Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Feb 4;9(1):1328.
doi: 10.1038/s41598-018-36675-8.

Distributed Acoustic Sensing Using Dark Fiber for Near-Surface Characterization and Broadband Seismic Event Detection

Affiliations
Free PMC article

Distributed Acoustic Sensing Using Dark Fiber for Near-Surface Characterization and Broadband Seismic Event Detection

Jonathan B Ajo-Franklin et al. Sci Rep. .
Free PMC article

Abstract

We present one of the first case studies demonstrating the use of distributed acoustic sensing deployed on regional unlit fiber-optic telecommunication infrastructure (dark fiber) for broadband seismic monitoring of both near-surface soil properties and earthquake seismology. We recorded 7 months of passive seismic data on a 27 km section of dark fiber stretching from West Sacramento, CA to Woodland, CA, densely sampled at 2 m spacing. This dataset was processed to extract surface wave velocity information using ambient noise interferometry techniques; the resulting VS profiles were used to map both shallow structural profiles and groundwater depth, thus demonstrating that basin-scale variations in hydrological state could be resolved using this technique. The same array was utilized for detection of regional and teleseismic earthquakes and evaluated for long period response using records from the M8.1 Chiapas, Mexico 2017, Sep 8th event. The combination of these two sets of observations conclusively demonstrates that regionally extensive fiber-optic networks can effectively be utilized for a host of geoscience observation tasks at a combination of scale and resolution previously inaccessible.

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1
Figure 1
Map of a section of the ESNet Dark Fiber Testbed (https://www.es.net/network-r-and-d/experimental-network-testbeds/100g-sdn-testbed/terms-and-conditions/). (A) The regional network within CA and western NV; zone of panel (B) shown in black dashed box. (B) The subsection of the network used in this study. The red segment in (B) is the area of focus for ambient noise analysis; W1 and W2 are reference wells for water table and soil horizons, respectively. The study fiber (blue) is aproximately co-linear with an active rail line. Dashed green line labeled I-5 is Interstate 5, a major source of ambient noise beyond the rail corridor.
Figure 2
Figure 2
Illustration of data processing workflow for ambient noise interferometry. (a) Example of train noise shown via an 8 second time domain slice. The red box in (a) highlights subsection of the array used for (b) noise correlation gather, (c) dispersion analysis, and (d) inversion of the shear-wave velocity (VS) profiles. Black and white markers in (c) denote observed and model-predicted multimodal dispersion curves respectively. In (d), the yellow dashed lines denote upper and lower bounds of the parameter space used in Monte Carlo sampling; the bold red line marks the best-fit VS profile; the yellow/blue lines denote the top 0.1% best-fitting VS profiles (color coded by their corresponding inversion misfits); Misfit denotes normalized misfit values (min-max normalized by misfits of the top 0.1% VS profiles).
Figure 3
Figure 3
Shear-wave velocity (Vs) inversion results and ground truth comparison. (a) Depths of groundwater levels (GWL) (upper) and Vs30 estimates (lower) extracted from the surface wave inversion results. (b) Pseudocolor display of Vs profiles and comparisons against well data. In (b), W1 and W2 mark surface locations of the two reference wells shown in Fig. 1; the blue dashed line denotes the depth of the ground water level provided by water well W1; the black dashed lines correspond to lithological horizons obtained from geotechnical well W2.
Figure 4
Figure 4
Time-lapse repeatability demonstration of ambient noise analysis in (a) space-time domain, (b) frequency-velocity domain, and (c) in terms of groundwater levels obtained from the model domain. Color sequence of red, green, and blue denote chronological orders of the monitoring period. In (c), the median, min-max range, and percentiles are calculated based upon all the topmost best-fitting models associated with the monitoring period.
Figure 5
Figure 5
Example earthquakes recorded by the Sacramento Dark Fiber DAS array. The recorded data are plotted as strain-rate after multiplying by the gauge length (10 m) to convert to units proportional to velocity (1e-6 m/s), and have been averaged over 100 m of linear fiber length (50 traces) and then bandpass filtered in the 0.1–0.4 Hz range for regional events, and 0.01–0.1 Hz for teleseisms. Events are sorted by increasing epicentral distance from Sacramento. Earthquake amplitudes for the Peru and Honduras events are scaled by the factors in parentheses.
Figure 6
Figure 6
Teleseismic DAS recording of the M8.1 Chiapas, Mexico 2017-Sep-08 earthquake. (a) Seismic data for [black trace] one location and [red and blue] all locations from 0.0–7.6 km at a 2 m spacing (4001 traces total); top right inset shows surface waves arriving at the [black] south and [pink] north end locations of the array (backazimuth 120°), bottom left inset shows body waves arriving coincidently at both locations. A two-corner, zerophase, f = 0.01–0.5 Hz bandpass filter was applied. (b) Stacking 400 m or 200 consecutive DAS channels, color-coded by the bandpass filter applied to emphasize the broadband observation (1–100 seconds). Gray background traces show the single trace recording for cases that make a significant difference. Each of the traces is normalized to peak amplitude.
Figure 7
Figure 7
(a) Locations and focal mechanisms of the M4.2 2018-Jan-18 Geysers (red) and M4.4 2018-Jan-04 Berkeley (blue) earthquakes, which occurred approximately 100 km from the Sacramento Dark Fiber DAS array (black line). (b,c) Raw and lowpass filtered DAS strain-rate waveforms for these events averaged over 100 m (50 channels) at the yellow circle position shown in (a) (channel 4975 +/− 50 channels). Note the similarity between seismic and non-seismic signal amplitudes and the differences in frequency content.
Figure 8
Figure 8
(a) Illustration of different installation geometries. (b) Earthquake (M4.2 Geysers 2018-Jan-18) trace comparison for each installation mode at Sacramento – trenched conduit (green), cased conduit (blue), attached conduit (red); strain-rate data are stacked over 100 m and filtered (BP 0.5–2 Hz n 4 p 2). (c) Normalized Fourier amplitude spectra for the waveforms shown in b.

Similar articles

See all similar articles

Cited by 2 articles

References

    1. Quesney A, et al. Estimation of watershed soil moisture index from ERS/SAR data. Remote sensing of environment. 2000;72:290–303. doi: 10.1016/S0034-4257(99)00102-9. - DOI
    1. Le Hégarat-Mascle S, Zribi M, Alem F, Weisse A, Loumagne C. Soil moisture estimation from ERS/SAR data: Toward an operational methodology. IEEE Transactions on Geoscience and Remote Sensing. 2002;40:2647–2658. doi: 10.1109/TGRS.2002.806994. - DOI
    1. Chaussard E, et al. Remote sensing of ground deformation for monitoring groundwater management practices: Application to the Santa Clara Valley during the 2012–2015 California drought. Journal of Geophysical Research: Solid Earth. 2017;122:8566–8582.
    1. Wahr, J., Swenson, S., Zlotnicki, V. & Velicogna, I. Time-variable gravity from GRACE: First results. Geophysical Research Letters31(2004).
    1. Xiao M, et al. How much groundwater did California’s central valley lose during the 2012–2016 drought? Geophysical Research Letters. 2017;44:4872–4879. doi: 10.1002/2017GL073333. - DOI
Feedback