Microanchored fiber-optic DSS in boreholes allows strain profiling of the 1 shallow subsurface 2

19 Vertical deformation profiles of subterranean geological formations are conventionally measured by 20 borehole extensometry. Distributed strain sensing (DSS) paired with fiber-optic cables installed in the 21 ground opens up possibilities of acquiring high-resolution static and quasistatic strain profiles of 22 deforming strata, but it is currently limited by reduced data quality due to complicated patterns of 23 interaction between the buried cables and their surroundings, especially in upper soil layers under low 24 confining pressures. Extending recent DSS studies, we present an improved approach for strain 25 determination along entire lengths of vertical boreholes by using microanchored fiber-optic cables 26 designed to optimize ground-to-cable coupling at the near surface. We proposed a novel criterion for soil– 27 cable coupling evaluation based on the geotechnical bearing capacity theory. We applied this enhanced 28 methodology to monitor groundwater-related vertical motions in both laboratory and field experiments. 29 Corroborating extensometer recordings, acquired simultaneously, validated fiber optically determined 30 displacements, suggesting microanchored DSS as an improved means for detecting and monitoring 31 shallow subsurface strain profiles. 32 Non-peer reviewed preprint submitted to EarthArXiv 2 / 24 Introduction 33 Shallow geohazards, such as landslides, debris flows, ground subsidence, and sinkhole collapses, can 34 have devastating effects on populations, economies, and landscapes across the world. The initiation and 35 evolution of these near-surface hazards are often accompanied by measurable deformation, and 36 therefore measuring and monitoring their spatio-temporal displacements is essential to implementing 37 early warning systems. Of the methods for vertical deformation acquisition, interferometric synthetic 38 aperture radar (InSAR) and global navigation satellite system (GNSS) are commonly used to detect land39 surface elevation changes. These ground-based or remotely sensed techniques have proved to be 40 effective in mapping large-scale ground motions, but they do not allow for subsurface deformation 41 profiles to be obtained. Drilling is a common means to determine lithology; by installing extensometers in 42 drilled boreholes, deformations occurring at certain depths below the ground surface can be observed. 43 While highly precise measurements can be made using borehole extensometry, the spatial resolution for 44 such systems is often constrained by discretely instrumented measuring points (markers), commonly 45 deployed at depths corresponding to critical layers. 46 Fiber-optic sensing has advanced significantly in the past few years for strain determination in many 47 areas of earth science and engineering. Fiber-optic sensing technologies are normally categorized 48 according to the measurand or the optical scattering mechanism whereby the measurement is made. 49 The fiber sensing method utilized for static strain detection is often referred to as distributed strain 50 sensing (DSS) while for dynamic strain acquisition as distributed acoustic/vibration sensing 51 (DAS/DVS). An attractive feature of the broad category of fiber-optic sensing technologies is their 52 ability to make spatially continuous strain (strain-rate) recordings along a fiber-optic cable up to tens of 53 kilometers in aperture. This advantage has been instrumental, for example, in localizing accurately active 54 compaction zones resulting from subsurface resources exploitation and better characterizing 55 hydromechanical responses. 56 The mechanical coupling between fiber-optic cables and Earth, depending on both cable construction 57 and installation, is an important influencing factor to carrying out successful fiber-optic monitoring 58 campaigns. Many have reported that the quality of fiber-optic data is strongly conditioned by the degree 59 of rigid ground–cable coupling, for either DSS or DAS (hereafter we will focus on DSS to limit 60 this study’s extent). This is especially the case when the deformation of low-confined upper layers is of 61 particular interest, and can be exacerbated by highly saturated weak strata such as those containing large 62 amounts of soft soils. In this respect, correction of measured strains via rigorous ground-to-fiber strain 63 transfer analysis has been proposed to be a potential solution, but it would be better for field 64 applications to have enhanced fiber-optic instrumentation, such as a specialty cable that can be rigidly 65 coupled to its surroundings. 66 Non-peer reviewed preprint submitted to EarthArXiv 3 / 24 Using anchors to improve interface bonding between reinforcements and surroundings is a common 67 practice in geotechnical engineering. This has inspired the DSS community to attach anchor-like 68 elements mechanically to outer coatings or jackets of fiber-optic cables, forming dedicated cables capable 69 of detecting displacements of laboratory physical models or in a field setting via horizontally70 trenched direct burial. Pullout tests and shear zone simulation tests were performed to confirm the 71 performance of a shallowly trenched, three-dimensional microanchored cable for landslide monitoring. 72 As to theory, the interaction of tube-anchored cables with surrounding soils has been interpreted from the 73 perspective of interface shearing, extending the framework developed primarily for unanchored 74 DSS. While this allows the overall interface shear strength between soil and anchored cables to be 75 estimated, it precludes the consideration of passive earth pressure effects commonly observed during soil– 76 anchor interaction. 77 We describe here an improved fiber-optic DSS approach for sensing of vertical ground 78 displacements with microanchored strain sensing cables deployed in boreholes. We fabricated three 79 microanchors to enhance soil–cable interlocking effects adding on previous work. We proposed a new 80 criterion for assessing soil–cable coupling based on the geotechnical bearing capacity theory. We 81 examined the effects of confining pressure, soil and interface strength parameters, and anchor type and 82 dimension on the performance of the microanchored DSS system. We demonstrated the feasibility of this 83 improved methodology through elementary testing, physical modeling, and a field experiment conducted 84 in a coastal setting. 85 86 DSS measurement principle 87 Figure 1a shows schematically a microanchored fiber-optic cable buried in a borehole for the detection of 88 vertical displacements of geological formations resulting from subsurface resources extraction. DSS 89 techniques used for fiber strain acquisition are based on Brillouin or Rayleigh scattering. These include 90 Brillouin optical time-domain reflectometry (BOTDR), Brillouin optical time-/frequency-domain analysis 91 (BOTDA/BOFDA), optical frequency-domain reflectometry (OFDR), and tunable‐wavelength coherent 92 optical time‐domain reflectometry (TW‐COTDR). Taking the BOTDR technique with single-ended 93 deployment as an example (Fig. 1b), an external strain (referred to axial strain if not otherwise stated) 94 acting on a fiber-optic cable will induce a shift in frequency B   of the Brillouin backscattered light 95 inside the fiber detectable by a BOTDR interrogator. The strain change   can be determined according 96 to: 97 B T e 1 ( ) C T C        (1) 98 Non-peer reviewed preprint submitted to EarthArXiv 4 / 24 where e C is the frequency shift–strain coefficient, T C is the frequency shift–temperature coefficient, and 99 T  is the change in temperature that can be quantified using a colocated temperature sensing cable 100 insensitive to mechanical strains. Because Brillouin backscattering is generated at each point of the fiber, 101 by repeatedly launching light pulses into the fiber a complete strain profile of the deforming strata along 102 the entire borehole length can be mapped. 103 Durability is a central concern for any instrument installed in a subsurface environment. 104 Theoretically, borehole-embedded fiber-optic DSS systems can be permanently used for deformation 105 observation as fiber-optics are inherently corrosion resistant. In practice, fiber-optic cables may break due 106 to large stratum deformation (the ultimate tensile strain of fiber-optics is ~2%, i.e., 20,000 με). Our first 107 borehole DSS system was deployed in Shengze (Southern Yangtze Delta, China) in 2012 and strain 108 acquisition has been performed routinely for nearly ten years. We anticipate such systems would survive 109 and function properly for at least several decades; a robust yet strain-sensitive cable is crucial. 110 111 Fabrication of microanchored cables 112 Anchor-like elements are viewed as essential to ensuring sufficient ground–cable coupling and hence the 113 DSS measurement quality can be improved. For this purpose, we fabricated three types of 114 microanchors—disc, cylinder, and spindle. These anchors were attached at discrete points to 115 commercially available fiber-optic strain sensing cables using epoxy resin adhesives. In doing so, three 116 dedicated cables were developed, covering both field and laboratory application scenarios; their features 117 and properties are summarized in Table 1. The disc-anchored cable is well suited for low-confined 118 laboratory physical modeling, as the 0.9-mmor 2-mm-diameter thermoplastic polyurethane (TPU)119 jacketed cable (NZS-DSS-C07 by NanZee Sensing Ltd.) can readily be integrated into loose media, 120 owing to its relatively low Young’s modulus (E = ~1 GPa), and the discs can enhance considerably soil– 121 cable interlocking effects. The cylinderor spindle-anchored cable utilizes a 5-mm-diameter steel strand122 reinforced, polyethylene (PE)-jacketed cable (NZS-DSS-C02; E = ~8 GPa). This ensures high survival 123 rates during sensor deployment. Moreover, the small-diameter cylinders and spindles (compared to discs) 124 render the fabricated cables suitable for direct burial installations in field monitoring boreholes. 125 126 Interaction mechanism between soil and microanchored cable 127 Pullout resistance mechanism of bearing microanchored cable. 128 We first elaborated on the interaction mechanism between soil and a buried microanchored cable through 129 a concise theoretical analysis (Fig. 2), which is a first step toward the successful application of the 130 proposed methodology. The analysis builds on the bearing capacity theory presented by Jewell and 131 Bergado et al., originating from geotechnical engineering. 132 Non-peer reviewed preprint submitted to EarthArXiv 5 / 24 During pullout, the resistance of the microanchored cable is composed mainly of two parts (Fig. 2a): 133 the frictional force component caused by sliding between the cable surface and soil, and the bearing 134 capacity component generated by extrusion between the microanchors and soil. Hence, the ultimate 135 pullout resistance r F of the microanchored cable can be expressed by: 136 r fr br F F F   (2) 137 where fr F is the soil–cable interface friction resistance that can be determined according to the Mohr– 138 Coulomb theory, and br F is the bearing resistance of microanchors. Note that fr F may be further divided 139 into the friction resistance between the soil and the anchor fr1 F and that between the soil and the 140 unanchored cable segment fr2 F . 141 The microanchor bearing resistance br F can be evaluated as follows: 142 c br b s L F S L   (3) 143 where c L is the embedment length of the microanchored cable; s L is the spacing between the 144 microanchors; S is the surface area of the microanchor; and b  is the bearing stress of a single 145 microanchor that can be evaluated by: 146 b n q c = N cN    (4) 147 where n  is the applied stress normal to the cable axis; c is the soil cohesion; and q N and c N are the 148 bearing capacity factors associated with the bearing failure mode. 149 Existing pullout bearing failure mechanisms include the general shear failure, punching shear failure, 150 and modified punching shear failure. Among the three failure modes, the general and punching shear 151 failures form the upper and lower bounds of the problem, while the modified punching failure can well 152 describe the bearing failure characteristics of grid reinforcements such as geogrids and geotextiles. 153 Hence, the modified punching failure mode was employed herein to describe the bearing mechanism of 154 microanchored fiber-optic cables, and q N and c N can be respectively expressed as: 155 2 tan q 1 1 1 π sin(2 ) tan 2 2 cos 4 2 k k N e                         (5) 156 2 tan c 1 π tan cot sin 4 2 N e               (6) 157 Non-peer reviewed preprint submitted to EarthArXiv 6 / 24 where  is the soil internal friction angle; k is the lateral earth pressure coefficient; and  is the angle 158 of the rotational failure zone (Fig. 2b). For k = 1 and  = /2, theoretical predictions were found to 159 agree well with laboratory test data, and q N and c N are thus reduced to: 160 π tan q 1 π tan cos 4 2 N e            (7) 161 π tan c 1 π tan cot sin 4 2 N e              (8) 162 Validation of bearing resistance equations via laboratory pullout testing. 163 To explore whether the bearing capacity theory is suitable for describing cable anchor failure, we 164 performed laboratory pullout tests on disc-anchored fiber-optic cables at variable anchor diameters. The 165 setup of the pullout tests is sketched in Supplementary Fig. S1a. The soil used was a poorly graded 166 medium sand. Its physical property parameters are: Gs = 2.65, d10 = 0.140 mm, d60 = 0.472, Cu = 3.371, Cc 167 = 1.144, dmax = 1.82 Mg m, and wopt = 7.82%. Four anchor diameters were investigated: 10, 20, 30, and 168 40 mm (Fig. S1b,c). For each test, a microanchored cable was buried in the testing soil at a density of 169 1.70 Mg m in the 500 mm × 160 mm × 160 mm chamber, and was pulled out at a velocity of 0.05 mm/s 170 while recording pullout forces (± 0.1 N). The test was terminated when pullout failure occurred. As the 171 tests lasted for only one hour, the variation of room temperature was negligible and temperature 172 compensation was thus not necessary. 173 A comparison between the measured pullout resistances and those predicted using the bearing 174 resistance theory (equations (2)–(8)) was carried out; the results are depicted in Fig. 3. Note that in 175 addition to modified punching shear failure, upperand lower-bound values constrained from general and 176 punching shear failure mechanisms were also computed. The parameters used for theoretical modeling are 177 shown in the caption of Fig. 3. It can be observed that the modified punching shear failure mechanism 178 presently used can better describe the bearing failure behavior of disc-anchored cables compared to the 179 general or punching shear failure. Although these results verified preliminarily the bearing resistance 180 equations, more laboratory testing should be conducted to further validate the proposed method, 181 especially its suitability for describing cylinderand spindle-anchor cables. 182 183 Criterion for soil–microanchored cable coupling evaluation 184 Criterion establishment. 185 Iten et al. argued that the contact between soil and a buried anchored cable is a combination of overall 186 bonding and point fixation. Experimental evidence further showed that this combination depends on the 187 Non-peer reviewed preprint submitted to EarthArXiv 7 / 24 deformation stage of the soil–cable interface. Specifically, tube anchors will continue to contribute to the 188 overall interface shear strength after the interface between soil and unanchored segments fails, converting 189 the contact from overall bonding to point fixation. Point fixation may reduce the spatial resolution of 190 DSS, but it is commonly sufficient to obtain a detailed strain profile of subsurface strata. Hence, for 191 ground motion sensing the acquired strain data can be considered as credible provided that the capacity of 192 microanchors has not been reached. In this sense, of particular importance in coupling assessment is the 193 evaluation of stress states of microanchors, especially for those buried in shallow strata. Force equilibrium 194 of a single microanchor yields (Fig. 2c): 195 a f1 b 1 2 F F F N N     (9) 196 where a F is the interaction force between the soil and microanchor; f1 F is the friction force; b F is the 197 bearing force; and 1 N and 2 N are the tensions or compressions provided by the unanchored cable 198 segments, which can be calculated using the measured fiber strain: 199 2 c c π ( ) ( ) 4 N x D E x   (10) 200 where c D and c E are the diameter and Young’s modulus of the unanchored cable segment, and ( ) x  is 201 the fiber-optic strain measurement. 202 Combining Eq. (9) with Eq. (10) yields: 203 2 a c c c π 4 F D E    (11) 204 where c   is the difference in strain measured by the two adjacent unanchored cable segments. Note that 205 if there is no evident step change in strain across the anchors, the strains of the unanchored cable 206 segments may be averaged to obtain c   , which is the case for our laboratory and field monitored data. 207 For the three microanchor types presented in the current work, the ultimate soil–anchor interaction 208 force ar F can be readily derived from the bearing capacity theory as: 209


33
Shallow geohazards, such as landslides, debris flows, ground subsidence, and sinkhole collapses, can 34 have devastating effects on populations, economies, and landscapes across the world. The initiation and 35 evolution of these near-surface hazards are often accompanied by measurable deformation [1][2][3] , and 36 therefore measuring and monitoring their spatio-temporal displacements is essential to implementing 37 early warning systems. Of the methods for vertical deformation acquisition, interferometric synthetic 38 aperture radar (InSAR) and global navigation satellite system (GNSS) are commonly used to detect land-39 surface elevation changes 4 . These ground-based or remotely sensed techniques have proved to be 40 effective in mapping large-scale ground motions 5 , but they do not allow for subsurface deformation 41 profiles to be obtained. Drilling is a common means to determine lithology; by installing extensometers in 42 drilled boreholes, deformations occurring at certain depths below the ground surface can be observed 6 . 43 While highly precise measurements can be made using borehole extensometry, the spatial resolution for 44 such systems is often constrained by discretely instrumented measuring points (markers), commonly 45 deployed at depths corresponding to critical layers. 46 Fiber-optic sensing has advanced significantly in the past few years for strain determination in many 47 areas of earth science and engineering 7-14 . Fiber-optic sensing technologies are normally categorized 48 according to the measurand or the optical scattering mechanism whereby the measurement is made 15,16 . 49 The fiber sensing method utilized for static strain detection is often referred to as distributed strain 50 sensing (DSS) while for dynamic strain acquisition as distributed acoustic/vibration sensing 51 (DAS/DVS) 17 . An attractive feature of the broad category of fiber-optic sensing technologies is their 52 ability to make spatially continuous strain (strain-rate) recordings along a fiber-optic cable up to tens of 53 kilometers in aperture. This advantage has been instrumental, for example, in localizing accurately active 54 compaction zones resulting from subsurface resources exploitation 18-20 and better characterizing 55 hydromechanical responses 21,22 . 56 The mechanical coupling between fiber-optic cables and Earth, depending on both cable construction 57 and installation 17 , is an important influencing factor to carrying out successful fiber-optic monitoring 58 campaigns. Many have reported that the quality of fiber-optic data is strongly conditioned by the degree 59 of rigid ground-cable coupling, for either DSS 7,23-25 or DAS 26 (hereafter we will focus on DSS to limit 60 this study's extent). This is especially the case when the deformation of low-confined upper layers is of 61 particular interest, and can be exacerbated by highly saturated weak strata such as those containing large 62 amounts of soft soils. In this respect, correction of measured strains via rigorous ground-to-fiber strain 63 transfer analysis has been proposed to be a potential solution 27 , but it would be better for field 64 applications to have enhanced fiber-optic instrumentation, such as a specialty cable that can be rigidly 65 coupled to its surroundings. 66 Non-peer reviewed preprint submitted to EarthArXiv

/ 24
Using anchors to improve interface bonding between reinforcements and surroundings is a common 67 practice in geotechnical engineering 28,29 . This has inspired the DSS community to attach anchor-like 68 elements mechanically to outer coatings or jackets of fiber-optic cables, forming dedicated cables capable 69 of detecting displacements of laboratory physical models 30-33 or in a field setting via horizontally-70 trenched direct burial 34 . Pullout tests and shear zone simulation tests were performed to confirm the 71 performance of a shallowly trenched, three-dimensional microanchored cable for landslide monitoring 35 . 72 As to theory, the interaction of tube-anchored cables with surrounding soils has been interpreted from the 73 perspective of interface shearing 36,37 , extending the framework developed primarily for unanchored 74 DSS 37 . While this allows the overall interface shear strength between soil and anchored cables to be 75 estimated, it precludes the consideration of passive earth pressure effects commonly observed during soil-76 anchor interaction 29 . 77 We describe here an improved fiber-optic DSS approach for sensing of vertical ground 78 displacements with microanchored strain sensing cables deployed in boreholes. We fabricated three 79 microanchors to enhance soil-cable interlocking effects adding on previous work 37 . We proposed a new 80 criterion for assessing soil-cable coupling based on the geotechnical bearing capacity theory. We 81 examined the effects of confining pressure, soil and interface strength parameters, and anchor type and 82 dimension on the performance of the microanchored DSS system. We demonstrated the feasibility of this 83 improved methodology through elementary testing, physical modeling, and a field experiment conducted 84 in a coastal setting.  (Fig. 1b) where e C is the frequency shift-strain coefficient, T C is the frequency shift-temperature coefficient, and 99 T  is the change in temperature that can be quantified using a colocated temperature sensing cable 100 insensitive to mechanical strains. Because Brillouin backscattering is generated at each point of the fiber, 101 by repeatedly launching light pulses into the fiber a complete strain profile of the deforming strata along 102 the entire borehole length can be mapped. 103 Durability is a central concern for any instrument installed in a subsurface environment. 104 Theoretically, borehole-embedded fiber-optic DSS systems can be permanently used for deformation 105 observation as fiber-optics are inherently corrosion resistant. In practice, fiber-optic cables may break due 106 to large stratum deformation (the ultimate tensile strain of fiber-optics is ~2%, i.e., 20,000 με). Our first 107 borehole DSS system was deployed in Shengze (Southern Yangtze Delta, China) in 2012 24 and strain 108 acquisition has been performed routinely for nearly ten years. We anticipate such systems would survive 109 and function properly for at least several decades; a robust yet strain-sensitive cable is crucial. 110

112
Anchor-like elements are viewed as essential to ensuring sufficient ground-cable coupling and hence the 113 DSS measurement quality can be improved 35 . For this purpose, we fabricated three types of 114 microanchors-disc, cylinder, and spindle. These anchors were attached at discrete points to 115 commercially available fiber-optic strain sensing cables using epoxy resin adhesives. In doing so, three 116 dedicated cables were developed, covering both field and laboratory application scenarios; their features 117 and properties are summarized in Table 1. The disc-anchored cable is well suited for low-confined 118 laboratory physical modeling, as the 0.9-mm-or 2-mm-diameter thermoplastic polyurethane (TPU)-119 jacketed cable (NZS-DSS-C07 by NanZee Sensing Ltd.) can readily be integrated into loose media, 120 owing to its relatively low Young's modulus (E = ~1 GPa), and the discs can enhance considerably soil-121 cable interlocking effects. The cylinder-or spindle-anchored cable utilizes a 5-mm-diameter steel strand-122 reinforced, polyethylene (PE)-jacketed cable (NZS-DSS-C02; E = ~8 GPa). This ensures high survival 123 rates during sensor deployment. Moreover, the small-diameter cylinders and spindles (compared to discs) 124 render the fabricated cables suitable for direct burial installations in field monitoring boreholes. 125 126 Interaction mechanism between soil and microanchored cable 127 Pullout resistance mechanism of bearing microanchored cable. 128 We first elaborated on the interaction mechanism between soil and a buried microanchored cable through 129 a concise theoretical analysis (Fig. 2), which is a first step toward the successful application of the 130 proposed methodology. The analysis builds on the bearing capacity theory presented by Jewell 28 and 131 Bergado et al. 29 , originating from geotechnical engineering. 132 Non-peer reviewed preprint submitted to EarthArXiv

/ 24
During pullout, the resistance of the microanchored cable is composed mainly of two parts (Fig. 2a)  Hence, the modified punching failure mode was employed herein to describe the bearing mechanism of 154 microanchored fiber-optic cables, and q N and c N can be respectively expressed as: 155 Non-peer reviewed preprint submitted to EarthArXiv where  is the soil internal friction angle; k is the lateral earth pressure coefficient; and  is the angle 158 of the rotational failure zone (Fig. 2b). For k = 1 and  = /2, theoretical predictions were found to 159 agree well with laboratory test data 29 , and q N and c N are thus reduced to: 160 π tan q 1 π tan cos 4 2

Validation of bearing resistance equations via laboratory pullout testing. 163
To explore whether the bearing capacity theory is suitable for describing cable anchor failure, we 164 performed laboratory pullout tests on disc-anchored fiber-optic cables at variable anchor diameters. The Iten et al. 39 argued that the contact between soil and a buried anchored cable is a combination of overall 186 bonding and point fixation. Experimental evidence 37 further showed that this combination depends on the 187 Non-peer reviewed preprint submitted to EarthArXiv 7 / 24 deformation stage of the soil-cable interface. Specifically, tube anchors will continue to contribute to the 188 overall interface shear strength after the interface between soil and unanchored segments fails, converting 189 the contact from overall bonding to point fixation. Point fixation may reduce the spatial resolution of 190 DSS 17 , but it is commonly sufficient to obtain a detailed strain profile of subsurface strata. Hence, for 191 ground motion sensing the acquired strain data can be considered as credible provided that the capacity of 192 microanchors has not been reached. In this sense, of particular importance in coupling assessment is the 193 evaluation of stress states of microanchors, especially for those buried in shallow strata. Force equilibrium 194 of a single microanchor yields (Fig. 2c): 195 where a F is the interaction force between the soil and microanchor; f1 F is the friction force; b F is the 197 bearing force; and 1 N and 2 N are the tensions or compressions provided by the unanchored cable 198 segments, which can be calculated using the measured fiber strain: 199  To ensure the quality of field monitored fiber-optic strains, a large ar F value is desirable. A concise 219 parametric analysis was conducted to investigate the influences of normal stress, microanchor type and 220 dimension, and soil and soil-anchor interface strength parameters on ar F . The parameters used in the 221 analysis are listed in Supplementary Table S1. 222 It can be observed that ar F increased with increasing n  or a D , but differed across microanchors 223 ( Fig. 4a,b). Because of anchor side friction, the spindle-shaped microanchor had higher ar F than the other 224 two microanchors, especially at high n  . For field applications, a strain of 1% (corresponding to a a F of 225 14.9 N under the current parameters) is usually taken as the maximum strain value considering the long-226 term working performance of the fiber-optic. This strain limit can be used for determining the minimum 227 microanchor diameter required, which is instructive for the design of cable anchors (dashed line, Fig. 4b). 228 The effects of soil and soil-anchor interface strength parameters on ar F are illustrated in Fig. 4c,d. ar  This experiment was performed in a cylindrical box with an internal diameter of 420 mm and a height of 241 1000 mm (Fig. S2). The model box consists primarily of three segmented plexiglass cylinders with a wall 242 thickness of 10 mm and a height of 300 mm per segment. The bottom of the model box is composed of a 243 square plexiglass plate with a side length of 500 mm and a 100-mm-high plexiglass cylinder (Fig. S3a). 244 Non-peer reviewed preprint submitted to EarthArXiv

/ 24
We used a sand as an analogue for the aquifer and a clayey soil for the aquitard. The specific gravity 245 of the sand is 2.65, the internal friction angle is 32°, and the permeability coefficient is 7.71×10 -2 mm/s. 246 The specific gravity of the clayey soil is 2.73, the liquid limit is 34.4%, the plastic limit is 20.0%, and the 247 plastic index is 14.4. Given the low confining pressure present in the model, we chose to use the disc-248 anchored fiber-optic cable for vertical strain sensing (Fig. S3b). The diameter of the unanchored cable is 249 1.2 mm with a Young's modulus of 1.01 GPa. The diameter of the disc is 50 mm, the thickness is 1 mm, 250 and the spacing is 100 mm. An NBX-6050A BOTDA interrogator (Neubrex, Japan; Fig. S3c) was 251 employed to record at a 50 mm sample interval with a 100 mm spatial resolution; the resulting strain 252 accuracy is ±7.5 με. A settlement gauge was also utilized to measure settlements of soil layers with a 253 measurement range of 0-10 mm and an accuracy of ±0.01 mm (Fig. S3d). 254

Experimental procedure. 255
The physical model was constructed following the procedure described below. Before filling 256 soils in the model box, the microanchored cable was pretentioned (~7000 με) and vertically deployed 257 (Fig. S4a). Note that prestrain of the cable allowed compressive deformation to be measured. A 200-mm-258 thick sand layer, a 300-mm-thick clayey soil layer, and a 100-mm-thick sand layer were then successively 259 compacted in the model box (Figs. S4b,c). The water contents of the sand and clayey soil layers were 260 18.6% and 16.1%, respectively, whereas the compaction densities were 1.68 g/cm 3 and 1.60 g/cm 3 , 261 respectively. To prevent fine particles from flowing into sand layers, a geotextile was laid at the interface 262 between sand and clayey soil layers (Figs. S4d) (2) Recharge. Connect the water valve to the water tank and gradually inject water into the model box. 276 In this process, vertical strains and settlements were also monitored. Similar to drainage, the recharge 277 experiment was stopped after the water level and soil strain were basically stable. 278 Note that in addition to the experiment described above, an additional experiment having an 279 unanchored cable as the distributed strain sensor was also conducted for comparison purposes. 280

Results. 281
Figure 5a-d shows the fiber-optic data measured by the microanchored cable (averaged over the two 282 buried cable segments) at different periods during the drainage experiment. Figure 5a depicts the original 283 Brillouin frequency shifts, which can be converted to strains by multiplying a calibrated frequency shift-284 strain coefficient. After deducting the initial strain measurements, actual strain change curves were 285 obtained (Fig. 5b). Note that negative (or positive) strains denote compression (respectively, tension).  (Fig. 5a,c), the stress condition of each anchor 300 at the above depths was determined (Fig. 5e). The mobilized interaction forces did not reach their 301 maximum values, indicating that the strain data monitored by the microanchored cable were credible. To 302 further validate the fiber-optic strain measurements, we integrated the measured strains to yield the soil 303 layer deformation at 50-600 mm depth, which was compared with the settlement gauge measurements 304 (Fig. 5f). It can be found that both the trend and magnitude of deformation obtained by the two methods 305 were essentially consistent, thus proving the feasibility of microanchored DSS for monitoring vertical soil 306 deformation at a laboratory scale. Notably, strain profiles measured with the unanchored fiber-optic cable 307 can barely reflect the deformation response of the soil layers due to poor data quality (Fig. S5). This could 308 result from slippage between the soil and the bare cable, owing to insufficient soil-cable coupling in a 309 high soil moisture, low-confined environment. These results highlight the role of soil-cable interface in 310 soil deformation sensing and underscore the importance of microanchorage in an unfavorable 311 environment. The shallow strata are composed of loose clay, sub-clay, and medium-fine sand, with a thickness of 200-319 1600 m. 320 In recent years, ground subsidence in Yancheng had become more and more serious due to the 321 unreasonable exploitation of subsurface resources and the construction of high-rise buildings 40 . It was 322 reported that the area with a cumulative settlement greater than 200 mm has reached 10.86 km 2 , with the 323 largest settlement being ~700 mm. In view of this, we employed the fiber-optic DSS technology to 324 examine the deformation characteristics of subsurface strata and help policy makers cope with the 325 subsidence hazard in the region. 326 Monitoring system deployment and data acquisition. 327 In July 2016, a fiber-optic DSS instrumented borehole was constructed in a development zone in 328 Yancheng (33°21'19.38"N, 120°10'36.39"E; Fig. S6). The development zone has suffered from severe 329 subsidence because of extensive construction and subsurface mining activities. The monitoring borehole 330 has a depth of ~240 m and a diameter of 129 mm. The microanchored fiber-optic cable was deployed in 331 the borehole following the procedure described below. 332 Drill a vertical borehole in the selected site and perform hole sweeping and washing using clean 333 water. Thread the microanchored cable into the head of a weight guide (Fig. S7a), and wind the cable on a 334 pay-off reel (Fig. S7b). Slowly lower the weight guide and cable into the borehole by controlling the wire 335 rope attached to the cable (Fig. S7c). Backfill the borehole with the prepared fine sand-gravel-bentonite 336 mixture. Keep the cable in a straightened state during this period. Retain the fixator after borehole 337 backfilling and build a monitoring station to achieve long-term deformation sensing. To evaluate whether the fiber-optic strains were reliable, the stress state of the microanchors at 0−20 357 m depth was analyzed. Figure 6c shows that with the increase of microanchor depth, the degree of 358 mobilization of soil−anchor interaction force decreased dramatically. This is expected because the 359 ultimate force increased significantly with depth. Although the average value of a ar / F F reached 360 approximately 33% for the microanchor at 2 m depth, all these microanchors remained good working 361 condition during the whole process. To further verify the measured fiber-optic data, a comparison 362 between extensometer measurements and fiber optically determined deformation at 0-140, 140-240, and 363 0-240 m depths was conducted (Fig. 6d, Fig. S8). pressures. In this study, we developed an improved DSS approach by using a dedicated fiber-optic cable 374 with microanchors attached to its surface whereby coupling can be improved. We first probed the 375 ground-cable interaction mechanism via theoretical analysis and proposed a bearing capacity-based 376 criterion for data reliability assessment. We then applied the proposed technique to both laboratory and 377 field experiments for the detection of vertical motions. As demonstrated by our results, no buried 378 microanchors failed even at limited confining pressures. We proved the feasibility of microanchored DSS 379 Non-peer reviewed preprint submitted to EarthArXiv 13 / 24 further through comparisons of fiber optically determined deformation with extensometer measurements. 380 We underscore this method's potential for retrieving high-resolution static and quasistatic strain profiles 381 with a single ground-buried microanchored fiber-optic cable. In particular, the improved quality of strain 382 data acquired in the near surface environment may provide new opportunities for geomechanics and 383 hydrology research. Future studies should aim to achieve higher measurement precision of microanchored 384 DSS via evaluating quantitatively the impact of anchorage on the ground-to-fiber strain transfer. 385 Moreover, future work to assess the suitability of proposed bearing resistance equations for a variety of 386  parameters (c, ϕ), and (d) soil-anchor interface strength parameters (ci, ϕi). c D is the diameter of 532 unanchored cable; a F is the interaction force between soil and microanchor (14.9 N corresponds to a 1% 533 tensile strain). Parameters used in the analysis are summarized in Supplementary Table S1.  Table 1. Three microanchored fiber-optic cables developed for deformation sensing in the near surface 552 environment. The optical fiber depicted in the cross-section illustration is comprised of a fiber core (silica 553 core + cladding) and a coating. The unanchored strain sensing cables are commercially available (NanZee 554 Sensing Ltd.): the TPU-jacketed (NZS-DSS-C07); the PE-jacketed (NZS-DSS-C02). Note that no 555 anchor-cable interface debonding was found in any of the applications presented. TPU = thermoplastic 556 polyurethane; PE = polyethylene. Refer to ref. 35 for a cable with special three-dimensional "dead" anchors 557 suitable especially for detection of shear deformation such as a creeping landslide. 558