George Sand França

Earthquake frequency–magnitude distributions in subduction zones provide fundamental constraints on the physical processes governing rupture nucleation and propagation. Although the classical Gutenberg–Richter law offers an approximate description of these data, systematic deviations have motivated the use of generalized statistical models grounded in non-extensive statistical mechanics and fragmentation frameworks. Here, we perform a Bayesian comparative analysis of four models applied to twelve subduction zones along the circum-Pacific belt, using National Earthquake Information Center catalogs with and without correction for magnitude of completeness. We show that catalog incompleteness exerts a first-order control on both model ranking and inferred parameters. After completeness correction, fragmentation-based models exhibit superior stability in describing moderate-to-large magnitude seismicity, highlighting the importance of accounting for detection biases in comparative statistical seismology.

The structural assessment of extended railway systems presents persistent challenges for operators responsible for maintaining safety and serviceability across hundreds of kilometers of track and associated infrastructure. Conventional inspection methods offer limited spatial coverage and are poorly suited to detecting geotechnical anomalies at early stages, particularly in complex geological settings. Distributed Acoustic Sensing addresses these limitations by converting optical fiber cables into dense seismic arrays capable of recording ground motion at thousands of points simultaneously, enabling continuous and spatially comprehensive monitoring along entire railway corridors, including the potential reuse of existing telecommunication fiber networks already installed alongside many rail lines.This study was carried out along one of the longest and most operationally demanding freight railways in northern Brazil, a corridor dedicated to the transport of iron ore in large volumes and to passenger service over extensive distances. Much of the existing distributed sensing literature has focused on high-speed passenger trains as seismic sources. Here, the operative trains are ore freight consists reaching lengths on the order of 3.3 km traveling at speeds between approximately 15 and 30 km/h, alongside shorter maintenance trains, representing a substantially different excitation regime that requires adapted analysis strategies.An optical fiber approximately one kilometer long was installed parallel to an active section of the railway. Data were acquired during multiple train passages under normal service conditions, capturing the seismic wavefield generated by each transit. Spectral analysis of the recorded signals revealed meaningful variations in the frequency content and energy distribution of surface waves along the fiber, providing evidence of lateral heterogeneities in the mechanical properties of the foundation materials.The monitored section encompasses a tunnel where geotechnical conditions are of particular concern. Within the tunnel, sandstone and banded iron formation units occur in direct contact alongside basalt dikes, forming a geologically heterogeneous rock mass with quality classifications ranging. The spatial mapping of subsurface stiffness derived from this dataset proved capable of distinguishing zones where support conditions differed from those observed in adjacent sections, pointing to areas of potential concern for long term performance. These observations align with geotechnical expectations regarding the role of differential mechanical behavior across geological contacts in the progressive degradation of tunnel foundations.Beyond the technical findings, this work contributes to the broader discussion of how fiber optic sensing networks can be integrated into monitoring programs for critical railway infrastructure. The scalability of distributed sensing, the relatively low operational cost of using trains in regular service as seismic sources, and the potential for nearly continuous data acquisition position this approach as a practical option for geotechnical surveillance in complex geological environments. Remaining challenges related to signal coupling and wavefield interpretation in heterogeneous rock masses are discussed in the context of future system development.

The South Atlantic Ocean and Margin (SAOM) serves as a crucial natural laboratory for studying passive continental margin formation, continental breakup, and the interactions between tectonic, magmatic, and sedimentary processes. This project investigates with DAS and short perido seimographs stations the tectonic evolution and magmatism of SAOM, with particular focus on seaward-dipping reflectors (SDRs), volcanic formations, and the continent-ocean transition zones. The primary objective is to generate high-resolution structural models of SAOM through an integrated approach combining advanced techniques including surface wave tomography, anisotropic analysis, receiver functions, reflected phase analysis, and joint inversion. This innovative methodology will enable precise mapping of subsurface architecture, particularly the crucial crustal transition zone. SDRs, which record volcano-magmatic processes during rifting, will be investigated as key markers to reconstruct the region's magmatic history, including its relationship to crustal thinning, oceanic crust formation, and tectonic activity. This project will examine the role of magmatism in SAOM evolution, analyzing its influence on fault development, fluid migration, and resource accumulation. Spatio-temporal analysis of these processes, from initial rifting through oceanic spreading, will provide new insights into passive margin dynamics. We will present how the project will be carried out and the first steps to be implemented.

Developed after the breakup of Gondwana in the Early Cretaceous, the southeastern margin of Brazil comprises a large, economically important hydrocarbon province that includes two of the world’s most prolific offshore sedimentary basins: Santos and Campos. At the same time, this passive margin hosts the country's main offshore seismic zone, characterised by frequent small-magnitude earthquakes and occasional larger events (M ≥ 4.8), occurring on average every 20-25 years, mostly within the extended and submerged continental crust. The larger events include the 1955 Vitória earthquake (mb 6.1), the 1939 Tubarão earthquake (mb 6.0), the 1972 Campos earthquake (M 4.8), the 1990 Porto Alegre earthquake (mR 5.2), and the 2008 São Vicente earthquake (mR 5.2). Due to insufficient instrumentation, limited to a few distant and unevenly distributed coastal stations of the Brazilian Seismographic Network (RSBR), and the low frequency of larger events, both the origin of this seismic activity and the risk it poses to key offshore infrastructure remain poorly understood. To better understand the processes related to the seismic activity in southeastern Brazil, the Brazilian Seismographic Network at the Sea Project (RSBR-Mar) intends to improve coverage by deploying: i - six temporary broadband coastal stations; ii - five Ocean Bottom Seismometers (OBSs); and iii - eight Mobile Earthquake Recorder in Marine Areas by Independent Divers (MERMAIDs) between São Paulo and Espírito Santo states. To date, two temporary land stations have been installed in Espírito Santo in June 2025. During a cruise mission between late September and early October 2025, we deployed all five OBSs and eight MERMAIDs. The OBSs will be recovered by September 2026, after one year of data acquisition, at which time the batteries will be replaced and the OBSs redeployed for an additional year. MERMAIDs periodically surface and establish satellite communication, allowing them to send important data back to us, collected during each operational cycle. To date, we have collected 50 waveforms detected by the MERMAIDs. Although MERMAIDs were originally designed to record high-quality P-waveforms for teleseismic tomography, rather than to detect small local events, we will explore this possibility by requesting data stored in their one-year internal storage. If successful, MERMAIDs could improve the localisation of earthquakes recorded in continuous records from land stations and OBSs. First, we plan to automatically detect and locate earthquakes in continuous seismograms from land stations and OBSs using machine-learning-based algorithms (e.g., LOC-FLOW, MALMI). These preliminary locations allow us to identify which data windows to request for each MERMAID. Then, we can relocate the small earthquakes with the acoustic data. Finally, relocation methods may help us to delineate possible seismogenic offshore structures (e.g., HypoDD).

The horizontal-to-vertical spectral ratio (HVSR) is a passive seismic method that uses ambient noise to estimate impedance contrasts at shallow depths, such as sedimentary layers and faults. Its resolution depends on impedance contrast and geological context, decreasing with depth owing to high-frequency attenuation. We conducted 58 measurements along five profiles in the Água Bonita Basin, a strike-slip basin in central–north Brazil, using short-period and broadband seismometers. The basin includes the Vereda Verde and Água Bonita formations, deposited during the Ordovician–Silurian after reactivation of the Transbrasiliano Lineament. HVSR curves were calculated and inverted under the assumption of a diffuse field to constrain the basin structure. Additionally, continuous seismic data from a broadband station were analysed with receiver functions, complemented by topographic and magnetic data. HVSR results indicate a maximum depth of 520 m, and receiver functions yield 435 m for the sedimentary package. Both estimates are shallower than previous models. A 3D geological model suggests two opposing half-graben formed by transtensive sinistral reactivation of the Transbrasiliano Lineament. The Água Bonita Basin thus serves as an analogue for early Phanerozoic basin formation in Brazil and may represent a failed rift stage in the Early Ordovician.

Being an important economic and industrial hub, Arequipa is the second most populous city in Peru, with over one million inhabitants. Within its metropolitan area, high-amplitude vibrations are generated by heavy construction machinery as well as by urban and commercial transportation. These vibrations pose a potential risk of damage to both urban and archaeological infrastructures. To assess possible structural impacts, vibration measurements were collected at significant sites and analyzed according to the DIN 4150- 3 standard. Ground vibrations attenuation was evaluated using the Levenberg-Marquardt inversion method. Regression adjustment models were developed, and the analysis of geometric and material damping coefficients was automated with the application of neural network algorithms, in areas with lower data density. The results indicate that vibrations caused by heavy machinery can damage sensitive or residential structures, as the peak particle velocity (PPV) exceeded 20 mm/s in these cases. However, for sources such as vehicular traffic on bridges and railway transport, the PPV only exceeded 10 mm/s. Attenuation analysis revealed that maximum particle velocity decreases with increasing distance from the source, and that the strongest damping occurs in the vertical component. Geometric damping (γ) values ranged from 0.1 to 1.65, and material damping (α) values ranged from 0.001 to 0.03 for the various sources evaluated. In cohesive soils, wave propagation is faster, which reduces tension in the foundations, whereas in sandy-silty soils, slower propagation increases the tension transmitted to the foundations. This research also aims to lay the foundation for the development of Peruvian regulations framework to assess structural damage caused by anthropogenic vibrations, since there are no existing regulations governing these effects on buildings. This study can help propose guidelines for the city of Arequipa.

In academia, science often seems disconnected from art, creating knowledge that is isolated or distant from society. However, this connection is essential and can be fostered through various artistic forms, including the figure of the clown. Since 2019, through an action-research project, we have been exploring the integration of the clown’s universe with that of science, enabling what we call a “back-and-forth” dynamic. This term reflects the bidirectional nature of the interaction: society benefits from receiving presentations that convey scientific ideas in an accessible way, while the artist-scientists on stage experience the challenge and practice of sharing knowledge in a playful, unpretentious manner rooted in fun. In this work, we present a diagnosis and consolidation of this initiative through the performance “A Journey to the Center of the Earth,” which involves researchers, undergraduate and graduate students, high school teachers, and staff, demonstrating how the union of art and science can enrich both academia and society.

The University of São Paulo (USP) initiated its international collaborations in seismology during the 1980s by assisting in the installation of the Geoscope very-broadband SPB station in Brazil. This marked the beginning of a sustained effort to establish and operate seismic stations under international partnerships. In 2024, an additional Geoscope station was installed in western Brazil, further enhancing very-broadband seismic coverage across South America.The Brazilian Seismographic Network (RSBR), established in 2010, now encompasses over 100 permanent broadband stations managed by four universities: USP (BL network), the National Observatory (ON network), the University of Brasília (BR network), and the Federal University of Rio Grande do Norte (NB network). RSBR plays a critical role in monitoring seismicity across Brazil, offering real-time open data (via Seedlink) and on-demand access (via FDSNws). Its impact is evident in the Brazilian seismic catalog, which recorded 820 earthquakes between 1900 and 2010. From 2010 to 2024, this number increased to over 2,800 events. The BL network alone serves more than 70 institutions in real-time and shares terabytes of data with researchers worldwide.Another noteworthy initiative is the Vale Seismographic Network, a public-private partnership between USP and Vale S.A. This network monitors local seismicity in the mining regions of the Quadrilátero Ferrífero (southeastern Brazil) and Carajás (northern Brazil) using 11 broadband stations. Data from these networks is openly available for regional and teleseismic events outside mining zones.In addition to permanent networks, USP’s Seismology Center deploys temporary stations to address induced seismicity in regions such as Taquaritinga/SP, Sales Oliveira/SP, and Gramado/RS. Following the M 5.7 intraplate earthquake on January 31, 2022, in Guyana, USP partnered with the Guyana Geological Commission to install four stations in Guyana's Deep South.Managing the extensive database of these networks poses significant challenges. Reliable metadata and synchronized waveform archives are maintained, even for temporary deployments that require frequent equipment updates. USP hosts South America's only FDSN server for waveform and metadata dissemination, ensuring open data access wherever feasible. Arrival times from all USP networks are regularly contributed to the International Seismological Centre (ISC) for inclusion in its revised bulletin.Despite operational and maintenance challenges, USP remains committed to fostering collaborations with Brazilian and neighboring institutions to promote the sustainable development of permanent, independent, and open seismic networks.

The art of clowning, an ancient practice, reveals remarkable potential to be explored in scientific outreach. In 2020, França et al. shared their experience integrating clowning into a show about scientific denial, which led to the creation of an innovative scientific workshop spanning two weeks. The aim is to reach high school teachers, undergraduate and graduate students, intending to disseminate geosciences unconventionally. In this context, we will present the experience in detail, highlighting the adopted approach, observed impacts, and the target audience we aim to reach. Additionally, we will explore strategies to expand this project and make it accessible to various locations. It is worth mentioning that during the pandemic, we ventured into our first foray into scientific outreach training through the scientific theater workshop. Despite being virtual, this hands-on approach proved surprisingly effective (link to the video: https://youtu.be/q0dkz8wtlxw?si=K63xn8XvzeVux3ri). Now, we will share the practical results obtained in the in-person and clown-focused format. We hope this initiative inspires new forms of scientific communication and contributes to the innovative and engaging promotion of geosciences. França et al., 2020, Geoscience Communication - https://doi.org/10.5194/gc-4-297-202