5:15 PM - 6:45 PM
[HDS11-P07] Tsunami Forecast Method Using Tsunami Data Assimilation with Real-time Source Estimation
Keywords:Tsunami real-time forecast, Tsunami Data Assimilation, 1896 Sanriku tsunami earthquake
The 2011 Tohoku earthquake and tsunami caused catastrophic damage along Japan’s Pacific coast, underscoring the urgent need for more robust tsunami early warning systems. In response, the Seafloor Observation Network for Earthquakes and Tsunamis along the Japan Trench (S-net) was established. This network, comprising 150 observation devices interconnected by approximately 5,500 km of optical seabed cables, enables real-time transmission of observed data to onshore stations. After establishing of such a seafloor observation network, the Tsunami Data Assimilation Method (Maeda et al., 2015) has been developed. One advantage of this tsunami data assimilation method is that tsunami wavefields are computed without earthquake source information. Because tsunamis are observed by ocean bottom pressure sensors in S-net, the method's effectiveness is hampered by the unavailability of data above the source area due to strong ground motions and the inability to measure the initial tsunami height in the source area. Consequently, those problems cause a significant time delay for accurate tsunami forecasts. To solve these problems and shorten the time required for accurate tsunami forecast, we propose an approach that combines tsunami data assimilation with real-time source estimation result such as REGARD (Kawamoto et al., 2017) using GNSS data observed by GEONET.
In this paper, the effectiveness of our method was tested for the 1896 Sanriku tsunami earthquake case. The slip distribution of the 1896 Sanriku tsunami earthquake was estimated by Satake et al. (2017). The tsunami waveforms at S-net stations numerically computed from that slip distribution was used as the observation data. The stations above the source area were eliminated because of the above problems. The tsunami data assimilation was first performed using that data set. The data assimilation was stopped at a certain time to forecast the tsunami. The tsunami computation was continued to obtain forecasted tsunami waveforms along the Sanriku coast. The tsunami waveforms along the Sanriku coast were compared with the reference tsunami waveforms which computed directly from the slip distribution estimated by Satake et al. (2017). Then, the data assimilations with simple rectangular source models were performed to shorten the data assimilation time for a tsunami forecast. The rectangular fault model is expected to be obtained in real-time. We first used the rectangular fault model estimated by Tanioka and Satake (1996). We also varied the width of faults from that rectangular fault and the location of the fault by considering the uncertainty of the real-time source model estimation using GSNN data. The results showed that the data assimilations with the expected real-time source model shorten the data assimilation time to forecast accurate tsunami heights along the Sanriku coast.
In conclusion, the tsunami data assimilation with the real-time source estimation is a powerful technique to shorten the tsunami forecast time.
References
Kawamoto S, Ohta Y, Hiyama Y et al (2017) REGARD: a new GNSS-based real-time finite fault modeling system for GEONET. J Geophys Res [solid Earth] 122:1324–1349. https://doi.org/10.1002/2016jb013485
Maeda, T., Obara, K., Shinohara, M., Kanazawa, T., & Uehira, K. (2015). Successive estimation of a tsunami wavefield without earthquake source data: A data assimilation approach toward real-time tsunami forecating. Geophys. Res. Lett., 42, 7923–7932. https://doi.org/10.1002/2015GL065588
Satake, K., Fujii, Y. & Yamaki, S. Different depths of near-trench slips of the 1896 Sanriku and 2011 Tohoku earthquakes. Geosci. Lett. 4, 33 (2017). https://doi.org/10.1186/s40562-017-0099-y
Tanioka Y, Satake K (1996) Fault parameters of the 1896 Sanriku tsunami earthquake estimated from tsunami numerical modeling. Geophys Res Lett 23:1522–1549
In this paper, the effectiveness of our method was tested for the 1896 Sanriku tsunami earthquake case. The slip distribution of the 1896 Sanriku tsunami earthquake was estimated by Satake et al. (2017). The tsunami waveforms at S-net stations numerically computed from that slip distribution was used as the observation data. The stations above the source area were eliminated because of the above problems. The tsunami data assimilation was first performed using that data set. The data assimilation was stopped at a certain time to forecast the tsunami. The tsunami computation was continued to obtain forecasted tsunami waveforms along the Sanriku coast. The tsunami waveforms along the Sanriku coast were compared with the reference tsunami waveforms which computed directly from the slip distribution estimated by Satake et al. (2017). Then, the data assimilations with simple rectangular source models were performed to shorten the data assimilation time for a tsunami forecast. The rectangular fault model is expected to be obtained in real-time. We first used the rectangular fault model estimated by Tanioka and Satake (1996). We also varied the width of faults from that rectangular fault and the location of the fault by considering the uncertainty of the real-time source model estimation using GSNN data. The results showed that the data assimilations with the expected real-time source model shorten the data assimilation time to forecast accurate tsunami heights along the Sanriku coast.
In conclusion, the tsunami data assimilation with the real-time source estimation is a powerful technique to shorten the tsunami forecast time.
References
Kawamoto S, Ohta Y, Hiyama Y et al (2017) REGARD: a new GNSS-based real-time finite fault modeling system for GEONET. J Geophys Res [solid Earth] 122:1324–1349. https://doi.org/10.1002/2016jb013485
Maeda, T., Obara, K., Shinohara, M., Kanazawa, T., & Uehira, K. (2015). Successive estimation of a tsunami wavefield without earthquake source data: A data assimilation approach toward real-time tsunami forecating. Geophys. Res. Lett., 42, 7923–7932. https://doi.org/10.1002/2015GL065588
Satake, K., Fujii, Y. & Yamaki, S. Different depths of near-trench slips of the 1896 Sanriku and 2011 Tohoku earthquakes. Geosci. Lett. 4, 33 (2017). https://doi.org/10.1186/s40562-017-0099-y
Tanioka Y, Satake K (1996) Fault parameters of the 1896 Sanriku tsunami earthquake estimated from tsunami numerical modeling. Geophys Res Lett 23:1522–1549