Borehole Tiltmeters Research Articles

Point stability at shallow depths: experience from tilt measurements in the Lower Rhine Embayment, Germany, and implications for high-resolution GPS and gravity recordings

SUMMARY From 1996 to 1999, we have studied ground tilts at depths of between 2 m and 5 m at three sites in the Lower Rhine Embayment (LRE), western Germany. The LRE is a tectonically active extensional sedimentary basin roughly 50 kmr100 km. The purpose of the tilt measurements was (a) to provide insight into the magnitude, nature and variability of background tilts and (b) to assess possible limitations of high-resolution GPS campaigns and microgravity surveys due to natural ground deformation. The tilt readings, sensed by biaxial borehole tiltmeters of baselength 0.85 m, cover a frequency range from 10 x8 Hz to 10 x2 Hz (periods from minutes to years). Assuming that the tilt signals represent ground displacements on a scale typically not larger than several times the tiltmeters’ baselength, and that tilt signals at shallow depth could in a simple geometric way be related to changes in surface elevation and gravity, we try to estimate the magnitude level of point movements and corresponding Bouguer gravity effects that is generally not surpassed. The largest tilt signals observed were some t50 mrad yr x1 .I f they were observable over a ground section of extension, e.g. 10 m, the converted rates may correspond to about t0.5 mm per 10 m yr x1 in vertical ground displacement, and t0.1 mgal yr x1 in Bouguer gravity effect, respectively. Large signals are mostly related to seasonal effects, probably linked to thermomechanical strain. Other causes of ground deformation identified include seepage effects after rainfalls (order of t10 mrad) and diurnal strains due to thermal heating and/or fluctuations in the water consumption of nearby trees (order of t1 mrad). Episodic step-like tilt anomalies with amplitudes up to 22 mrad at one of the observation sites might reflect creep events associated to a nearby active fault. Except for short-term ground deformation caused by the passage of seismic waves from distant earthquakes, amplitudes of non-identified tilt signals in the studied frequency range seem not to exceed t2 mrad. As the larger tilt signals are close to the precision achieved with modern GPS systems and superconducting gravimeters when converted into height and gravity changes, further enhancement in resolution of these techniques may require simultaneous recording of local ground deformation at the observation sites.

Detectability for GPS, EDM and tiltmeters of the Hot Springs Research Institute in the western Kanagawa prefecture, Japan

To better understanding the observation data, we attempt to assess the detectability for the crustal deformation observation network of the Hot Springs Research Institute of Kanagawa prefecture (HSRI). The HSRI has carried out continuous monitoring of crustal deformation with GPS, EDM, borehole tiltmeters, groundwater leveling since 1989. Our observation purpose is to detect the pre-, co- and post-seismic deformations of the hypothetical M7 class earthquake designated as the “Western Kanagawa Prefecture Earthquake (WKE)”. In order to estimate the detectability, weapplya simple calculation method, thatiscomparing between co-seismic deformation (displacement and tilt) and the accuracy of observation instruments. Co-seismic deformation is calculated using the Okada’s method, which is a formulation of the expected crustal deformation due to seismic faulting as a function of an earthquake magnitude and a hypocentral distance without assuming focal mechanism. It is indicated that the highest detectability area is located in the center of our HSRI observation network. Comparing detectability contour maps for several kinds of networks, we can find that the detectability of the co-seismic deformation by the tiltmeter network is higher than those by the GPS and EDM networks in the case of the M5 class earthquake with a focal depth of 10 km, and that the detectability by the GPS network is almost equal to that by the EDM network in the case of the M6 class earthquake with a focal depth of shallower than 15 km. For M4 class earthquakes, it is difficult for the GPS, EDM and tiltmeter networks to detect co-seismic deformation. The present result shows advantageous characteristics of the HSRI geodetic network and an important constraint on investigating the pre-, co- and post-seismic deformation of the WKE. To monitor the pre-, co- and post-seismic deformation of the WKE, it is necessary to combine these observations in the most effective way.

Coswarm and preswarm crustal deformation in the eastern Izu Peninsula, central Japan

Remarkable crustal deformation has repeatedly been observed in the eastern Izu Peninsula, central Japan, synchronized to the occurrence of extensive earthquake swarms in the past 20 years. Among the instrumentation, a borehole tiltmeter installed at station KWN, which is located within a few kilometers of the swarm region, has faithfully recorded large signals of ground tilt associated with more than 10 swarm sequences in the past 10 years. In addition to coswarm signals it became clear that precursory tilt signals preceded major swarm activities. The amplitudes of the precursory tilt signals are of order of 0.1 μrad, and they precede the onset of major swarms by several hours to half a day. This signal is convincing because the same phenomena appeared at least 7 times and were also detected by other independent observations. We speculate that a dike intrusion at depth causes crustal deformation and only as it reaches depths shallower than ∼10 km does seismic activity begin. The precursory tilt signal can be utilized for practical prediction of the start of swarm activity in the eastern Izu Peninsula.

Tidal tilt observations in the Netherlands using shallow borehole tiltmeters

A small scale experiment with borehole tiltmeters above a gasfield in the Netherlands is described. Pressure variations in the gasfield may cause subsurface deformations leading to tilt. We describe the technique of the installation of the tiltmeters and the influence of the surface conditions on the measurements. For this case study we quantify the minimum pressure variation which is required to result in observable, induced tilt. Tidal tilt observations are shown and used for the verification of the measurements. Finally, we compare the observations with computed ocean tide loading tilt for the M2 tidal component.

Monitoring ground deformation at the decade volcano Gunung Merapi, Indonesia

Pressure changes within a volcano lead to deformations of his edifice. To discriminate local and regional ground movements, four deformation stations with clusters of continuous borehole tiltmeters have been installed at the hillsides of Gunung Merapi. In addition, recording local environmental parameters allows the recognition of interfering local tilt signals. First analyses of the data records show: The significant tidal variations of the tilt signal validate the successful coupling between the instruments and their surroundings. Transient tilt signals correlated to rainfalls are probably caused by poroelastic deformation of the hillsides. Four volcanic events occurred during the observation period; all of them were accompanied, if not triggered, by heavy rainfalls. Tilt anomalies, considerably larger than the rain induced tilts, were recorded around the event of October 31, 1996, coincident with an extraordinary high seismic activity. If this activity reflects internal fracture processes prior to the eruption, the accompanying deformations might partially explain the observed tilt signal.

Studies of in Situ Pore Pressure Fluctuations At Various Scales

Pore pressure fluctuations in fluid saturated geological formations, either of natural or anthropogenic origin, can be observed at different scales. Natural fluctuations, e. g. , due to tidal, barometric or seismogenic forcing, or man-made effects as through use of underground fluid reservoirs, or initial filling and cyclic loading of lake reservoirs may have wavelengths from meters to kilometers. In situ monitoring of processes, in which both rock deformation and pore pressure changes are significant, improves our knowledge on the mechanical behaviour and the role of pore pressure in porous rocks and sedimentary layers. Pressure transducers for continuous recording of fluid level variations in wells, reflecting pore pressure changes at depth, or borehole tiltmeters that are sensitive to ground deformation caused by gradients of pore pressure fluctuations are relatively simple means to trace the dynamics of such rock-fluid interactions. The obtained data series are usually interpreted in two ways: by application of analytical solutions-adopting homogeneous poroelastic conditions or single fracture models in a uniform, elastic medium-and by simulation through numerical calculations allowing for some heterogeneity in the model volume. Field cases presented in this article include tilt measurements in the vicinity of pumped wells (1 to 100 m scale), fluid level monitoring in wells (borehole scale), and studies of pore pressure effects induced by seismic events (1 to 100 km scale). Specific rock parameters that can be constrained are the Skempton ratio, the hydraulic diffusivity, and the type of the effective rheology. In cases of tiltmeter studies, anisotropy of pore fluid flow can also be detected. Keywords: fluids in rocks, pore pressure, poroelasticity, hydrology.

An investigation into shallow borehole tiltmeters

Shallow borehole tiltmeters are frequently used for ground deformation monitoring in volcanic areas, where they are usually installed within a few meters of the surface. A major problem concerns the noise affecting the installation at shallow depth where large tilt and strain are caused by temperature effects. The long term stability of these instruments and the best shallow depth of installation, in terms of best cost/benefit, are the crucial points. We conducted an experiment in order to compare the signals of three shallow borehole electronic tiltmeters installed in the same place at different depths. We verified the signal reliability in revealing the long term slow deformation and looked into the reduction of the temperature effects with depth. We present three years of data and discuss the limits and the advantages of the different installation depths.

Comparison of long-period tilt measurements: results from the two clinometric stations Metsähovi and Lohja, Finland

The records of an ASKANIA borehole tiltmeter and two water tube tiltmeters are compared. The borehole tiltmeter was operating at the Geodetic Observatory Metsähovi in a 63 m deep borehole and the long water tube tiltmeters are installed in a limestone mine at Lohja, both observatories are some 15 km apart west of Helsinki, Finland. Regarding the tides the comparison yields a coherence within 0.5% . 3% for the diminishing factors, which proves their regional validity within the error bars. Air pressure induced tilts in both stations reach about 10 msec of similar signals, pointing into the same direction of about 150°N. They are caused by the horizontal air pressure gradients as well as by the local air pressure variations in interaction with the coast line of the nearby sea (Gulf of Finland). In the borehole, hydraulic induced tilt signals are reaching the amplitudes of the tides (about 20 msec), even 20 m below an aquifer in solid granite. Reproduceable short-period (minutes to hours) pore pressure induced tilts are caused by pumping ground water. These effects can be explained by poroelastic modeling and thus, they can be partly corrected. They are corresponding to the main directions of permeability in a system of clefts. The seasonal drift directions in the borehole point to pore pressure variations in the aquifer interacting with a nearby swamp, correlated with the freeze. As the physics of the hydrological and air pressure induced tilts can be explained, instrumental effects are excluded at both stations. Thus, continuous tilt measurements are suitable not only for the investigation of small scale but also for regional scale crustal dynamics. But in order to avoid local disturbances the borehole tiltmeter needs to be installed deeper than now, in boreholes of 100 . 500 m depth.

Interpretation of long-period tilt records at BlåSjø, southern Norway, with respect to variations in the lake level

The impoundment of BlåSjø in Southern Norway was accompanied by repeated precise levelling and microgravity measurements, seismic studies, and deformation monitoring with borehole tiltmeters. The results of the continuous tilt measurements are presented here. The tiltmeters were installed in 40 m depth at distances of about 12 km and 20 km west of BlåSjø. The finite-element modeling of the tilt changes caused by the water load revealed that the geological faults are crucial for the deformation pattern. Due to a higher elasticity strain is transformed to north-south direction. During the recording period of about two years elastic deformation dominated. Expected retarded deformations as a result of slow changes in underground pore pressure were not significant. This is proved by numerical modeling of fluid migration by pore pressure changes along the fault system which shows time delays of only some days instead of months or even years.

Measurement and interpretation of tidal tilts in a small array

Strain‐induced tilts caused by variations in material properties can produce large perturbations in the tilt field near a material discontinuity. This is evidenced in data from a closely spaced array of borehole tiltmeters that has been operating at Piñon Flat Geophysical Observatory in southern California since 1987. The array is composed of three borehole tiltmeters located at depths of 24, 36, and 120 m. The tiltmeters have a sensitivity of a few nanoradians. Tidal measurements show differences between the boreholes of up to 50%. The largest percent difference occurs in the semidiurnal band where the tilt field measured in the 36‐m borehole differs by ∼10 nrad at the M2 frequency. This difference has been modeled as a strain‐induced tilt caused by the weathering layer.

Outstanding student paper award

Mary Kohl's presentation, “Using a Short‐Baseline Borehole Tiltmeter to Measure the Regional Tilt,” was named the Outstanding Student Paper in the Geodesy Section at AGU's 1992 Fall Meeting in San Francisco, Calif.Kohl is a doctoral student in physics at the Joint Institute for Laboratory Astrophysics at the University of Colorado. Her research is in experimental geophysics, and her dissertation will focus on understanding and interpreting tidal tilt measurements in southern California. Kohl received a B.S. in physics from Union College, Schenectady, N.Y., in 1985.

Measuring Low Frequency Tilts.

A borehole tiltmeter with a sensitivity of a few nanoradians is described. It is composed of two orthogonal horizontal pendulums with free periods of 1 s. The pendulums are insensitive to barometric pressure fluctuations, and the measured temperature coefficient is less than 30 nrad/°C. The range of the pendulums is about ±5 μ rad, and their response is linear within 1% and stable over several years. The performance of the tiltmeter in the field was evaluated using tidal data obtained from a closely spaced array of boreholes in Southern California. The long-term stability of the tiltmeter is generally better than 1 μ rad/yr. The data also indicate that instruments in boreholes at least 24 m deepare independent of surface effects. Several different capsules designed to couple the instrument to the surrounding material have been tested. In addition, an experimental method for estimating the magnitudes of local perturbation in the regional tilt field is described.

Interpretation of the tilt response of two tiltmeters to water injection into a tar sands formation

Near-surface tilt associated with fracture growth caused by water injection into a tar sands formation during an 18-day period was monitored by two biaxial mercury-level borehole tiltmeters. A nonlinear inversion method has been used to determine the fracture dimensions and position of its centre for five data groups from within that time. The inversion is limited by having data from only two tiltmeters, and a horizontal rectangular fracture must be assumed. The fracture expands rapidly initially, and more slowly as time progresses. The area of the fracture increases from approximately 5000 to about 240 000 m2 during the 18 days of injection and its centre migrates approximately 50 m. Key words : tiltmeters, tilt response, enhancement monitoring.

A tilt and seismicity episode in the New Hebrides (Vanuatu) Island Arc

Tilt and seismicity have been monitored in the central New Hebrides island arc since 1978 using bench mark arrays, long tube water tiltmeters, borehole tiltmeters, and a local seismometer network. Releveling of the bench mark array on Efate island in late November 1986 revealed a 10 μrad tilt up to the NNW since the previous leveling in April 1986. The tilt event was preceded by a magnitude 5.9 thrust event that occurred on October 25, 1986, at a depth of 48 km and about 11 km NW of the tiltmeter instruments. Six days later, a shallow (<20 km) swarm of earthquakes occurred 12 km NNW of the tiltmeter instruments and 5 km north of the epicenter of the magnitude 5.9 earthquake. Closely coincident in time with the swarm, a 5 μrad tilt up to the NNW that occurred over a period of 5 days was recorded on both the 100 m baseline water tube tiltmeter and the borehole bubble level tiltmeter. A composite focal mechanism of 191 earthquakes selected from the swarm indicates a thrust mechanism with some component of strike slip. Calculations show that the seismic slip associated with a swarm of this magnitude is apparently inadequate to cause the observed surface deformation. Two similar shallow swarms in November 1987 and July 1988 have occurred within 15 km of the 1986 swarm but with no apparent surface deformation. The most likely explanation, supported by simple modeling, is that the swarm and tilt are the result of a magmatic intrusion from island arc volcanism. An alternate hypothesis is that both the seismicity and the tilt are due to an episode of largely aseismic creep in the upper crust.

Dyke intrusion model for the 1989 seismovolcanic activity off Ito, central Japan

In the last 10 years, extensive earthquake swarms have occurred intermittently in the region east of Izu Peninsula, central Japan. In July 1989 an active swarm again took place off Ito City, and among all past activities, it was nearest to land and shallowest. This activity eventually became a weak submarine eruption, the first of its kind in this region. Associated with this event, a borehole tiltmeter at 92 m depth, installed several kilometers from the source area, detected distinct continuous tilt change with the following features: (1) Tilt of 20 μrad was observed within 1 week preceding the eruption. (2) Tilt change showed a good correlation to swarm activity. (3) No obvious coeruption tilt change was observed. On the basis of the tilt record and swarm distribution, we constructed a dislocation model to interpret the seismovolcanic activity. The model consists of two tensile faults, each of which corresponds to magma intrusions in May and July, and a right‐lateral reverse fault representing the largest earthquake (M5.5) at 1109 LT on July 9. This model can simulate not only the above tilt data but also the record of a nearby dilatometer, the results from daily laser distance measurements and daily Global Positioning System monitoring, as well as results from other conventional distance surveys and leveling surveys in the surrounding region. From this model, we conclude that the intersection of the shear fault with the tensile fault caused magma to come up into soft sedimentary layers at the seabottom, creating a lava dome. Later an instability brought about a magma‐phreatic explosion which did not generate any substantial crustal deformations at deeper levels.

Near-surface tilt response to steam injection into a tar sands formation

Two biaxial mercury-level borehole tiltmeters located at moderate depth (20 m) and 91 m horizontally distant from the injection well have been used to monitor the effects of a fast hydraulic fracture and subsequent steam injection in a tar sands formation at a depth of 230 m. Tilt vectors are determined for the maximum tilts during the fracture, and the long-term tilt migration associated with the steam-injection process is monitored. The tilt associated with the fast fracture is of the order of 1–2 μrad, and the long-term tilt increased as much as 240 μrad over the 500 day monitoring period, and appeared to approach a limit. The long-term tilt migration generally follows the same orientation as the initial tilt due to the fast fracture.

Tilt monitoring at wairakei geothermal field

The withdrawal of fluids from the Wairakei geothermal field has induced substantial deformation of the ground surface, with subsidence rates as great as 400 mm/yr. Substantial changes to pressures within the upper steam zone of the reservoir would be expected to be recognised in corresponding changes in the deformation rate and in particular in the tilting. Two types of tiltmeter have been tested for this purpose at Wairakei geothermal field. A 5 month test of borehole tiltmeters showed short duration tilt differences as great as 35 μrad (microradians) between meters 5 m apart. This effect is believed to be caused by differences in local response to regional subsidence. Over 5 months of testing the tilt rates from two borehole meters were in good agreement with surface tilt rates derived from a local levelling network (520 μrad/yr). The short-term test of a two-component bubble tiltmeter mounted on a horizontal concrete block gave results that were dominated by a large amplitude (150 μrad) diurnal tilt variation attributed to the response of the block to air temperature variations. However the repeatability of the diurnal response suggests that the tilt rate could be determined from as little as 3 days' observations provided temperatures of the site and instrument were also monitored. In the absence of site temperature data, daily temperature extremes (measured at about 1 km from the site) were used to eliminate the temperature-dependent response. The resulting secular tilt rate (measured over 36 days) of 950 ± 130 μrad/yr, is larger than the long-term estimate determined from precise levelling surveys.

Tilt observations using borehole tiltmeters: 1. Analysis of tidal and secular tilt

We have designed a borehole tiltmeter using two horizontal pendulums which have periods of 1 s. We have installed the instruments at seven sites in Colorado and Wyoming to evaluate the secular tilt, the tides, and the coherence between nearby instruments. Using 28 days of data from Boulder, Colorado, the standard deviations of the tidal estimates are 5% for the semidiurnal component M2 and 9% for the diurnal component O1. The estimates agree with models that include the body tide, the ocean load, and the topographic correction to better than the estimated uncertainty. Tidal measurements at Erie, Colorado, have larger, possibly nonrandom variability that may be caused by a coupling between the tides and long‐period tilts. The coherence between nearby instruments is ≳0.5 for frequencies ranging from 0.5 to 2 cycles per day, but drops to a much smaller value outside of this range. The secular tilt rate is quite variable and ranges from 0.1 to 1 μrad/yr. Measurements at Erie, Colorado, and in Yellowstone National Park, Wyoming, exhibit an annual or biannual periodicity with an amplitude of about 1 μrad, which is consistent with a 1/f2 extrapolation of the diurnal noise power to longer periods.

Tilt observations using borehole tiltmeters: 2. Analysis of data from Yellowstone National Park

We have installed borehole tiltmeters at five sites in Yellowstone National Park, Wyoming, and have used these instruments to measure the spatial variation of the amplitude and phase of the principal semidiurnal tide. The measured tides vary both with position and azimuth and differ from the sum of the body tide and the ocean load by up to 50%. The difference predicted by a finite element model constructed from seismic, refraction, and gravity data has a maximum value of only 12%, although the discrepancy between our observations and the model is only marginally significant at some sites. The disagreement between the model and our observations is much larger than we observed using the same instruments at other sites and cannot be attributed to an instrumental effect. We have been unable to modify the model to explain our results while keeping it consistent with the previous observations.

Twelve years of geophysical research in the Charlevoix seismic zone

Various and extensive geophysical monitoring began in 1974 in the Charlevoix Seismic Zone (CSZ). The historical record and a major earthquake in 1925 characterize this zone as the likely site of a future damaging earthquake (i.e. M = 5.5 or larger) and the aim of the monitoring was to develop the capability of predicting such earthquakes. The parameters studied included: microseismicity, seismic travel times, electrical impedance, vertical movement, horizontal movement, tilt, gravity change, and strain through water well level changes. Although no clear precursor was detected in the months before the largest seismic event that occurred during the observation period ( M = 5.0), the multiparameter experiment provided new insights into the structure and the mechanics of this active region. Observations of several hundred well-located microearthquakes have permitted the delineation of ancient rift faults that are now being activated by high horizontal compression. The geographical extent of the zone is sharply defined by the microseismicity which occurs in an area of 35 km × 80 km straddling the St. Lawrence River. Observation of P-waves from accurately-timed explosions and S-waves from earthquakes support the hypothesis of Extensive Dilatancy Anisotropy (EDA) involving saturated parallel cracks in the crust. The data have permitted an estimate of both crack density and the direction for the principal compressive stress in the region (N 30° E). Changes in electrical parameters with periods of months to years were detected at five of the six magnetotelluric stations installed in the region. Changes in the polarization angle of the earth's electrical (telluric) field correlated well (for seven years) with calculated groundwater accumulation levels at one site. Electric impedance changes showed no clear correlation from station to station, but changes at one station appear to correlate with a seismic P-wave travel time change anomaly and with the disappearance of a negative gravity anomaly at the time of an unusual number of earthquakes in the range M = 2 to 3 from mid 1976 to mid 1977. It is suggested that these correlated changes could be related to aseismic stress redistribution and fluid flow along one of the NE striking rift faults about two years before the M = 5.0 earthquake in 1979. Continuous monitoring of tilt (surface and borehole) and well-water levels was carried out at a single site (Charlevoix Geophysical Observatory). Although there were unexplained near-surface tilts of the order of the earth tide amplitude within two or three days before some earthquakes, they were either not spatially coherent or were related to small unexplained temperature changes in the tiltmeter vault. No intermediate-term (one week to one year) precursors were seen against a noise level of 1–3 μrad (RMS) for near-surface tilt observations and a noise level of 0.2–0.3 μrad (RMS) for borehole tilt measurements at a depth of 50–100 m. The threshold of detectable tidal tilt admittance changes for the borehole tiltmeters was lowered to 2% but no significant changes were seen. Four coseismic changes in the levels of water in boreholes were observed (all were drops in water level) which were much larger than expected on the basis of elastic dislocation theory. It is suggested that these could be explained either by the collapse of a dilatant zone around the respective hypocentres or by triggered aseisnuc movement on a nearby active fault. Borehole water level tides observed from 1980 to 1986 exhibit significant variations in amplitude and phase which are thought to be the result of time-varying elastic properties in a nearby structure. Some constraints can now be placed on the long-term deformation of the CSZ as a whole from tiltmeter, levelling and horizontal surveys. Data from a levelling array observed from 1976 to 1986 puts a limit of 0.54 μ rad/yr on the secular tilt rate at the observatory: a further limit of 0.1 μ rad/yr is provided by borehole tilt measurements over the period 1983–1986. A limit of 2 cm is placed on vertical movements northwest of the St. Lawrence River by levelling surveys in 1965, 1978, 1980 and 1982. No significant (90% confidence) deviatoric strain has been detected across the CSZ from horizontal control surveys carried out in 1912–1919, 1965, 1978 and 1983: the maximum allowable long-term strain rate across the CSZ from 1919 to 1983 is 0.16 μ strain/yr.