Geophysical Research Letters

RESEARCH LETTER Vertically Distributed Sensing of Deformation Using Fiber

10.1029/2018GL080428
Optic Sensing
Key Points: Cheng-Cheng Zhang1,2 , Bin Shi1 , Kai Gu1, Su-Ping Liu1 , Jing-Hong Wu3, Song Zhang1,
• Distributed fiber optic sensing with
Brillouin scattering provides a clear Lei Zhang1, Hong-Tao Jiang4, and Guang-Qing Wei5
subsurface strain profile using a 1
single fiber optic cable School of Earth Sciences and Engineering, Nanjing University, Nanjing, China, 2Department of Civil and Environmental
• We find a threshold in confining Engineering, University of California, Berkeley, CA, USA, 3School of Civil Engineering, Suzhou University of Science and
pressure to achieve a strong fiber Technology, Suzhou, China, 4School of Geographic and Oceanographic Sciences, Nanjing University, Nanjing, China,
optic cable-soil coupling 5
Suzhou NanZee Sensing Technology Co., Ltd., Suzhou, China
• Distributed fiber optic sensing can
complement with existing
techniques for improved monitoring
of vertical deformation Abstract Vertical deformation can be revealed by various techniques such as precise leveling, satellite
imagery, and extensometry. Despite considerable effort, recording detailed subsurface deformation using
traditional extensometers remains challenging when attempting to detect localized deformation. Here we
Supporting Information:
• Supporting Information S1 introduce distributed fiber optic sensing based on Brillouin scattering as a geophysical exploration method
for imaging distributed profiles of vertical deformation. By examining fiber optic cable-soil interaction we
Correspondence to: found a threshold in confining pressure to achieve a strong cable-soil coupling, thus validating data
B. Shi,
collected from a borehole-embedded fiber optic cable deployed in Shengze, southern Yangtze Delta, China.
[email protected]
Clear-cut strain profiles acquired from November 2014 to December 2016 allowed us to pinpoint where
compaction or rebound was actively occurring and examine strain responses at various locations along the
Citation:
entire cable length. We suggest that distributed fiber optic sensing can complement with extensometry and
Zhang, C.-C., Shi, B., Gu, K., Liu, S.-P., Wu,
J.-H., Zhang, S., et al. (2018). Vertically remote sensing techniques for improved monitoring of vertical deformation.
distributed sensing of deformation
using fiber optic sensing. Geophysical Plain Language Summary Recording detailed subsurface deformation using traditional methods
Research Letters, 45, 11,732–11,741. (e.g., extensometers) is sometimes difficult due to limited measuring points. This dilemma may be
https://doi.org/10.1029/2018GL080428
overcome by using the emerging distributed fiber optic sensing technology, which transforms common
telecommunication fiber optic cables into sensors capable of making distributed strain measurements. We
Received 11 SEP 2018
Accepted 23 OCT 2018 report the use of this technology for monitoring distributions of vertical deformation resulting from
Accepted article online 29 OCT 2018 groundwater abstraction in Shengze, southern Yangtze Delta, China. An evaluation of the performance of a
Published online 13 NOV 2018
borehole-embedded fiber optic cable helps us to validate the in situ strain data. The advantage of using this
technology for vertical deformation sensing is the ability to locate any strata undergoing compaction or
rebound and look at strain responses at any depth of interest. Moreover, recording subsurface changes in this
fashion may also be useful in other geophysical and engineering applications that require refined monitoring
of the media.

1. Introduction
Ground deformation often accompanies a host of geological and anthropogenic processes such as earth-
quakes, volcanic activities, and subsurface mining and fluid withdrawal. Especially, overexploitation of
groundwater can generate land subsidence (Galloway & Burbey, 2011; Poland & Davis, 1969), which is
increasingly a problem in many areas of the world (Erban et al., 2013; Famiglietti, 2014; Higgins et al.,
2013; Ortiz-Zamora & Ortega-Guerrero, 2010; Shi et al., 2008; Shirzaei & Bürgmann, 2018). Vertical deforma-
tion is currently measured by various techniques such as precise leveling, interferometric synthetic aperture
radar (InSAR), global navigation satellite system (GNSS), and extensometry (Galloway & Burbey, 2011).
Satellite imagery and space geodetic techniques measuring land-surface elevation changes have allowed
for mapping and monitoring detailed surface deformation over large areas (Amelung et al., 1999;
Massonnet & Feigl, 1998; Shirzaei & Bürgmann, 2018), whereas subsurface strata deformation is often iden-
tified and characterized using extensometers (Hwang et al., 2008; Metzger et al., 2001; Riley, 1969). Despite
considerable effort, however, recording detailed distributions of vertical deformation using traditional
©2018. American Geophysical Union. extensometers is sometimes challenging, mainly due to the inadequate information provided by spatially
All Rights Reserved. discrete measuring points.

ZHANG ET AL. 11,732

 19448007, 2018, 21, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018GL080428 by CAPES, Wiley Online Library on [19/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters 10.1029/2018GL080428

This spatial undersampling problem may be overcome by using the emerging distributed fiber optic sen-
sing (DFOS) technology, which allows for distributed strain measurements along a common fiber optic
(FO) cable (Bao & Chen, 2012; Habel & Krebber, 2011; Schenato, 2017). DFOS with Raman scattering has
already been known as distributed temperature sensing and utilized for a host of hydrological and geother-
mal applications (Briggs et al., 2012; Carlino et al., 2016; Curtis & Kyle, 2011; Selker et al., 2006). Recently, by
exploiting changes in Brillouin and Rayleigh scattering induced by external strains, the DFOS technology
has been utilized for geohazard sensing including earthquake observations (Dou et al., 2017; Jousset
et al., 2018; Lindsey et al., 2017) and landslide detection (Huntley et al., 2014; Lienhart, 2015; Michlmayr
et al., 2017; Picarelli et al., 2015; Schenato et al., 2017). While a few studies have explored the feasibility
of DFOS for subsidence and strata deformation sensing (Murai et al., 2013; Wu et al., 2015), the understand-
ing of data collected from borehole-embedded FO cables has remained elusive and hence precluded its
use in many contexts.
In this study, we demonstrate the utilization of DFOS based on Brillouin scattering for measurement of dis-
tributed vertical deformation in Shengze, southern Yangtze Delta, China. By examining FO cable-soil interac-
tion we find a threshold in confining pressure to achieve a strong FO cable-soil coupling, which validates
distributed strain measurements collected from a borehole-embedded FO cable. Data acquired from
November 2014 to December 2016 allow us to obtain a complete subsurface profile and characterize the ver-
tical deformation in detail.

2. Methods
2.1. DFOS Based on Brillouin Scattering
DFOS is a technology that enables spatially continuous, long-distance, and near-real-time measurements
along a FO cable. The FO cable is both the transmission medium and the sensing element, so no extra sensors
are needed in the optical path. When a light wave generated from a FO analyzer travels through the core of
an optical fiber, backscattered lights (Rayleigh, Raman, or Brillouin) are generated at any point along the fiber,
as a result of the interaction between the incident light and impurities of the fiber core (Habel & Krebber,
2011). The exploitation of spontaneous and stimulated Brillouin scattering for distributed sensing resulted
in the invention of Brillouin optical time domain reflectometry (BOTDR) and Brillouin optical time domain
analysis (BOTDA), respectively (Bao & Chen, 2012). Both technologies leverage the linearity between the
Brillouin scattering shift and the change of mechanical strain (either extensional or compressional) or tem-
perature, which is given by the following (Horiguchi et al., 1989):

∂v B ∂v B
Δv B ¼ Δε þ ΔT (1)
∂ε ∂T

where ΔvB is the change in Brillouin frequency shift; Δε and ΔT are the changes in strain and temperature,
respectively; and ∂vB/∂ε and ∂vB/∂T are the coefficients for strain and temperature changes, respectively.
Note that all strains discussed in this letter are axial strains if not otherwise specified. According to the above
correlation (equation (1)), the strain and temperature changes along the entire length of the fiber can be
determined provided that the change in Brillouin frequency shift is recorded. Further information about
the fundamentals of the DFOS technology may be found in Habel and Krebber (2011), Bao and Chen
(2012), and Schenato (2017).
The case study presented in this letter used a N8511 BOTDR analyzer (Advantest, Tokyo, Japan) for inter-
rogation of distributed sensors. The coefficients for strain and temperature changes were measured via
laboratory tensile tests, and their values were 0.0493 MHz/με and 1.43 MHz/°C, respectively. The analyzer
provides a minimum readout resolution of 0.05 m with a spatial resolution of 1 m and a noise level of
40 με, allowing for distributed measurements along an optical fiber of 1 km. These settings can be chan-
ged to realize distributed sensing up to 80 km. For the laboratory testing of FO cable-soil interfacial cou-
pling, a NeubreScope NBX-6050A BOTDA analyzer (Neubrex, Kobe, Japan) was utilized. The readout
resolution and spatial resolution were set to 0.05 and 0.1 m, respectively; the resulting measurement accu-
racy was ±7.5 με. The coefficients for strain and temperature changes were determined to be 0.0497 MHz/
με and 1.07 MHz/°C, respectively.

ZHANG ET AL. 11,733

 19448007, 2018, 21, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018GL080428 by CAPES, Wiley Online Library on [19/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters 10.1029/2018GL080428

2.2. Testing of FO Cable-Soil Coupling

First, we sought to test the interfacial coupling between FO cable and soil for the successful deployment
of a DFOS system for vertical deformation sensing. We devised a displacement-controlled pullout appara-
tus to investigate the cable-soil interaction mechanism (Figure S1). The apparatus consists essentially of
three parts: (a) a pressure chamber, in which the oil pressure is directly applied to a specimen up to
20 MPa by controlling the hydraulic pump; (b) a tensile tester, through which the pullout displacement
is automatically applied to a FO cable run from the specimen, whereas the pullout force is measured
using a force gauge (±0.1 N); and (c) a FO analyzer, through which the distribution of strain along the
FO cable is recorded during testing. The advantage over a former apparatus (Zhang et al., 2016) lies in
the fact that the cable-soil interaction can be studied at a large scale, along with the convenience of
applying uniform confining pressures using a hydraulic oil. Because the applied confining pressure is per-
pendicular to the cable axis, results obtained from this test can simulate the behavior of a cable buried in
a vertical borehole.
Tests were carried out on FO cables embedded in saturated clayey and sandy soils (Text S1) at confining pres-
sures (σ c) up to 1.6 MPa using the devised apparatus. The tested FO cables were 2-mm-diameter tight-buffed
strain cables (Nanzee Sensing, Suzhou, China; Text S2 and Figure S2). In each test, a FO cable was embedded
in a 60-mm-diameter heat-shrink tube, with the embedment cable length being 1 m. A soil was then com-
pacted in the tube using a miniature hammer (Figure S3). After preparation, the specimen was installed
within the pressure chamber and the FO cable was pulled out at a velocity of 0.1 mm/s. The tensile tester
was temporarily stopped under each displacement step of 0.97 mm, thereby permitting the measurement
of strain distribution along the FO cable using the analyzer. Each measurement took approximately 67 s
under the current measurement setting. Each test was terminated either when the FO cable was completely
pulled out (failure of the cable-soil interface) or when the strain at the cable head reached approximately
15,000 με. Note that temperature compensation for measured strains was not considered as the variation
of temperature (~0.38 °C) was negligible during testing. Tensile forces calculated from measured strains
agreed quite well with pullout forces measured from the force gauge (Figure S4), hence validating the mea-
sured strain data and suitability of the test setup and procedure.

2.3. Vertically Distributed Sensing of Deformation in Shengze

Next, we deployed the DFOS technology to monitor distributed vertical deformation in Shengze, Suzhou
(southern Yangtze Delta, China). Figure 1a shows the total subsidence of the southern Yangtze Delta over
the period 1956–1999; this area is one of the most developed areas in China and has been directly affected
by land subsidence primarily due to groundwater abstraction (Shi et al., 2008; Xue et al., 2005). Particular
attention has been given to Shengze because it subsided at an abnormal rate (~36 mm/year during
2004–08) in contrast with downtown Suzhou (<5 mm/year; Shi et al., 2014). Although subsidence rate has
been largely reduced due to measures implemented to restrict groundwater use in this area since 2000
(Wu et al., 2018), knowledge of a detailed vertical strain profile would be helpful to pinpoint where compac-
tion is still actively occurring. The overall hydrogeology is described in Text S3 and Table S1.
Figure 1b shows a schematic representation of the DFOS system based on BOTDR. The basic principle is that
the compaction (or rebound) of strata will exert on the vertically buried FO cable a compressional strain
(respectively, extensional strain), which will induce a change in the Brillouin frequency shift detectable using
a FO analyzer. Once the strain profile is obtained using equation (1), the amount of compaction or rebound
can be estimated for any section of interest by direct integration of the spatially continuous strain data along
that section. In particular, the vertical surface displacement (i.e., the total amount of subsidence or uplift) can
be obtained by integrating the measured strain along the entire borehole length, provided that strata defor-
mation is negligible below the embedment depth of the cable.
The field test was performed within the campus of Shengze Middle School. First, a vertical borehole
(30°530 23.96″N, 120°400 47.11″E) with a depth of 200 m and a diameter of 129 mm was drilled. The drilled core
samples indicated that the local lithology consists almost entirely of Quaternary deposits alternating sandy,
silty, and clayey soils (Table S1). Once the borehole was dug, a FO strain cable was attached to a cone guide,
which was lowered into the borehole by controlling two wire ropes. In doing this, the FO cable was run along
the entire depth of the borehole. Afterward, the borehole was backfilled with a sand-clay mixture, together

ZHANG ET AL. 11,734

 19448007, 2018, 21, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018GL080428 by CAPES, Wiley Online Library on [19/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters 10.1029/2018GL080428

Figure 1. (a) A map showing the total subsidence of the southern Yangtze Delta, China (modified from Zhang & Wei, 2005).
Also marked is the position of a borehole instrumented with a distributed fiber optic sensing system for vertically distrib-
uted sensing of deformation in Shengze. The overall hydrogeology is described in Text S3 and Table S1. (b) A schematic
representation of vertically distributed sensing of deformation using distributed fiber optic sensing based on Brillouin
scattering. In the case of groundwater-related vertical deformation, either compressional strain (due to strata compaction)
or extensional strain (due to strata rebound) exerts on the vertically buried fiber optic (FO) cable will induce a change in
the Brillouin frequency shift of a backscattered light in the FO cable. This can be detected and interrogated using a FO
analyzer set in a nearby monitoring station. The analyzer and the borehole-embedded vertical cable are connected using a
horizontal cable, which however is not utilized for deformation sensing. The actual deformation in millimeters can be
estimated by direct integration of spatially continuous strain data (see section 2 for details). Note that the presented
scheme is a single-ended deployment based on Brillouin optical time domain reflectometry. The scheme is slightly dif-
ferent for Brillouin optical time domain analysis, which requires an additional probing light at the other end of the cable
and hence a loop deployment.

with some gravels. Note that the cable was kept tensioned throughout the whole process. In addition, a fiber
Bragg grating osmometer was also deployed at 87.7-m depth to monitor pore fluid pressures. FO
temperature cables were, however, not installed in the borehole in this particular case. Finally, the cables
ran from the top of the borehole to a FO analyzer set in a nearby monitoring station. Note that the
horizontal cable was not utilized for deformation sensing; it simply connected the analyzer with the
borehole-embedded vertical cable.

ZHANG ET AL. 11,735

 19448007, 2018, 21, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018GL080428 by CAPES, Wiley Online Library on [19/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters 10.1029/2018GL080428

The first measurement was performed in December 2012, one month after borehole backfilling. Details of
field instrumentation and first results acquired between December 2012 and November 2014 were reported
by Wu et al. (2015). Afterward, further measurements were carried out until December 2016. The time
required for making each measurement was approximately 15 min. Initial data acquired in November 2014
were used as the zero reference.

3. Threshold in Confining Pressure

Strain profiles along the FO cable embedded in both clayey and sandy soils under a variety of confining pres-
sures were obtained from the pullout tests (Figures 2a, 2b, S5, and S6). The strain increased and propagated
toward cable toe with increasing displacement steps under a low confining pressure, indicative of an evident
progressive interface failure behavior and a weak mechanical coupling between the cable and the soil. In
contrast, the propagation of strain was restrained around the cable head (<0.2 m) under a high confining
pressure, indicating a strong cable-soil interfacial coupling.
To test whether there exists a critical threshold in confining pressure to assure the FO cable-soil interfacial
coupling and quality of FO strain measurements, we proposed an index ξ c–s to quantitatively evaluate the
cable-soil coupling condition, which is given by the following:
 
∫εðl Þdl
ξ c–s ¼ 100 (2)
u0

where u0 is the pullout displacement applied at the cable head and ε(l) is the distribution of strain along the
cable length. Apparently, a large index ξ c–s corresponds to a strong cable-soil interfacial coupling and,
accordingly, there will be less slippage between cable and soil. According to the index ξ c–s, we suggested
a preliminary classification of the FO cable-soil coupling condition: an index of 81–100 is rated as excellent,
61–80 as good, 41–60 as fair, and 0–40 as poor.
For monitoring vertical deformation in the long run, a strain of 10,000 με was regarded as the maximum
detectable value considering the working condition of a common FO cable. Therefore, this index was calcu-
lated based on the displacement step inducing a strain of approximately 10,000 με at the cable head
(Figure 2c). Piecewise linear regression of the data provided good results and showed that the index
increased first with the confining pressure and eventually approached a constant value, demonstrating the
existence of a critical confining pressure. For the cable embedded in the clayey soil, the index was rather
small under low confining pressures. However, the index increased rapidly with the confining pressure, inter-
secting with ξ c–s = 60 at σ c = 0.26 MPa and ξ c–s = 80 at σ c = 0.36 MPa, and approached a constant value of
98.0. On the contrary, for the cable embedded in the sandy soil, the index was as high as 67.9 at zero confin-
ing pressure; it intersected with ξ c–s = 80 at σ c = 0.36 MPa as well but tended to a relatively smaller constant
value of 87.9. This was expected due to the different mechanical behaviors of clayey and sandy soils.
Therefore, in our field test we employed a sand-clay mixture to achieve an optimum coupling between FO
cable and backfilled soil. In either case, our data showed that there exists an identical threshold in confining
pressure (~0.36 MPa) where an excellent FO cable-soil coupling condition can be achieved. Note that this
value was also confirmed by the strain propagation coefficient (defined as propagation length of strain
divided by cable length), where the critical value was within the range 0.2–0.4 MPa (Figure 2d).

4. Strain Profiles and Validation

New strain profiles along the Shengze borehole obtained by DFOS acquired from November 2014 to
December 2016 were presented here. Strain profiles obtained from BOTDR usually show a visible waviness.
Therefore, prior to data analysis the strains were filtered using a second-order Savitzky-Golay filter (Figure
S7); the resulting effective spatial resolution may be estimated as the corresponding window length.
Figure 3a depicts the original Brillouin frequency shift, and Figure 3b is a color map showing the strain
changes along the borehole. Note that data measured below a depth of 145.3 m were aborted due to high
signal loss of the cable. In addition, abnormal strain values were registered at a depth of ~0–6 m as shown in
Figure S8b (shaded area). This could be due to the variation of shallow soil temperature, which was also
observed during the monitoring of a landslide triggered by rainfall infiltration using the coherent optical

ZHANG ET AL. 11,736

 19448007, 2018, 21, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018GL080428 by CAPES, Wiley Online Library on [19/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters 10.1029/2018GL080428

Figure 2. Threshold in confining pressure to achieve a strong fiber optic (FO) cable-soil coupling. (a and b) Representative distributions of strain along a FO cable
installed in a clayey soil and a sandy soil, respectively, as obtained from a laboratory pullout test. Sketch of the test setup and additional figures are in
Figures S1–S6. (c) Influence of confining pressure on the FO cable-soil coupling condition quantified using equation (2). An index of 81–100 is rated as excellent (E),
61–80 as good (G), 41–60 as fair (F), and 0–40 as poor (P).” The dashed lines represent piecewise linear regressions of the test data, intersecting with ξ c–s = 60 (80) at
σ c = 0.26 MPa (0.36 MPa). (d) Correlation between strain propagation coefficient and confining pressure.

time domain reflectometry technique (Kogure & Okuda, 2018). Because strain changes in this near-surface
zone were already comparable to those induced by groundwater abstraction, they were not included in
the color map.
To determine whether the strains registered by the DFOS system were valid, we evaluated the coupling con-
dition between the distributed sensors and the surrounding soil in the Shengze borehole. With the critical
confining pressure in mind, we sought to estimate the confining pressure applied on the distributed sensors.
Because confining pressures were not directly measured in this case, we assumed that the sensors were only
subject to hydrostatic pressures (~10.3 kPa/m; Figure S9), which can provide a sufficient first estimate
(Lumens, 2014). With this simplified assumption, the critical confining pressure was readily converted to an
equivalent depth of ~35 m, implying that sensors below this depth would bond excellently to the surround-
ing soils. Note that the confining pressure acting on the vertically installed cable may be accurately deter-
mined using geostatic pressures. Because Af-II had been the main aquifer for groundwater pumping in this
region (Text S3 and Table S1; Wu et al., 2018) and the resulting deformation occurred mainly in Af-II and
two adjacent aquitards At-II and At-III (41.2–137.9 m; Figure 3b), our analysis indicated that the strains
acquired by DFOS were reliable.

ZHANG ET AL. 11,737

 19448007, 2018, 21, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018GL080428 by CAPES, Wiley Online Library on [19/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters 10.1029/2018GL080428

Figure 3. Monitored data profiles along the Shengze borehole acquired from 13 November 2014 to 27 December 2016. (a) Brillion frequency shifts (BFSs). The inset
plot shows the change in BFS during the 2-year period. (b) A color map showing vertical strain changes along the borehole. Positive values mean rebound;
negative values mean compaction. Also plotted is a simplified lithology indicating aquifer and aquitard units. Full strain profiles and detailed lithology are shown
respectively in Figure S8a and Table S1. (c) Strain changes at depths of 44.7, 63.4, 80, and 90 m, representative of At-II (red), Af-II (green), and At-III (black). Also plotted
is the variation of pore fluid pressure (PFP) at 87.7-m depth. UA: unconfined aquifer; At: aquitard; Af: confined aquifer.

To further validate the FO strain measurements, we compared subsidence rates estimated from FO data with
available corroborating data. The surface displacements were calculated by integrating FO strain profiles
along the borehole (Figure S10a). Because FO strains showed visible fluctuations in the near-surface zone

ZHANG ET AL. 11,738

 19448007, 2018, 21, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018GL080428 by CAPES, Wiley Online Library on [19/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters 10.1029/2018GL080428

(0- to 6-m depth), they were not included in the calculation. Deformation below 150 m was neglected as the
sediments were either semilithified or hard and completely lithified according to borehole logs. The compar-
ison indicated that the calculated subsidence rates from FO strains agreed generally with leveling data
(Figure S10b), which further validated the DFOS technique.

5. Do We Need Distributed Strain?

As illustrated in Figures 3b and S8a, vertically distributed sensing based on Brillouin scattering can provide a
complete subsurface strain profile with a single FO cable. The strain varying with time and depth can offer
helpful insights into the vertical distribution of strata deformation and evolution.
An obvious advantage of distributed strain measurements is to accurately locate a layer undergoing compac-
tion or rebound. With the distributed strain we were able to ascertain that most of the aquifer-system defor-
mation occurred at a depth of approximately 60–100 m during the period November 2014 to December
2016. Positive strains were primarily observed in Af-II, indicative of an overall rebound of the pumped aquifer.
Despite a steady recovery of water levels and a reduced rate of subsidence (Text S3), we pinpointed the loca-
tion of a 7.6-m-thick small zone (60.3- to 67.9-m depth) where compaction was still actively occurring. This
was not unexpected because the drainage of aquitards can exhibit pronounced time dependence due to
their low hydraulic conductivity (Helm, 1975, 1976), resulting in years of continued compaction (Chaussard
et al., 2014; Miller & Shirzaei, 2015). This was especially evident for this zone, probably because it corre-
sponded to a layer of very soft silty clay with silty sand according to borehole logs (Table S1). The compaction
of this zone was ~1.83 mm over the 2-year period, which contributed dominantly to the compaction of At-II
(Figure S11). In addition, there were several minor compaction zones visible in Figure 3b; their compaction
(e.g., ~0.2 mm from 121.65- to 125.35-m depth) was however not comparable to that of the main one.
A second benefit of distributed information is to examine the strain response of soil at various locations along
the entire cable length. We illustrated the desirability of doing so by looking at strain changes at depths of
44.7, 63.4, 80, and 90 m, representative of At-II (red), Af-II (green), and At-III (black) (Figure 3c). The change
in strain at 80-m depth reflected the strain response of the pumped aquifer Af-II, which varied nearly synchro-
nously with the pore fluid pressure. A decrease in pore fluid pressure led to a decrease in strain, and vice
versa; this can be well described by the theory of hydrodynamic consolidation (Terzaghi, 1943). The strain
at 63.4-m depth, which was 11 m above the top surface of Af-II, was the maximum negative strain observed
in the 7.6-m-thick zone (also At-II). The negative strain recorded in this depth peaked in December 2016,
revealing a strain change of ~642 με over the 2-year period. In contrast to this, the data recorded at 44.7-
m depth (located 29.65 m above the top surface of Af-II) showed a positive strain up to ~92 με, representative
of the localized feature of a silty clay with calcareous concretions (Table S1). Similar to the delayed compac-
tion of the 7.6-m-thick zone in At-II occurring after groundwater abstraction had stopped, negative strains
were expected for some zones within the underlying aquitard At-III. However, although our first data
acquired from December 2012 to November 2014 showed some negative strains in the upper part of At-III
(Wu et al., 2015), we did not observe such phenomena from the new data collected during the period
November 2014 to December 2016. The strain registered at 90-m depth fluctuated around a positive strain
of ~203 με, implying that delayed deformation of At-III had nearly ended. Combined, these results demon-
strate that the spatial continuity of FO strains allows for identifying very localized deformation, whereas such
features may not be captured by traditional instrumentation such as extensometers (Hwang et al., 2008;
Metzger et al., 2001; Riley, 1969).

6. Discussion and Conclusions

A novel technique for vertically distributed sensing of deformation has been demonstrated by application of
DFOS based on Brillouin scattering. Detailed vertical strain profiles obtained during the period November
2014 to December 2016 (Figures 3b and S8a) revealed a 7.6-m-thick zone actively undergoing compaction
despite a steady recovery of water levels. Based on the analysis in the prior section, we conclude that
aquifer-system compaction during this time period was primarily ascribed to the delayed compaction of
At-II, which was dominated by the compression of the 7.6-m-thick zone.
Our data showed a near-surface zone at approximately 0- to 6-m depth where the Brillouin frequency shift
fluctuated markedly over time (Figure S8b), which was primarily attributed to the variation of shallow soil

ZHANG ET AL. 11,739

 19448007, 2018, 21, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018GL080428 by CAPES, Wiley Online Library on [19/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters 10.1029/2018GL080428

temperature (Bense & Kooi, 2004; Kogure & Okuda, 2018). This can be readily addressed by employing an
additional FO temperature cable insensitive to mechanical strains (Habel & Krebber, 2011; Schenato, 2017).
By subtracting thermally induced Brillouin frequency shifts (equation (1)), the actual strain due to strata defor-
mation can be obtained.
Using BOTDA we found that there exists a threshold in confining pressure (~0.36 MPa) under which a strong
FO cable-soil coupling can be achieved (Figure 2c). This critical confining pressure ensured the quality of
strain data acquired from the FO strain cable deployed in the Shengze borehole. In the case where corrobor-
ating data are not available, the FO cable-soil coupling index may help to assess the validity of measured
strains, but further investigation will be needed to examine how this index correlates with actual
strain measurements.
Previous research showed vertically distributed strain sensing of a shallow landslide up to ~15 m (Kogure &
Okuda, 2018). In their study, the measured strains were not converted to actual deformation, probably due to
the complicated correlation between axial linear strain and shear deformation. In our case where the defor-
mation of strata was principally vertical and parallel to the cable axis, a direct integration of the spatially con-
tinuous strain data should yield the actual strata deformation. The estimated subsidence rate from
distributed strains correlated with available corroborating data (Figure S10b). However, until standards are
well established pertaining to cable selection, field deployment, and data processing, a combination of
DFOS with other techniques (e.g., extensometry) would be necessary.
In summary, we have shown the usefulness of DFOS based on Brillouin scattering for examining subsurface
deformation in Shengze, southern Yangtze Delta, China. The complete subsurface strain profile obtained
using BOTDR with a single FO cable provided helpful insights into the vertical distributions of strata deforma-
tion. Although our case study has focused on groundwater-related aquifer-system deformation, this technol-
ogy can be extended to other areas where measurable subsidence or uplift is actively occurring. Furthermore,
recording subsurface changes in this fashion may also be useful in other geophysical and engineering appli-
cations that require refined monitoring of the media.

Acknowledgments References
We thank Nanzee staff for field support;
C. Liu, X. Wang, C.-X. Feng, and J.-A. Amelung, F., Galloway, D. L., Bell, J. W., Zebker, H. A., & Laczniak, R. J. (1999). Sensing the ups and downs of Las Vegas: InSAR reveals structural
Chen for laboratory assistance; and K. control of land subsidence and aquifer-system deformation. Geology, 27(6), 483. https://doi.org/10.1130/0091-7613(1999)027<0483:
Soga, H.-H. Zhu, and D. Zhang for fruitful STUADO>2.3.CO;2
discussions. We also acknowledge the Bao, X., & Chen, L. (2012). Recent progress in distributed fiber optic sensors. Sensors, 12(7), 8601–8639. https://doi.org/10.3390/
constructive reviews provided by the s120708601
Editor and two reviewers. This work was Bense, V. F., & Kooi, H. (2004). Temporal and spatial variations of shallow subsurface temperature as a record of lateral variations in
supported by the National Natural groundwater flow. Journal of Geophysical Research, 109, B04103. https://doi.org/10.1029/2003JB002782
Science Foundation of China (NSFC) Briggs, M. A., Lautz, L. K., & McKenzie, J. M. (2012). A comparison of fibre-optic distributed temperature sensing to traditional methods of
grants 41427801 and 41230636. C.-C. Z. evaluating groundwater inflow to streams. Hydrological Processes, 26(9), 1277–1290. https://doi.org/10.1002/hyp.8200
was supported by the Postgraduate Carlino, S., Mirabile, M., Troise, C., Sacchi, M., Zeni, L., Minardo, A., Caccavale, M., et al. (2016). Distributed-temperature-sensing using optical
Research and Practice Innovation methods: A first application in the offshore area of Campi Flegrei caldera (southern Italy) for volcano monitoring. Remote Sensing, 8(8),
Program of Jiangsu Province grant 674. https://doi.org/10.3390/rs8080674
KYCX17_0031, along with a scholarship Chaussard, E., Bürgmann, R., Shirzaei, M., Fielding, E. J., & Baker, B. (2014). Predictability of hydraulic head changes and characterization of
from the China Scholarship Council. K. aquifer-system and fault properties from InSAR-derived ground deformation. Journal of Geophysical Research: Solid Earth, 119, 6572–6590.
G. was supported by NSFC grant https://doi.org/10.1002/2014JB011266
41502274. Data to support this research Curtis, A., & Kyle, P. (2011). Geothermal point sources identified in a fumarolic ice cave on Erebus volcano, Antarctica using fiber optic dis-
are available on repository (https://fig- tributed temperature sensing. Geophysical Research Letters, 38, L16802. https://doi.org/10.1029/2011GL048272
share.com/s/d9fa3ebbd104621ce138). Dou, S., Lindsey, N., Wagner, A. M., Daley, T. M., Freifeld, B., Robertson, M., Peterson, J., et al. (2017). Distributed acoustic sensing for seismic
monitoring of the near surface: A traffic-noise interferometry case study. Scientific Reports, 7(1), 11,620. https://doi.org/10.1038/s41598-
017-11986-4
Erban, L. E., Gorelick, S. M., Zebker, H. A., & Fendorf, S. (2013). Release of arsenic to deep groundwater in the Mekong Delta, Vietnam, linked to
pumping-induced land subsidence. Proceedings of the National Academy of Sciences, 110(34), 13,751–13,756. https://doi.org/10.1073/
pnas.1300503110
Famiglietti, J. S. (2014). The global groundwater crisis. Nature Climate Change, 4(11), 945–948. https://doi.org/10.1038/nclimate2425
Galloway, D. L., & Burbey, T. J. (2011). Review: Regional land subsidence accompanying groundwater extraction. Hydrogeology Journal, 19(8),
1459–1486. https://doi.org/10.1007/s10040-011-0775-5
Habel, W. R., & Krebber, K. (2011). Fiber-optic sensor applications in civil and geotechnical engineering. Photonic Sensors, 1(3), 268–280.
https://doi.org/10.1007/s13320-011-0011-x
Helm, D. C. (1975). One-dimensional simulation of aquifer system compaction near Pixley, California: 1. Constant parameters. Water
Resources Research, 11, 465–478. https://doi.org/10.1029/WR011i003p00465
Helm, D. C. (1976). One-dimensional simulation of aquifer system compaction near Pixley, California: 2. Stress-Dependent Parameters. Water
Resources Research, 12, 375–391. https://doi.org/10.1029/WR012i003p00375
Higgins, S., Overeem, I., Tanaka, A., & Syvitski, J. P. M. (2013). Land subsidence at aquaculture facilities in the Yellow River delta, China.
Geophysical Research Letters, 40, 3898–3902. https://doi.org/10.1002/grl.50758

ZHANG ET AL. 11,740

 19448007, 2018, 21, Downloaded from https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2018GL080428 by CAPES, Wiley Online Library on [19/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Geophysical Research Letters 10.1029/2018GL080428

Horiguchi, T., Kurashima, T., & Tateda, M. (1989). Tensile strain dependence of Brillouin frequency shift in silica optical fibers. IEEE Photonics
Technology Letters, 1(5), 107–108. https://doi.org/10.1109/68.34756
Huntley, D., Bobrowsky, P., Qing, Z., Sladen, W., Bunce, C., Edwards, T., Hendry, M., et al. (2014). Fiber optic strain monitoring and evaluation of
a slow-moving landslide near Ashcroft, British Columbia, Canada. In K. Sassa, P. Canuti, & Y. Yin (Eds.), Landslide Science for a Safer
Geoenvironment, The International Programme on Landslides (IPL), (Vol. 1, pp. 415–421). Cham: Springer International Publishing. https://
doi.org/10.1007/978-3-319-04999-1_58
Hwang, C., Hung, W.-C., & Liu, C.-H. (2008). Results of geodetic and geotechnical monitoring of subsidence for Taiwan High Speed Rail
operation. Natural Hazards, 47(1), 1–16. https://doi.org/10.1007/s11069-007-9211-5
Jousset, P., Reinsch, T., Ryberg, T., Blanck, H., Clarke, A., Aghayev, R., Hersir, G. P., et al. (2018). Dynamic strain determination using fibre-optic
cables allows imaging of seismological and structural features. Nature Communications, 9(1), 2509. https://doi.org/10.1038/s41467-018-
04860-y
Kogure, T., & Okuda, Y. (2018). Monitoring the vertical distribution of rainfall-induced strain changes in a landslide measured by distributed
fiber optic sensing with rayleigh backscattering. Geophysical Research Letters, 45, 4033–4040. https://doi.org/10.1029/2018GL077607
Lienhart, W. (2015). Case studies of high-sensitivity monitoring of natural and engineered slopes. Journal of Rock Mechanics and Geotechnical
Engineering, 7(4), 379–384. https://doi.org/10.1016/j.jrmge.2015.04.002
Lindsey, N. J., Martin, E. R., Dreger, D. S., Freifeld, B., Cole, S., James, S. R., Biondi, B. L., et al. (2017). Fiber-optic network observations of
earthquake wavefields. Geophysical Research Letters, 44, 11,792–11,799. https://doi.org/10.1002/2017GL075722
Lumens, P. G. E. (2014). Fibre-optic sensing for application in oil and gas wells. Technische Universiteit Eindhoven, Eindhoven. https://doi.org/
10.6100/IR769555
Massonnet, D., & Feigl, K. L. (1998). Radar interferometry and its application to changes in the Earth’s surface. Reviews of Geophysics, 36,
441–500. https://doi.org/10.1029/97RG03139
Metzger, L. F., Ikehara, M. E., & Howle, J. F. (2001). Vertical-deformation, water-level, microgravity, geodetic, water-chemistry, and flow-rate
data collected during injection, storage, and recovery tests at Lancaster, Antelope Valley, California, September 1995 through September
1998. U.S. Geological Survey Open-File Report. https://doi.org/10.3133/ofr01414
Michlmayr, G., Chalari, A., Clarke, A., & Or, D. (2017). Fiber-optic high-resolution acoustic emission (AE) monitoring of slope failure. Landslides,
14(3), 1139–1146. https://doi.org/10.1007/s10346-016-0776-5
Miller, M. M., & Shirzaei, M. (2015). Spatiotemporal characterization of land subsidence and uplift in Phoenix using InSAR time series and
wavelet transforms. Journal of Geophysical Research: Solid Earth, 120, 5822–5842. https://doi.org/10.1002/2015JB012017
Murai, D., Kunisue, S., Higuchi, T., & Kokubo, T. (2013). In-Situ Formation Compaction Monitoring in Deep Reservoirs by Use of Fiber Optics. In
EGU General Assembly 2013 (pp. EGU2013–3860). Vienna, Austria.
Ortiz-Zamora, D., & Ortega-Guerrero, A. (2010). Evolution of long-term land subsidence near Mexico City: Review, field investigations, and
predictive simulations. Water Resources Research, 46, W01513. https://doi.org/10.1029/2008WR007398
Picarelli, L., Damiano, E., Greco, R., Minardo, A., Olivares, L., & Zeni, L. (2015). Performance of slope behavior indicators in unsaturated pyro-
clastic soils. Journal of Mountain Science, 12(6), 1434–1447. https://doi.org/10.1007/s11629-014-3104-3
Poland, J. F., & Davis, G. H. (1969). Land subsidence due to withdrawal of fluids. In D. J. Varnes, & G. Kiersch (Eds.), Reviews in Engineering
Geology, (pp. 187–269). Boulder, Colorado: Geological Society of America. https://doi.org/10.1130/REG2-p187
Riley, F. S. (1969). Analysis of borehole extensometer data from central California. In L. Tison (Ed.), Land subsidence: Proceedings of the Tokyo
symposium, (pp. 423–431). Tokyo: IAHS Publications.
Schenato, L. (2017). A review of distributed fibre optic sensors for geo-hydrological applications. Applied Sciences, 7(9), 896. https://doi.org/
10.3390/app7090896
Schenato, L., Palmieri, L., Camporese, M., Bersan, S., Cola, S., Pasuto, A., Galtarossa, A., et al. (2017). Distributed optical fibre sensing for early
detection of shallow landslides triggering. Scientific Reports, 7(1), 14,686. https://doi.org/10.1038/s41598-017-12610-1
Selker, J., van de Giesen, N., Westhoff, M., Luxemburg, W., & Parlange, M. B. (2006). Fiber optics opens window on stream dynamics.
Geophysical Research Letters, 33, L24401. https://doi.org/10.1029/2006GL027979
Shi, X., Feng, Z., Yao, B., Huang, X., & Wu, J. (2014). Study on the deformation characteristics of soil layers after banning groundwater pumping
in Su-Xi-Chang area. Quaternary Sciences, 34(5), 1062–1071.
Shi, X., Xue, Y., Wu, J., Ye, S., Zhang, Y., Wei, Z., & Yu, J. (2008). Characterization of regional land subsidence in Yangtze Delta, China: The
example of Su-Xi-Chang area and the city of Shanghai. Hydrogeology Journal, 16(3), 593–607. https://doi.org/10.1007/s10040-007-0237-2
Shirzaei, M., & Bürgmann, R. (2018). Global climate change and local land subsidence exacerbate inundation risk to the San Francisco Bay
Area. Science Advances, 4(3). https://doi.org/10.1126/sciadv.aap9234
Terzaghi, K. (1943). Theoretical soil mechanics. Hoboken, NJ, USA: John Wiley & Sons, Inc. https://doi.org/10.1002/9780470172766
Wu, J., Hu, Y., & Luo, Z. (2018). Impact of building load on land subsidence in Shengze area of Wujiang, Jiangsu Province. Journal of Geology,
42(1), 167–174.
Wu, J., Jiang, H., Su, J., Shi, B., Jiang, Y., & Gu, K. (2015). Application of distributed fiber optic sensing technique in land subsidence monitoring.
Journal of Civil Structural Health Monitoring, 5(5), 587–597. https://doi.org/10.1007/s13349-015-0133-8
Xue, Y.-Q., Zhang, Y., Ye, S.-J., Wu, J.-C., & Li, Q.-F. (2005). Land subsidence in China. Environmental Geology, 48(6), 713–720. https://doi.org/
10.1007/s00254-005-0010-6
Zhang, A. G., & Wei, Z. X. (2005). Land subsidence in China. Shanghai, China: Shanghai Scientific & Technical Publishers.
Zhang, C.-C., Zhu, H.-H., & Shi, B. (2016). Role of the interface between distributed fibre optic strain sensor and soil in ground deformation
measurement. Scientific Reports, 6(1), 36,469. https://doi.org/10.1038/srep36469.

ZHANG ET AL. 11,741