Earthquake catalogs for: Improving microearthquake detection in the Val d'Agri region (Southern Italy) with deep learning

This repository contains two earthquake catalogs (dubbed PRN and QS) obtained from the application of two deep-learning-based detection workflows to continuous seismic data recorded in the Val d’Agri region (Southern Italy). These catalogs have been generated using the PhaseNet neural network for seismic phase picking (Zhu & Beroza, 2019). The workflows used to generate the catalogs are described in detail in: Caredda et al. (2025). These datasets offer a more comprehensive representation of local seismicity compared to manually generated, STA/LTA-based catalogs (available in the open periodic monitoring reports accessible at: https://cms.ingv.it/sperimentazioni/val-d-agri [last accessed on 18/09/2025]). The datasets include event origin times, locations, magnitudes, location uncertainties, and phase arrival times with corresponding PhaseNet “pick probabilities” (for the PRN catalog), providing an enriched representation of local seismicity compared to conventional STA/LTA-based catalogs. These catalogs can serve as valuable resources for further research on seismicity, induced processes, Earth structure, and seismic hazard assessment in the Val d’Agri region. References: Caredda, E., Isken, M. P., Cesca, S., Errico, M., Zerbinato, G., & Morelli, A. (2025). Improving microearthquake detection in the Val d’Agri region (Southern Italy) with deep learning. Seismica, 4(2). Zhu, W., and Beroza, G. C. (2018). 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本仓库包含两份地震目录（分别命名为PRN与QS），二者均通过将两种基于深度学习的检测流程应用于意大利南部瓦莱达格里（Val d’Agri）地区记录的连续地震数据生成。上述目录均采用用于震相拾取的PhaseNet神经网络生成（Zhu & Beroza, 2019）。用于生成该目录的检测流程已由Caredda等人（2025）详细阐述。相较于人工生成的基于STA/LTA（短时段平均/长时段平均比）的地震目录（可从公开定期监测报告中获取，访问链接：https://cms.ingv.it/sperimentazioni/val-d-agri，最后访问时间：2025年9月18日），本数据集能够更全面地刻画当地地震活动特征。 本数据集包含地震事件的发震时刻、空间位置、震级、定位不确定性，以及带有对应PhaseNet“拾取概率”的震相到时信息（适用于PRN目录），相较传统基于STA/LTA的地震目录，实现了对当地地震活动更丰富的表征。 上述两份地震目录可作为瓦莱达格里地区地震活动、诱发过程、地球内部结构以及地震危险性评估相关后续研究的宝贵资源。 参考文献： Caredda, E., Isken, M. P., Cesca, S., Errico, M., Zerbinato, G., & Morelli, A. (2025). Improving microearthquake detection in the Val d’Agri region (Southern Italy) with deep learning. Seismica, 4(2). Zhu, W., and Beroza, G. C. (2018). PhaseNet: A Deep-Neural-Network-Based Seismic Arrival Time Picking Method. 国际地球物理期刊（Geophysical Journal International）, 216(1), 261–273. https://doi.org/10.1093/gji/ggy423