Distributed Acoustic Sensing as a Tool for Geotechnical Monitoring Using Seismic Waves Generated by Train Traffic in Northern Brazil

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.