Data assimilation system for seismoacoustic waves

The Great East Japan Earthquake, which took place on March 11, 2011, caused giant tsunamis and seriously damaged the Pacific coast area of the Tohoku district, Japan. Some studies reported that microbarometers in East Asia recorded atmospheric pressure perturbations, which are believed to be evidence of acoustic waves excited by the tsunamis, several tens of minutes after the main shock. This observation suggests a way to establish an early warning system that utilizes the fact that acoustic velocity is much faster than the propagation speed of a tsunami. Such a system would require data assimilation (DA), which is a fundamental technique to combine a numerical simulation and observation data, in order to predict the magnitude and arrival time of a tsunami and to warn people in coastal areas. The present paper introduces a DA system for seismoacoustic waves, which has been under development at the Research and Development Center for Data Assimilation at the Institute of Statistical Mathematics, Japan. We also validate this system by using a twin experiment that assumes atmospheric pressure variations excited by a giant earthquake in the Pacific Ocean off the coast of West Japan, which is expected to take place in the near future. The proposed system will eventually contribute to the early warning system for tsunamis, although the time needed for the DA computation and for the handling of observation data still needs to be improved.