Data underlying the publication: Transdimensional ambient-noise surface wave tomography of the Reykjanes Peninsula, SW Iceland

This package includes a 3D shear wave velocity model of the Reykjanes Peninsula which is explained in detail in our accepted manuscript for publication in Geophysical Journal International (Rahimi Dalkhani et al., 2023). The ambient noise data were recorded in 2014-2015 in the context of the IMAGE (Integrated Methods for Advanced Geothermal Exploration) project. For details regarding the IMAGE project, we refer to Hersir et al. (2020) and Blanck et al. (2020). The open-access multi-institutional IMAGE data set is available at https://doi.org/10.14470/9Y7569325908. For further processing steps applied to the data, phase velocity retrieval algorithm, tomography algorithm, and geological interpretation of the recovered shear wave velocity model, see Rahimi Dalkhani et al. (2023).<br>In this study, we used a recently developed probabilistic tomographic algorithm (Zhang et al., 2018; Rahimi Dalkhani et al., 2021) to perform Ambient noise surface wave tomography of the Reykjanes peninsula. The shear wave velocities obtained in this study result from a 3D, one-step Bayesian tomographic inversion (Zhang et al., 2018), which has its roots in the transdimensional inversion algorithm introduced by Bodin &amp; Sambridge (2009). Rahimi Dalkhani et al. (2021) modified the algorithm in the sense that they update the ray paths less frequently (i.e., not at every perturbation step), while at the same time still honoring the non-linear aspect of the tomographic problem. They tested the modified algorithm on synthetic station-station travel times generated for the configuration of the extended IMAGE seismic network and the surface wave frequencies of interest (i.e., 0.1-0.5 Hz). In this study, we applied the modified algorithm to the extended IMAGE data set. First, we retrieved station-station surface wave phase travel times from the time-corrected ambient noise recordings (Weemstra et al., 2021). Then, we used these surface waves’ dispersion curves to generate 3D images of the RP subsurface’ shear wave velocity using the mentioned one-step transdimensional tomography algorithm. Finally, we interpret the recovered shear wave velocities, discuss how they compare to other recent geophysical studies, and list the most important conclusions (see Rahimi Dalkhani et al., 2023 for details).<br>In addition to the recovered shear wave velocity model of the Reykjanes peninsula, we included the following data and MATLAB scripts for the sake of the reproducibility of our results:Input files for the probabilistic inversion algorithm (i.e., MCTomo). The MCTomo software is open access and available at "https://blogs.ed.ac.uk/imaging/research/codes/" entitled "3D Monte Carlo tomography using both body and surface wave data". For more details see Rahimi Dalkhani et al. (2021, 2023). The modified package is available upon request.The retrieved dispersion curves that we used in our inversion algorithm both as frequency-dependent travel times and phase velocities.MATLAB data and scripts for reproducing all figures in our manuscript (Rahimi Dalkhani et al., 2023). It is worth mentioning that the MATLAB codes related to the phase velocity retrieval algorithm explained in Appendix A of Rahimi Dalkhani et al. (2023) are also included in the MATLAB scripts. See the script "FigureA1.m" for an example of how to use the codes.A "readme" file is included in each folder explaining the structure of the data and instructions. For the list of references we cited here see "ReferencesList.txt" in the main directory of the data.<br>

本数据包包含雷克雅内斯半岛（Reykjanes Peninsula）的三维剪切波速度（shear wave velocity）模型，相关细节已详细阐述于我们被《国际地球物理学报（Geophysical Journal International）》接收的待刊手稿（Rahimi Dalkhani等，2023）。该环境噪声数据于2014-2015年在IMAGE（Integrated Methods for Advanced Geothermal Exploration，先进地热勘探综合方法）项目框架下采集。关于IMAGE项目的详细信息，可参考Hersir等（2020）与Blanck等（2020）的研究。该多机构共建的开放获取IMAGE数据集可通过https://doi.org/10.14470/9Y7569325908获取。若需了解数据的后续处理步骤、相速度反演算法（phase velocity retrieval algorithm）、层析成像算法（tomography algorithm）以及反演得到的剪切波速度模型的地质解译方法，请参阅Rahimi Dalkhani等（2023）。<br>本研究采用新近开发的概率层析成像算法（probabilistic tomographic algorithm）对雷克雅内斯半岛开展环境噪声面波层析成像（Ambient noise surface wave tomography）研究（Zhang等，2018；Rahimi Dalkhani等，2021）。本研究得到的剪切波速度结果源自三维一步法贝叶斯层析反演（Bayesian tomographic inversion，Zhang等，2018），该方法的核心源于Bodin与Sambridge（2009）提出的跨维反演算法（transdimensional inversion algorithm）。Rahimi Dalkhani等（2021）对该算法进行了改进，减少了射线路径的更新频率（即并非每一步扰动步都更新射线路径），同时仍保留了层析成像问题的非线性特性。他们基于扩展IMAGE地震台网的台站间走时合成数据以及目标面波频段（0.1~0.5 Hz）对改进后的算法进行了测试。本研究将该改进算法应用于扩展IMAGE数据集。首先，我们从经过时间校正的环境噪声记录中提取台站间的面波相走时（Weemstra等，2021）；随后，利用这些面波的频散曲线（dispersion curves），通过前述一步法跨维层析成像算法生成雷克雅内斯半岛地下的三维剪切波速度图像；最后，我们对反演得到的剪切波速度进行解译，讨论其与其他近期地球物理研究的异同，并列出核心结论（详细内容参见Rahimi Dalkhani等，2023）。<br>除反演得到的雷克雅内斯半岛剪切波速度模型外，为确保研究结果可复现，我们还提供了以下数据与MATLAB脚本：概率反演算法（即MCTomo）的输入文件。MCTomo软件为开源软件，可在"https://blogs.ed.ac.uk/imaging/research/codes/"获取，该资源标题为"3D Monte Carlo tomography using both body and surface wave data"。更多细节可参阅Rahimi Dalkhani等（2021、2023）。修改后的软件包可按需获取。我们在反演算法中使用的提取得到的频散曲线，可作为频率相关走时与相速度使用。用于复现手稿中所有图表的MATLAB数据与脚本（Rahimi Dalkhani等，2023）。值得一提的是，Rahimi Dalkhani等（2023）附录A中提及的相速度反演算法相关MATLAB代码也已包含在本次提供的脚本中，可参考"FigureA1.m"脚本了解代码使用示例。每个文件夹中均包含"readme"文件，用于说明数据结构与操作指南。本文引用的完整参考文献列表可在数据主目录下的"ReferencesList.txt"文件中查看。