BUSCADOR RECOLECTA

The dataset consists of 14 directories, one for each earthquake analyzed, including raw data and results. A README.txt file describes the contents of the directories., Data files and modelling results of regional centroid moment tensors obtained for earthquakes in the 2013 CASTOR gas storage seismic crisis, and presented in the article "Fault reactivation by gas injection at an underground gas storage off the east coast of Spain”, by Antonio Villaseñor et al. During September-October of 2013 an intense swarm of earthquakes occurred off the east coast of Spain associated with the injection of the base gas in an offshore underground gas storage. Two weeks after the end of the injection operations, three moderate-sized earthquakes (Mw 4.0-4.1) occurred near the storage. These events were widely felt by the nearby population, leading to the indefinite shut-down of the facility. Here we investigate the source parameters (focal depth and mechanism) of the largest earthquakes in the sequence in order to identify the faults reactivated by the gas injection, and to help understand the processes that caused the earthquakes. Our waveform modeling results indicate that the largest earthquakes occurred at depths of 6-8 km beneath the sea floor, significantly deeper than the injection depth (~ 1800 m). Although we cannot undoubtedly discriminate the fault plane from the two nodal planes of the mechanisms, most evidence seems to favor a NW-SE striking fault plane. We propose that the gas injection reactivated unmapped faults in the Paleozoic basement, with regional orientation possibly inherited from the opening of the Valencia Trough., We thank the seismic networks that provided the waveform data used in this study: IGN
295 (https://doi.org/10.7914/SN/ES), and ICGC (https://doi.org/10.7914/SN/CA). This research was funded by project SEAL
(Ministerio de Ciencia e Innovación, Spain, CGL2017-88864-R), Peer reviewed

12 pages, 7 figures, 3 tables, supplementary material https://doi.org/10.5194/se-11-63-2020.-- Data availability http://dx.doi.org/10.20350/digitalCSIC/8966, During September-October of 2013 an intense swarm of earthquakes occurred off the east coast of Spain associated with the injection of the base gas in an offshore underground gas storage. Two weeks after the end of the injection operations, three moderate-sized earthquakes (Mw 4.0-4.1) occurred near the storage. These events were widely felt by the nearby population, leading to the indefinite shut-down of the facility. Here we investigate the source parameters (focal depth and mechanism) of the largest earthquakes in the sequence in order to identify the faults reactivated by the gas injection and to help understand the processes that caused the earthquakes. Our waveform modeling results indicate that the largest earthquakes occurred at depths of 6-8 km beneath the sea floor, significantly deeper than the injection depth (∼1800 m). Although we cannot undoubtedly discriminate the fault plane from the two nodal planes of the mechanisms, most evidence seems to favor a NW–SE-striking fault plane. We propose that the gas injection reactivated faults in the Paleozoic basement, with regional orientation possibly inherited from the opening of the Valencia Trough, This research has been supported by the Ministerio de Ciencia e Innovación, Spain (grant no. CGL2017-88864-R), With the funding support of the ‘Severo Ochoa Centre of Excellence’ accreditation (CEX2019-000928-S), of the Spanish Research Agency (AEI)

10 pages, 4 figures, supporting information https://doi.org/10.1029/2020GL092038.-- Data Availability Statement: The associated data is available at the repository DIGITAL.CSIC (https://digital.csic.es/handle/10261/216863, The offshore Castor Underground Gas Storage (UGS) project had to be halted after gas injection triggered three M4 earthquakes, each larger than any ever induced by UGS. The mechanisms that induced seismicity in the crystalline basement at 5–10 km depth after gas injection at 1.7 km depth remain unknown. Here, we propose a combination of mechanisms to explain the observed seismicity. First, the critically stressed Amposta fault, bounding the storage formation, crept by the superposition of well‐known overpressure effects and buoyancy of the relatively light injected gas. This aseismic slip brought an unmapped critically stressed fault in the hydraulically disconnected crystalline basement to failure. We attribute the delay between induced earthquakes to the pressure drop associated to expansion of areas where earthquakes slips cause further instabilities. Earthquakes occur only after these pressure drops have dissipated. Understanding triggering mechanisms is key to forecast induced seismicity and successfully design deep underground operations., The authors would like to acknowledge Álvaro González for sharing the catalogs that were used in Cesca et al. (2014). Funding: Víctor Vilarrasa acknowledges funding from the European Research Council (ERC) under the European Union’s Horizon 2020 Research and Innovation Program through the Starting Grant GEoREST (www.georest.eu) (grant agreement No. 801809). Antonio Villaseñor acknowledges funding from Spanish Ministry of Science and Innovation grant CGL2017‐88864‐R. IDAEA‐CSIC is a Center of Excellence Severo Ochoa (Spanish Ministry of Science and Innovation, Project CEX2018‐000794‐S). ICM‐CSIC is a Center of Excellence Severo Ochoa (Spanish Ministry of Science and Innovation, Project CEX2019‐000928‐S)., Peer reviewed

Data and figures for the manuscript “Performance of Deep Learning pickers in routine network processing application”. The dataset includes:
- phase arrival times of P and S waves for 3 test datasets obtained using different Deep Learning pickers
- figures showing the results of the comparisons between Deep Learning pickers, Computer resources for this study have been provided by Artemisa, funded by the European Union ERDF and Comunitat Valenciana.
Additional funding has been obtained from Spanish Ministry of Science and Innovation grants CGL2017-88864-R, PID2020-114682RB-C31, and Center of Excellence Severo Ochoa accreditation to ICM-CSIC (Project CEX2019-000928-S), Peer reviewed

A recent publication by Cesca et al.1 reanalyzes and expands seismic data to identify hypocenters of observed seismicity induced by the Castor Underground Gas Storage (UGS) operations. Their results confirm those of previous studies2,3 that earthquakes occurred below the storage formation on a fault dipping opposite from the Amposta fault, which bounds the reservoir. However, two important sets of disagreements require revising the conclusions by Cesca et al.1: the depth of hypocenters and the processes leading to seismicity. Inaccurate estimates of hypocenters location and partial consideration of the physical mechanisms that induce seismicity may imply endangering future deep underground projects., V.V. acknowledges funding from the European Research Council (ERC) under the European Union’s Horizon 2020 Research and Innovation Programme through the Starting Grant GEoREST (www.georest.eu) under grant agreement No. 801809. A.V. acknowledges funding from Spanish Ministry of Science and Innovation grant CGL2017-88864-R. IDAEA-CSIC and ICM-CSIC are Centres of Excellence Severo Ochoa (Spanish Ministry of Science and Innovation Grants CEX2018-000794-S and CEX2019-000928-S, respectively, funded by MCIN/AEI/ 10.13039/501100011033), Peer reviewed

14 pages, 9 figures, 3 tables, Picking arrival times of P and S phases is a fundamental and time‐consuming task for the routine processing of seismic data acquired by permanent and temporary networks. A large number of automatic pickers have been developed, but to perform well they often require the tuning of multiple parameters to adapt them to each dataset. Despite the great advance in techniques, some problems remain, such as the difficulty to accurately pick S waves and earthquake recordings with a low signal‐to‐noise ratio. Recently, phase pickers based on deep learning (DL) have shown great potential for event identification and arrival‐time picking. However, the general adoption of these methods for the routine processing of monitoring networks has been held back by factors such as the availability of well‐documented software, computational resources, and a gap in knowledge of these methods. In this study, we evaluate recent available DL pickers for earthquake data, comparing the performance of several neural network architectures. We test the selected pickers using three datasets with different characteristics. We found that the analyzed DL pickers (generalized phase detection, PhaseNet, and EQTransformer) perform well in the three tested cases. They are very efficient at ignoring large‐amplitude transient noise and at picking S waves, a task that is often difficult even for experienced analysts. Nevertheless, the performance of the analyzed DL pickers varies widely in terms of sensitivity and false discovery rate, with some pickers missing a significant percentage of true picks and others producing a large number of false positives. There are also variations in run time between DL pickers, with some of them requiring significant resources to process large datasets. In spite of these drawbacks, we show that DL pickers can be used efficiently to process large seismic datasets and obtain results comparable or better than current standard procedures, The authors gratefully acknowledge the computer resources at Artemisa, funded by the European Union ERDF and Comunitat Valenciana as well as the technical support provided by the Instituto de Física Corpuscular, IFIC (CSIC-UV). Antonio Villaseñor acknowledges funding from Spanish Ministry of Science and Innovation MCIN/AEI/10.13039/501100011033 grants CGL2017-88864-R, PID2020-114682RB-C31, and the Severo Ochoa Center of Excellence accreditation CEX2019-000928-S to ICM-CSIC, Peer reviewed

Trabajo presentado en la 3rd International Conference on Coupled Processes in Fractured Geological Media: "Observation, Modeling, and Application" (Coufrac 2022), celebrada en Berkeley (USA) entre el 14 y el 16 de noviembre de 2022., Pressure buildup is the standard, but often insufficient, explanation for induced seismicity, , especially when it comes to delayed effects. Coupled processes may play relevant roles, so that they must be accounted for to find the mechanisms triggering induced seismicity, especially when it comes to delayed effects. We illustrate the importance of coupled processes in induced seismicity through the case of the Underground Gas Storage (UGS) project of Castor, Spain, where, cushion gas injection induced hundreds of events, with maximum magnitudes of 4.1, leading to the cancellation of the project. Gas injection lasted for 15 days, and the largest earthquakes occurred 17 days after the stop of injection. Injection overpressures had dissipated by the time of the largest events. The gas was injected at 1.7 km depth, while the induced seismicity occurred at depths ranging from 4 to 10 km. These characteristics of the induced seismicity at Castor pose a challenge on explaining its causes. Coupled processes provide a plausible explanation when combining poromechanical stresses, buoyancy, and shear-slip stress transfer. If coupled processes had been considered in the assessment of the induced seismicity at Castor, the induced earthquakes at Castor could have been anticipated and managed., V.V. acknowledges funding from the European Research Council (ERC) under the European Union’s Horizon 2020 Research and Innovation Programme through the Starting Grant GEoREST (www.georest.eu) under grant agreement No. 801809. A.V. acknowledges funding from Spanish Ministry of Science and Innovation grant CGL2017-88864-R. IDAEA-CSIC and ICM-CSIC are Centres of Excellence Severo Ochoa (Spanish Ministry of Science and Innovation Grants CEX2018-000794-S and CEX2019-000928-S, respectively, funded by MCIN/AEI/ 10.13039/501100011033)., With funding from the Spanish government through the ‘Severo Ochoa Centre of Excellence’ accreditation (CEX2018-000794-S) y (CEX2019-000928-S), Peer reviewed

European Geosciences Union (EGU) General Assembly, 23-27 May 2022, Vienna, Austria, In recent years there have been a great progress in earthquake detection and picking arrival times of P and S phases using Deep Learning algorithms. However, the general adoption of these methods for the routine processing of monitoring networks has been held back by factors such as the availability of well documented software, computational resources, and a gap in knowledge of these methods. We have analyzed recent available Deep Learning pickers, comparing the results against data picked by a human operartor and against non-Deep Learning programs. We have used data recorded in several locations, with different characteristics and triggering mechanisms, such as volcanic eruptions, induced seismicity and local eartquakes, recorded using different types of instruments. We have found that the Deep Learning algorithms are able to achieve results comparables to a human operator, and several times better than a classical program, specially in data with a low signal to noise ratio. They are very efficient at ignoring large amplitude transient noise and at picking S waves, a task that is often difficult even for experienced analysts, and they require very few parameters to tune (often only the probability threshold) so an in-depth knowledge of neural networks is not required, This research has been funded by Spanish Ministry of Science and Innovation MICINN/AEI/10.13039/501100011033 grants CGL2017-88864-R and PRE2018-084986, Peer reviewed