-side inset shows the conceptual distribution of observational instruments in ‘GPS-shield’ for Sumatra. Red circles are real-time GPS stations (those with white rings are also equipped with broad-band seismometers and accelerometers). Red diamonds are GPS-equipped buoys. Right-side inset shows reconstructions of tsunamogenic sea-bottom displacements from inversion of synthetic observations at the three-station GPS array (triangles) for two different slip distributions at the fault.