Urban Seismology of a Popular Road Race Using Distributed Acoustic Sensing

Distributed Acoustic Sensing (DAS) has emerged as a powerful tool for monitoring human-induced seismic signals in urban environments, enabling dense, meter-scale observations of dynamic sources. Building on previous studies demonstrating the capability of DAS to image large public events, such as parades and other mass-participation activities, we present a novel experiment in which two different DAS technologies (ΦOTDR and Chirped-Pulse ΦOTDR) were simultaneously deployed to record a popular pedestrian road race held in the surroundings of the University of Zaragoza (Spain).

The experiment took advantage of an already deployed optical-fiber installation with a total effective length of approximately 2 km. The fiber layout captured three distinct geometrical configurations with respect to the race course: (1) a straight section coincident with the runners’ trajectory over the last 300 m of the first kilometer (outbound leg), (2) the same straight section during the return at kilometer 4 (inbound leg), and (3) a perpendicular crossing of the fiber with the race course at the finish line. This geometry provides a unique opportunity to analyze runner-induced ground vibrations under varying crowd densities, running speeds, and fiber–source orientations.

Waterfall representations of the strain-rate data reveal clear, coherent signatures associated with individual runners and runner groups in both DAS systems. Along the straight section, the outbound leg exhibits a compact, high-amplitude wavefield characterized by closely spaced, overlapping runner traces, consistent with the tightly packed peloton early in the race. In contrast, the inbound leg shows a markedly more dispersed pattern, reflecting the progressive spreading of participants according to performance and fatigue. These differences are consistently observed in both phase-based and chirped-pulse DAS data, although with distinct signal-to-noise characteristics across different frequency bands.

At the finish line, where the fiber crosses the race course perpendicularly, the DAS records provide exceptional temporal resolution of runner arrivals. The first five finishers are individually and unambiguously identified, with isolated signatures that can be robustly matched to official arrival times. This demonstrates the potential of DAS not only for bulk crowd characterization but also for resolving individual human-induced seismic sources in real-world conditions.

Our results highlight the complementarity of DAS technologies for urban seismology applications. The experiment underscores the sensitivity of DAS to subtle variations in crowd dynamics and source geometry and illustrates its potential for non-intrusive monitoring of mass-participation events, pedestrian flows, and urban activity. These observations contribute to the growing field of anthropogenic seismology and reinforce the role of optical fiber sensing as a scalable tool for high-resolution monitoring of human activity in cities.