4960 20241203105543.0 10.1029/2020GL089931 DOI SCART-2020-0146 Lindsey, N.J. City-scale dark fiber DAS measurements of infrastructure use during the COVID-19 pandemic 2020 Throughout the recent COVID-19 pandemic, real-time measurements about shifting use of roads, hospitals, grocery stores, and other public infrastructure became vital for government decision makers. Mobile phone locations are increasingly assimilated for this purpose, but an alternative, unexplored, natively anonymous, absolute method would be to use geophysical sensing to directly measure public infrastructure usage. In this paper, we demonstrate how fiber-optic distributed acoustic sensing (DAS) connected to a telecommunication cable beneath Palo Alto, CA, successfully monitored traffic over a 2-month period, including major reductions associated with COVID-19 response. Continuous DAS recordings of over 450,000 individual vehicles were analyzed using an automatic template-matching detection algorithm based on roadbed strain. In one commuter sector, we found a 50% decrease in vehicles immediately following the order, but near Stanford Hospital, the traffic persisted. The DAS measurements correlate with mobile phone locations and urban seismic noise levels, suggesting geophysics would complement future digital city sensing systems. NO Yuan, S. Lellouch, A. Gualtieri, L. Lecocq, T. Biondi, B. Geophysical Research Letters 47 2020 16 e2020GL089931 thomas.lecocq@observatoire.be https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2020GL089931 1154286 http://publi2-as.oma.be/record/4960/files/grl61048-fig-0002-m.jpg Vehicle observations from Stanford DAS-2 Experiment. (a) Example of DAS recordings bandpassed around 3–30 Hz showing vehicle surface waves and contrasting background energy between Stanford Hospital and Sand Hill Road sections. (b) Same as (a) but bandpassed around 0.1–1 Hz to highlight the high quality geodetic strain responses of the roadbed due to vehicle loading. Individual vehicles are numbered. (c) Continuous wavelet transform applied to spatial axis of unfiltered data shown in (a) and (b) highlighting dominant frequencies of different array segments. (d) Example processed strain data from DAS channel at 3.4 km, bandpass filtered as in (b) in gray, with a model of the horizontal strain for three vehicles passing the fiber on the southbound side of the road (black line), and three STA/LTA detections (red lines) for vehicles #5, #8, #10 shown in (b). A matched template algorithm was then applied using the median of approximately 200 detected vehicle signals and scanning over the full daily time series for the DAS channel 12102 http://publi2-as.oma.be/record/4960/files/grl61048-fig-0002-m.jpg?subformat=icon-180 icon-180 Vehicle observations from Stanford DAS-2 Experiment. (a) Example of DAS recordings bandpassed around 3–30 Hz showing vehicle surface waves and contrasting background energy between Stanford Hospital and Sand Hill Road sections. (b) Same as (a) but bandpassed around 0.1–1 Hz to highlight the high quality geodetic strain responses of the roadbed due to vehicle loading. Individual vehicles are numbered. (c) Continuous wavelet transform applied to spatial axis of unfiltered data shown in (a) and (b) highlighting dominant frequencies of different array segments. (d) Example processed strain data from DAS channel at 3.4 km, bandpass filtered as in (b) in gray, with a model of the horizontal strain for three vehicles passing the fiber on the southbound side of the road (black line), and three STA/LTA detections (red lines) for vehicles #5, #8, #10 shown in (b). A matched template algorithm was then applied using the median of approximately 200 detected vehicle signals and scanning over the full daily time series for the DAS channel 14879 http://publi2-as.oma.be/record/4960/files/grl61048-fig-0002-m.gif?subformat=icon icon Vehicle observations from Stanford DAS-2 Experiment. (a) Example of DAS recordings bandpassed around 3–30 Hz showing vehicle surface waves and contrasting background energy between Stanford Hospital and Sand Hill Road sections. (b) Same as (a) but bandpassed around 0.1–1 Hz to highlight the high quality geodetic strain responses of the roadbed due to vehicle loading. Individual vehicles are numbered. (c) Continuous wavelet transform applied to spatial axis of unfiltered data shown in (a) and (b) highlighting dominant frequencies of different array segments. (d) Example processed strain data from DAS channel at 3.4 km, bandpass filtered as in (b) in gray, with a model of the horizontal strain for three vehicles passing the fiber on the southbound side of the road (black line), and three STA/LTA detections (red lines) for vehicles #5, #8, #10 shown in (b). A matched template algorithm was then applied using the median of approximately 200 detected vehicle signals and scanning over the full daily time series for the DAS channel published in REFERD