Global Seismic Waveform Dataset from Raspberry Shake Geophones and MEMS Sensors for Machine Learning Applications

Global Seismic Waveform Dataset from Raspberry Shake Geophones and MEMS Sensors for Machine Learning Applications Naman Harish Nasara, Vipul Silwal*, N iv ika S in gh Department of Earth Sciences, Indian Institute of Technology Roorkee Email: [email protected] File Description 1. RSDB.zip – RSDB has three folders of the following name. a. RSGeo1C – single component geophone data b. RSGeo3C – 3component geophone data c. RSMESMS3C - 3component MEMS data Each of these folders has dataset file waveform.hdf5 and metadata.csv 2. Data_access.py – Python file for access each of the dataset. Detailed Description The dataset comprises of three subparts (detail in Table 1). Each waveform dataset is annotated and contains 35 attributes. The Pand S-wave arrivals are identified using PhaseNet (Zhu and Beroza, 2019) and SeisBench (Woollam et al., 2022). We employ the DiTing dataset (Zhao et al., 2023) for phase picking due to its superior performance compared to other available models. DiTing dataset was chosen because it performed significantly better than other dataset in detecting P and S arrivals (Table 2). Also after data curation, DiTing had relatively lesser erroneous detections (P and S retainment > 70%). Table 1: Summary of dataset present in RSDB dataset Table 2: Summary of decision P and S detection and percentage retainment after data curation. References: 1. Woollam, J., Münchmeyer, J., Tilmann, F., Rietbrock, A., Lange, D., Bornstein, T., Diehl, T., Giunchi, C., Haslinger, F., Jozinović, D., Michelini, A., Saul, J., & Soto, H. (2022). SeisBench—A toolbox for machine learning in seismology. Seismological Research Letters, 93(3), 1695–1709. https://doi.org/10.1785/0220210324 2. Zhao, M., Xiao, Z., Chen, S., & Fang, L. (2023). DiTing: A large-scale Chinese seismic benchmark dataset for artificial intelligence in seismology. Earthquake Science, 36(2), 84–94. https://doi.org/10.1016/j.eqs.2022.01.022 3. Zhu, W., & Beroza, G. C. (2019). PhaseNet: A deep-neural-network-based seismic arrival-time picking method. Geophysical Journal International, 216(1), 261–273. https://doi.org/10.1093/gji/ggy423