Description We converted the smooth depth-varying velocity structure model of Tonegawa et al. (2017) to a 5-layer model beneath each DONET station. The physical parameters of each layer are listed in the vmodel.csv. The thicknesses of each layer were determined by fitting the depth-averaged S-wave velocities derived by Tonegawa et al. (2017). For example, if the depth-averaged VS of Tonegawa et al. (2017) become the VS of layer 1 at a certain depth, this depth is considered as the bottom of layer 1. In DONET_layeredData.csv, the estimated bottom depths of each layer are listed. For seismic wave propagation simulation For simulations of seismic wave propagation along the Nankai Trough, the 3D model used in the simulations was basically constructed from the Japan Integrated Velocity Structure Model (JIVSM) (Koketsu et al., 2012). The JIVSM onshore and outer-rise sedimentary structures and structures beneath bedrock were fixed. To construct 3D model of the accretionary prism from layered S wave velocity models in "DONET_layeredData.csv," each station's bottom depths were interpolated and extrapolated via the ‘Surface’ gridding algorithm in Generic Mapping Tools software (GMT; Wessel et al. 2013). Interpolation and extrapolation were only applied within the region of the accretionary prism (Figure S1 of Takemura, Yabe & Emoto 2020). By using interpolated and extrapolated data of layer bottom depths and physical parameters (vmodel.csv), we can obtain 3D model of the accretionary prism along the Nankai Trough. We confirmed very similar simulation results between smooth depth-varying and layered accretionary prism models. Smooth depth varying model case Takemura, Kubo et al., 2019 Takemura, Matsuzawa et al., 2019 Layered model case Figures S3, S4 of Takemura, Yabe & Emoto 2020 Related papers For citing general information of this dataset, please include this data DOI and the following references Tonegawa, T., Araki, E., Kimura, T. et al. (2017). Sporadic low-velocity volumes spatially correlate with shallow very low frequency earthquake clusters. Nat Commun 8, 2048 https://doi.org/10.1038/s41467-017-02276-8 Takemura, S., Yabe, S., & Emoto (2020), K. Modelling high-frequency seismograms at ocean bottom seismometers: effects of heterogeneous structures on source parameter estimation for small offshore earthquakes and shallow low-frequency tremors, Geophys. J. Int., 223 (3), 1708-1723, https://doi.org/10.1093/gji/ggaa404