The Long – Short Term Memory (LSTM) algorithm have been applied to many fields, but its application to on-site earthquake early warning remains unexplored. This study aims to predict the Peak Ground Acceleration (PGA) of the upcoming seismic waves using LSTM. The inputs of the LSTM prediction model are the acceleration and velocity time histories of seismic waves truncated using six windows of 4-s length and a 0.5-s interval, and the LSTM model will predict the PGA time history based on the inputs. The grid search method is used to determine two hyper-parameters of the best LSTM model. The best LSTM model tends to over-predict PGA. However, when the threshold value is equal to 25 Gal, the classification performance is acceptable when one-level of intensity error is allowed. Compared to the prediction model that using 3-s of seismic wave after trigger, because once the LSTM model predicts PGA larger than the threshold, an alert is issued. Hence longer lead time could be achieved using the proposed LSTM model. Moreover, this study proposes an alternative algorithm to issue an alert. This alternative algorithm will reduce false positive alerts but also increase false negative alerts and shorten lead time. It is also found that the LSTM approach performs well when the threshold value is changed into 8 Gal or 250 Gal.

Alcik, H., Ozel, O., Apaydin, N., & Erdik, M. (2009). A study on warning algorithms for Istanbul earthquake early warning system. Geophysical Research Letters, 36(5).

Allen, R. M., Gasparini, P., Kamigaichi, O., & Bose, M. (2009). The status of earthquake early warning around the world: An introductory overview. Seismological Research Letters, 80(5), 682-693.

Allen, R. M., & Kanamori, H. (2003). The potential for earthquake early warning in southern California. Science, 300(5620), 786-789.

Behr, Y., Clinton, J. F., Cauzzi, C., Hauksson, E., Jónsdóttir, K., Marius, C. G., Pinar, A., Salichon, J., & Sokos, E. (2016). The Virtual Seismologist in SeisComP3: A new implementation strategy for earthquake early warning algorithms. Seismological Research Letters, 87(2A), 363-373.

Böse, M., Heaton, T., & Hauksson, E. (2012). Rapid estimation of earthquake source and ground‐motion parameters for earthquake early warning using data from a single three‐component broadband or strong‐motion sensor. Bulletin of the Seismological Society of America, 102(2), 738-750.

Crowell, B. W., Schmidt, D. A., Bodin, P., Vidale, J. E., Baker, B., Barrientos, S., & Geng, J. (2018). G‐FAST earthquake early warning potential for great earthquakes in Chile. Seismological Research Letters, 89(2A), 542-556.

Cua, G., Fischer, M., Heaton, T., & Wiemer, S. (2009). Real-time performance of the Virtual Seismologist earthquake early warning algorithm in southern California. Seismological Research Letters, 80(5), 740-747.

Cuéllar, A., Suárez, G., & Espinosa‐Aranda, J. (2017). Performance evaluation of the earthquake detection and classification algorithm 2 (t S–t P) of the Seismic Alert System of Mexico (SASMEX). Bulletin of the Seismological Society of America, 107(3), 1451-1463.

Given, D. D., Allen, R. M., Baltay, A. S., Bodin, P., Cochran, E. S., Creager, K., de Groot, R. M., Gee, L. S., Hauksson, E., & Heaton, T. H. (2018). Revised technical implementation plan for the ShakeAlert system—An earthquake early warning system for the West Coast of the United States (2331-1258).

Hebb, D. O. (1949). The organisation of behaviour: a neuropsychological theory. Science Editions New York.

Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural computation, 9(8), 1735-1780.

Hoshiba, M., Kamigaichi, O., Saito, M., Tsukada, S. y., & Hamada, N. (2008). Earthquake early warning starts nationwide in Japan. EOS, Transactions American geophysical union, 89(8), 73-74.

Hsiao, N. C., Wu, Y. M., Shin, T. C., Zhao, L., & Teng, T. L. (2009). Development of earthquake early warning system in Taiwan. Geophysical research letters, 36(5).

Hsu, T.-Y., & Huang, C.-W. (2021). Onsite Early Prediction of PGA Using CNN with Multi-Scale and Multi-Domain P-Waves as Input. Frontiers in Earth Science, 9, 247.

Hsu, T.-Y., Huang, S.-K., Chang, Y.-W., Kuo, C.-H., Lin, C.-M., Chang, T.-M., Wen, K.-L., & Loh, C.-H. (2013). Rapid on-site peak ground acceleration estimation based on support vector regression and P-wave features in Taiwan. Soil Dynamics and Earthquake Engineering, 49, 210-217.

Hsu, T.-Y., Lin, P., Wang, H., Chiang, H., Chang, Y., Kuo, C., Lin, C., & Wen, K. (2018). Comparing the performance of the NEEWS earthquake early warning system against the CWB system during the 6 February 2018 Mw 6.2 Hualien earthquake. Geophysical Research Letters, 45(12), 6001-6007.

Hsu, T. Y., Kuo, C. H., Wang, H. H., Chang, Y. W., Lin, P. Y., & Wen, K. L. (2021). The realization of an earthquake early warning system for schools and its performance during the 2019 ML 6.3 Hualien (Taiwan) earthquake. 92(1), 342-351.

Hsu, T. Y., Wu, R.-T., Liang, C.-W., Kuo, C.-H., & Lin, C.-M. (2020). Peak ground acceleration estimation using P-wave parameters and horizontal-to-vertical spectral ratios. Terr. Atmos. Ocean. Sci, 31, 1-8.

Hsu, T. Y., Wu, R. T., & Chang, K. C. (2016). Two novel approaches to reduce false alarm due to non‐earthquake events for on‐site earthquake early warning system. Computer‐Aided Civil and Infrastructure Engineering, 31(7), 535-549.

Ji, J., Gao, Y., Lü, Q., Wu, Z., Zhang, W., & Zhang, C. (2019). China's early warning system progress. Science, 365(6451), 332-332.

Kamigaichi, O., Saito, M., Doi, K., Matsumori, T., Tsukada, S. y., Takeda, K., Shimoyama, T., Nakamura, K., Kiyomoto, M., & Watanabe, Y. (2009). Earthquake early warning in Japan: Warning the general public and future prospects. Seismological Research Letters, 80(5), 717-726.

Kanamori, H. (2005). Real-time seismology and earthquake damage mitigation. Annu. Rev. Earth Planet. Sci., 33, 195-214.

Kumar, A., Mittal, H., Chamoli, B., Gairola, A., Jakka, R., & Srivastava, A. (2014). Earthquake early warning system for northern India. 15th symposium on earthquake engineering, Indian Institute of Technology, Roorkee,

Mărmureanu, A., Ionescu, C., & Cioflan, C. (2011). Advanced real-time acquisition of the Vrancea earthquake early warning system. Soil Dynamics and Earthquake Engineering, 31(2), 163-169.

Meier, M. A. (2017). How “good” are real‐time ground motion predictions from earthquake early warning systems? Journal of Geophysical Research: Solid Earth, 122(7), 5561-5577.

Milenkoski, M., Trivodaliev, K., Kalajdziski, S., Jovanov, M., & Stojkoska, B. R. (2018). Real time human activity recognition on smartphones using LSTM networks. 2018 41st International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO),

Nakamura, H., Horiuchi, S., Wu, C., Yamamoto, S., & Rydelek, P. A. (2009). Evaluation of the real‐time earthquake information system in Japan. Geophysical Research Letters, 36(5).

Nakamura, Y. (1988). On the urgent earthquake detection and alarm system (UrEDAS). Proc. of the 9th World Conference on Earthquake Engineering,

Nakamura, Y. (1996). Research and development of intelligent earthquake disaster prevention systems UrEDAS and HERAS. Doboku Gakkai Ronbunshu, 1996(531), 1-33.

Nakamura, Y., & Saita, J. (2007). UrEDAS, the earthquake warning system: Today and tomorrow. In Earthquake early warning systems (pp. 249-281). Springer.

Nof, R. N., & Allen, R. M. (2016). Implementing the ElarmS earthquake early warning algorithm on the Israeli seismic network. Bulletin of the Seismological Society of America, 106(5), 2332-2344.

Odaka, T., Ashiya, K., Tsukada, S. y., Sato, S., Ohtake, K., & Nozaka, D. (2003). A new method of quickly estimating epicentral distance and magnitude from a single seismic record. Bulletin of the Seismological Society of America, 93(1), 526-532.

Pazos, A., Romeu, N., Lozano, L., Colom, Y., López Mesa, M., Goula, X., Jara, J., Cantavella, J., Zollo, A., & Hanka, W. (2015). A regional approach for earthquake early warning in south west Iberia: A feasibility study. Bulletin of the Seismological Society of America, 105(2A), 560-567.

Roondiwala, M., Patel, H., & Varma, S. (2017). Predicting stock prices using LSTM. International Journal of Science and Research (IJSR), 6(4), 1754-1756.

Rosenblatt, F. (1958). The perceptron: a probabilistic model for information storage and organization in the brain. Psychological review, 65(6), 386.

Saita, J., & Nakamura, Y. (2003). UrEDAS: the early warning system for mitigation of disasters caused by earthquakes and tsunamis. In Early warning systems for natural disaster reduction (pp. 453-460). Springer.

Sheen, D. H., Park, J. H., Chi, H. C., Hwang, E. H., Lim, I. S., Seong, Y. J., & Pak, J. (2017). The first stage of an earthquake early warning system in South Korea. Seismological Research Letters, 88(6), 1491-1498.

Strauch, W., Talavera, E., Tenorio, V., Ramírez, J., Argüello, G., Herrera, M., Acosta, A., & Morales, A. (2018). Toward an earthquake and tsunami monitoring and early warning system for Nicaragua and Central America. Seismological Research Letters, 89(2A), 399-406.

Violos, J., Tsanakas, S., Androutsopoulou, M., Palaiokrassas, G., & Varvarigou, T. (2020). Next Position Prediction using LSTM Neural Networks. 11th Hellenic Conference on Artificial Intelligence,

Wang, C. Y. (2020). A LSTM Neural Network for On-site Earthquake Early Warning. EGU General Assembly Conference Abstracts,

Wang, H., & Raj, B. (2017). On the origin of deep learning. arXiv preprint arXiv:1702.07800.

Wu, Y.-M., & Kanamori, H. (2005). Rapid assessment of damage potential of earthquakes in Taiwan from the beginning of P waves. Bulletin of the Seismological Society of America, 95(3), 1181-1185.

Zollo, A., Colombelli, S., Elia, L., Emolo, A., Festa, G., Iannaccone, G., Martino, C., & Gasparini, P. (2014). An integrated regional and on-site Earthquake Early Warning System for Southern Italy: Concepts, methodologies and performances. In Early warning for geological disasters (pp. 117-137). Springer.