Taiwan Chelungpu-fault Drilling Project Research Articles

Time‐Dependent Deformation in the Damage Zone of the Chelungpu Fault System and Potential Stress Relaxation

AbstractHigh fracture density in fault damage zones not only reduces the elastic stiffness of rocks but may also promote time‐dependent bulk deformation through the sliding of fracture and thus alter the stress in fault zones. On comparing the damage zones of the three faults in the Chelungpu fault system encountered in the Taiwan Chelungpu fault Drilling Project (TCDP), the youngest damage zone showed pronounced sonic velocity reduction even though fracture density is the same for all three fault zones, consistent with the shorter time for velocity recovery in the youngest fault. Caliper log data showed a time‐dependent enlargement of the borehole wall at the damage zone. These damage zones record lower differential stress than the surrounding host rock, which cannot be explained by the reduced elastic stiffness in the damage zone. Stress relaxation caused by time‐dependent bulk deformation in the damage zone may be responsible for the observed low differential stress.

Lithology and Fault‐Related Stress Variations Along the TCDP Boreholes: The Stress State Before and After the 1999 Chi‐Chi Earthquake

AbstractUnderstanding the stress state before and after an earthquake is essential to study how stress on faults evolves during the seismic cycle. This study integrates wellbore failure analysis, laboratory experiments, and edge dislocation model to study the stress state before and after the Chi‐Chi earthquake. The post‐earthquake in‐situ stress state observed along boreholes of the Taiwan Chelungpu‐fault Drilling Project (TCDP) is heterogeneous due to lithological variations. Along the borehole, we observe that drilling‐induced tensile fractures are only present in sandstones, whereas breakouts are mostly present in silt‐rich rocks. Laboratory experiments on TCDP cores also show that tensile and compressive strength are weaker in sandstones than in silt‐rich rocks. These observations imply that both maximum and minimum horizontal principal stresses are higher in silt‐rich intervals. Extended leak‐off tests in the TCDP borehole also show lower minimum horizontal stress in sand‐rich intervals, consistent with the above observations. We combine these observations to estimate a profile of stress magnitudes along the well which explains the variability of stress states found in previous studies. The stress heterogeneity we observed underlines the importance of acknowledging the spatial scale that the stress data represent. We then use an edge dislocation model constrained by GPS surface displacements obtained during Chi‐Chi earthquake to calculate the coseismic stress changes. Our inferred pre‐earthquake stress magnitudes, obtained by subtracting the coseismic stress change from the post‐earthquake stress, suggest subcritical stress state before the earthquake despite the large displacements observed during the Chi‐Chi earthquake in the region where TCDP encountered the fault.

Changes in paleostress and its magnitude related to seismic cycles in the Chelung‐pu Fault, Taiwan

AbstractPaleostress analysis was conducted through a multiple stress inversion method using slip data recoded for the core samples from the Taiwan Chelung‐pu Fault Drilling Project (TCDP). Two stress fields were obtained; one of these had horizontally plunging σ1, and the other has horizontally plunging σ2 or σ3 in the compressional stress direction of the Chi‐Chi earthquake. Stress magnitude for both the stress fields was constrained by stress polygons, which indicated larger SHmax for horizontally plunging σ1 than that in the case of horizontally plunging σ2 or σ3. These differences in stress orientations and stress magnitude suggest that the change in stress filed can be caused by stress drop and stress buildup associated with seismic cycles. The seismic cycles recoded in the core samples from TCDP could include many events at geological timescale and not only the 1999 Chi‐Chi earthquake.

Ultrafine spherical quartz formation during seismic fault slip: Natural and experimental evidence and its implications

In recent works on the determination of pseudotachylyte within the principal slip zone (PSZ) of the Chelungpu fault (Taiwan), we demonstrated that frictional melting occurred at shallow depths during the 1999 Mw 7.6 Chi-Chi earthquake. Thus, the characteristics of melts are of paramount importance to investigate processes controlling dynamic fault mechanics during seismic slips. We conducted rock friction experiments on siltstone recovered from the Taiwan Chelungpu fault Drilling Project (TCDP) at a slip rate of 1.3m/s and a normal stress of 1MPa. Here we not only target to characterize experimental pseudotachylyte and evaluate the associated frictional behavior, but also compare it with natural frictional melts of TCDP. Our results show that (1) initial shear stress drop was related to the generation of low viscosity melt patches, (2) the evolution of shear stress in the postmelting regime was congruent with frictional melt rheology, and (3) the slip strengthening presumably resulted from the extremely low content of water of the frictional melt. In particular, the state-of-art of in situ synchrotron analyses (X-ray diffraction and Transmission X-ray Microscope) determine the presence of ultrafine spherical quartz (USQ) grains (~10nm to 50nm) in the glassy matrices presumably produced at high temperature. Our observations confirm that the USQ grains formed in rock friction experiments do occur in natural faults. We surmise the USQ is the result of frictional melting on siltstone and represents the latest slip zones of the Chelungpu fault, and further infer that the viscous melts may terminate seismic slips at shallow crustal conditions.

Quantitative modeling of the newly formed magnetic minerals in the fault gouge of 1999 Chi‐Chi earthquake (Mw 7.6), Taiwan

AbstractWhen an earthquake occurs, magnetic minerals are formed in the gouge under the combined action of frictional heating and fluid. Generally, the gouge is altered, and a detailed analysis of neoformed minerals is difficult. The Taiwan Chelungpu fault Drilling Project provided continuous records of the active Chelungpu fault gouges. By analyzing the magnetic parameters along the 16 cm thick gouge which houses the principal slip zone of the 1999 Chi‐Chi earthquake (Mw 7.6), we observed a 4 cm shift between the maximum of magnetic susceptibility and the maximum of remanence. The maximum magnetic susceptibility is localized along the principal slip zone. The main identified magnetic minerals are magnetite and goethite, which have very different magnetic parameters. We propose that the maximum of the concentration of magnetite and goethite corresponds to the maximum of magnetic susceptibility and remanence, respectively. By modeling the concentration of these two magnetic minerals, we explain satisfactorily the profiles of magnetic susceptibility and remanence. This quantitative modeling indicates that ~200 ppmv of magnetite formed in the principal slip zone and its main contact area. Similarly, ~1% of goethite is formed in the center of the gouge, where the fluids are more enriched in iron. We propose that the magnetite and goethite are formed and altered during seismic cycles.

Faulting processes in active faults – Evidences from TCDP and SAFOD drill core samples

The microstructures, mineralogy and chemistry of representative samples collected from the cores of the San Andreas Fault drill hole (SAFOD) and the Taiwan Chelungpu-Fault Drilling project (TCDP) have been studied using optical microscopy, TEM, SEM, XRD and XRF analyses. SAFOD samples provide a transect across undeformed host rock, the fault damage zone and currently active deforming zones of the San Andreas Fault. TCDP samples are retrieved from the principal slip zone (PSZ) and from the surrounding damage zone of the Chelungpu Fault. Substantial differences exist in the clay mineralogy of SAFOD and TCDP fault gouge samples. Amorphous material has been observed in SAFOD as well as TCDP samples. In line with previous publications, we propose that melt, observed in TCDP black gouge samples, was produced by seismic slip (melt origin) whereas amorphous material in SAFOD samples was formed by comminution of grains (crush origin) rather than by melting. Dauphiné twins in quartz grains of SAFOD and TCDP samples may indicate high seismic stress. The differences in the crystallographic preferred orientation of calcite between SAFOD and TCDP samples are significant. Microstructures resulting from dissolution–precipitation processes were observed in both faults but are more frequently found in SAFOD samples than in TCDP fault rocks. As already described for many other fault zones clay-gouge fabrics are quite weak in SAFOD and TCDP samples. Clay-clast aggregates (CCAs), proposed to indicate frictional heating and thermal pressurization, occur in material taken from the PSZ of the Chelungpu Fault, as well as within and outside of the SAFOD deforming zones, indicating that these microstructures were formed over a wide range of slip rates.

Seismic velocity variations at TCDP are controlled by MJO driven precipitation pattern and high fluid discharge properties

Using seismic noise based monitoring techniques we find that seismic velocity variations (dv/v) observed with the borehole array of the Taiwan Chelungpu-fault Drilling Project (TCDP) are controlled by strong precipitation events associated with the Madden–Julian Oscillation (MJO), a dynamic intraseasonal atmospheric pattern in the tropical atmosphere. High-frequency noise (>1 Hz) excited by steady anthropogenic activity in the vicinity of the TCDP allows daily resolution of dv/v time series. Relatively large fluid discharge properties control the equilibration of the ground water table and hence seismic velocities on time scales smaller than the average precipitation recurrence interval. This leads to the observed synchronous 50–80 day periodicity in dv/v and rainfall records in addition to the dominant annual component. Further evidence for the governing role of hydraulic properties is inferred from the similarity of observed dv/v timing, amplitude, and recovery properties with dv/v synthetics generated by a combined model of ground water table changes and diffusive propagation of seismic energy. The lapse time (τ) dependent increase of dv/v amplitudes is controlled by the sensitivity of the diffuse wave field sampled at 1100 m depth to shallower water level fluctuations. The significant vertical offset between stations and water level explains the direct τ dependence which is opposite to the trend previously inferred from measurements at the surface.

S-Wave Velocity Structure of the Taiwan Chelungpu Fault Drilling Project (TCDP) Site Using Microtremor Array Measurements

The Taiwan Chelungpu Fault Drilling Project (TCDP) drilled a 2-km-deep hole 2.4 km east of the surface rupture of the 1999 Chi–Chi earthquake (Mw 7.6), near the town of Dakeng. Geophysical well logs at the TCDP site were run over depths ranging from 500 to 1,900 m to obtain the physical properties of the fault zones and adjacent damage zones. These data provide good reference material for examining the validity of velocity structures using microtremor array measurement; therefore, we conduct array measurements for a total of four arrays at two sites near the TCDP drilling sites. The phase velocities at frequencies of 0.2–5 Hz are calculated using the frequency-wavenumber (f–k) spectrum method. Then the S-wave velocity structures are estimated by employing surface wave inversion techniques. The S-wave velocity from the differential inversion technique gradually increases from 1.52 to 2.22 km/s at depths between 585 and 1,710 m. This result is similar to those from the velocity logs, which range from 1.4 km/s at a depth of 597 m to 2.98 km/s at a depth of 1,705 m. The stochastic inversion results are similar to those from the seismic reflection methods and the lithostratigraphy of TCDP-A borehole, comparatively. These results show that microtremor array measurement provides a good tool for estimating deep S-wave velocity structure.

Clay mineralogy and geochemistry investigations in the host rocks of the Chelungpu fault, Taiwan: Implication for faulting mechanism

The Chelungpu fault, with northward propagating ruptures, was ruptured as a result of the Mw 7.6 earthquake which struck Central Taiwan on 21st September 1999. To understand the consequences of clays in fault gouges, we examined the clay mineralogy and major element geochemistry of the “host rocks” of the Chelungpu fault from four outcrops, shallow boreholes at Fengyuan (455.3m in depth), Nantou (211.9m in depth), and a deep borehole of the Taiwan Chelungpu fault Drilling Project (2003m in depth). Results revealed different degrees of chemical weathering, the intensity, as documented by differences in the smectite content, could be further understood through the relative clay percentage of smectite, illite chemistry index, and illite crystallinity. These mineralogical proxies combined with the chemical index of alteration (CIA) and the intensity of chemical weathering indicate that the degree of chemical weathering is a function of depth, i.e., the most severe on the surface and the mildest in the TCDP samples. The mineralogical and geochemical data obtained in this study also suggest that chemical weathering, rather than leaching, seems to be the main driving force for the phase changes of clays. The amount of smectite produced by chemical weathering varies with depth, and it suggests that the previous idea of weak-fault caused by the presence of smectite in the fault zone will be risky without considering weathering processes. The observations of clay mineralogy and major element geochemistry in this study indicate that the presence of smectite in the outcrops may not play a significant role during faulting, and suggest that fault-weakening as a result of the presence of smectite cannot be applied to the Chelungpu fault.

Anatomy of the high‐frequency ambient seismic wave field at the TCDP borehole

The Taiwan Chelungpu‐fault Drilling Project (TCDP) installed a vertical seismic array between 950 and 1270 m depth in an active thrust fault environment. In this paper we analyze continuous noise records of the TCDP array between 1 and 16 Hz. We apply multiple array processing and noise correlation techniques to study the noise source process, properties of the propagation medium, and the ambient seismic wave field. Diurnal amplitude and slowness patterns suggest that noise is generated by cultural activity. The vicinity of the recording site to the excitation region, indicated by a narrow azimuthal distribution of propagation directions, leads to a predominant ballistic propagation regime. This is evident from the compatibility of the data with an incident plane wave model, polarized direct arrivals of noise correlation functions, and the asymmetric arrival shape. Evidence for contributions from scattering comes from equilibrated earthquake coda energy ratios, the frequency dependent randomization of propagation directions, and the existence of correlation coda waves. We conclude that the ballistic and scattered propagation regime coexist, where the first regime dominates the records, but the second is weaker yet not negligible. Consequently, the wave field is not equipartitioned. Correlation signal‐to‐noise ratios indicate a frequency dependent noise intensity. Iterations of the correlation procedure enhance the signature of the scattered regime. Discrepancies between phase velocities estimated from correlation functions and in‐situ measurements are associated with the array geometry and its relative orientation to the predominant energy flux. The stability of correlation functions suggests their applicability in future monitoring efforts.

An earthquake slip zone is a magnetic recorder

During an earthquake, the physical and the chemical transformations along a slip zone lead to an intense deformation within the gouge layer of a mature fault zone. Because the gouge contains ferromagnetic minerals, it has the capacity to behave as a magnetic recorder during an earthquake. This constitutes a conceivable way to identify earthquakes slip zones. In this paper, we investigate the magnetic record of the Chelungpu fault gouge that hosts the principal slip zone of the Chi-Chi earthquake (Mw 7.6, 1999, Taiwan) using Taiwan Chelungpu-fault Drilling Project core samples. Rock magnetic investigation pinpoints the location of the Chi-Chi mm-thick principal slip zone within the 16-cm thick gouge at ~1 km depth. A modern magnetic dipole of Earth magnetic field is recovered throughout this gouge but not in the wall rocks nor in the two other adjacent fault zones. This magnetic record resides essentially in two magnetic minerals; magnetite in the principal slip zone, and neoformed goethite elsewhere in the gouge. We propose a model where magnetic record: 1) is preserved during inter-seismic time, 2) is erased during co-seismic time and 3) is imprinted during post-seismic time when fluids cooled down. We suggest that the identification of a stable magnetic record carried by neoformed goethite may be a signature of friction-heating process in seismic slip zone.

Lithological control on shear-wave velocity anisotropy in core samples from the Taiwan Chelungpu Fault Drilling Project

The present study provides laboratory constrains on the in situ source of shear wave birefringence in siltstone and sandstone lithologies from the Taiwan Chelungpu Fault Drilling Project (TCDP). Bench top measurements at ambient pressure and temperature are performed across the end faces of vertically drilled samples while the polarization angle of the source and receiver is rotated from 0 to 360° in azimuth. Shear wave birefringence in siltstone is low (∼3.5%) and controlled by low angle (∼30°) bedding dip. The sandstone samples exhibit moderate (∼15%) shear wave anisotropy due to a network of vertical microcracks with normal about the bedding strike direction. As related studies infer that the microcracks responsible for shear wave birefringence in sandstone are also open at depth, our results suggest that anisotropy in borehole measurements can simply be accounted for by the observed siltstone and sandstone microstructures, i.e., without invoking larger scale objects, nor randomly oriented cracks/fractures subjected to triaxial stress state.

Possible stress states adjacent to the rupture zone of the 1999 Chi-Chi, Taiwan, earthquake

We explore regional stress states in the vicinity of the rupture area of the 1999 Chi-Chi, Taiwan, earthquake by reconciling data from the Taiwan Chelungpu-Fault Drilling Project (TCDP) in-situ measurements and earthquake focal mechanisms. Given the background deviatoric stress in the range of 10–50 MPa and the horizontal NW–SE directed maximum principal stress axis, the predicted fault types show strike–slip and normal faulting near the coseismic surface rupture and thrust and strike–slip faulting in central Taiwan. Such predictions are able to fit TCDP in-situ observations in a local scale and aftershock earthquake focal mechanisms in a regional scale. Additionally, the proposed stress state explains remarkable rotations of the maximum stress axes observed near the northern segment of the Chelungpu Fault. This result provides key information for forecasting of consequent earthquakes and evaluation of focal mechanisms after the occurrence of a large earthquake.

ESR熱年代学による断層活動性評価：台湾チェルンプ断層掘削計画Hole Bコアを例として

ESR熱年代学による断層活動性評価：台湾チェルンプ断層掘削計画Hole Bコアを例として

Determining the maximum overburden along thrust faults using a porosity versus effective confining pressure curve

This study aims to quantify the tectonically induced thrusting and erosion. Here, we propose and evaluate a laboratory-based method for determining the maximum effective stress ever sustained by a sedimentary-rock formation. The porosities of nineteen samples collected from two deep boreholes of the Taiwan Chelungpu-Fault Drilling Project were measured under different confining stresses. The relation between measured porosity and confining pressure (compaction curve) is bilinear in log–log scale. The maximum effective stress of the tested rocks was determined from the intersection point of the two straight lines of the compaction curve. The maximum effective stresses of the rocks below the Sanyi Fault (reverse fault) are close to the in-situ vertical effective stress. Above the Sanyi Fault (hanging wall), where the rocks have been thrust, the maximum effective stresses of the rocks are significantly greater than the in-situ stress with an average value of 29.48MPa. That is, the eroded thickness of the formation on the hanging wall of the Sanyi Fault is 2482m, a value close to that determined from the geological profile (2773m). It is inferred that the total displacements along the thrust Sanyi Fault is 11.5km, a value close to that determined from the geological profile (14–16km). These results indicate that the proposed laboratory-based method is potentially useful for determining the thrusting and erosion in a fold–thrust belt and the total displacements along a thrust fault.

Magnetic inference of in situ open microcracks in sandstone samples from the Taiwan Chelungpu Fault Drilling Project (TCDP)

This paper presents a thorough magnetic study of centimeter-sized rock samples from Hole B of the Taiwan Chelungpu Fault Drilling Project (TCDP). Fabrics were obtained from magnetic susceptibility and anhysteretic remanent magnetization are interpreted using whole rock and magnetic mineralogy characterization. We infer the presence in the sandstone lithology of secondary magnetite coating a dense network of microcracks. Using other evidence from ferrofluid saturation experiments, we conclude that this network of microcracks is not sealed but open in situ, and should therefore be incorporated in the interpretation of seismic and borehole scale studies in the TCDP area.

Elastic anisotropy of core samples from the Taiwan Chelungpu Fault Drilling Project (TCDP): direct 3-D measurements and weak anisotropy approximations

The study of seismic anisotropy has become a powerful tool to decipher rock physics attributes in reservoirs or in complex tectonic settings. We compare direct 3-D measurements of P-wave velocity in 132 different directions on spherical rock samples to the prediction of the approximate model proposed by Louis et al. based on a tensorial approach. The data set includes measurements on dry spheres under confining pressure ranging from 5 to 200 MPa for three sandstones retrieved at a depth of 850, 1365 and 1394 metres in TCDP hole A (Taiwan Chelungpu Fault Drilling Project). As long as the P-wave velocity anisotropy is weak, we show that the predictions of the approximate model are in good agreement with the measurements. As the tensorial method is designed to work with cylindrical samples cored in three orthogonal directions, a significant gain both in the number of measurements involved and in sample preparation is achieved compared to measurements on spheres. We analysed the pressure dependence of the velocity field and show that as the confining pressure is raised the velocity increases, the anisotropy decreases but remains significant even at high pressure, and the shape of the ellipsoid representing the velocity (or elastic) fabric evolves from elongated to planar. These observations can be accounted for by considering the existence of both isotropic and anisotropic crack distributions and their evolution with applied pressure.

Thermal and pore fluid pressure history on the Chelungpu fault at a depth of 1111 m during the 1999 Chi-Chi, Taiwan, earthquake

[1] On 20 September 1999 the Ms7.6 Chi-Chi earthquake ruptured the Chelungpu fault in central Taiwan. In 2005 the Taiwan Chelungpu Fault Drilling Project was launched to drill two deep holes (holes A and B) cutting across the fault plane. The heat strength (= 7.0°C m), within a heated layer of ∼5 mm, on the fault at hole A due to frictional faulting is evaluated from the values of shear stress and thermal and mechanical parameters measured from the core samples. Based on a 1-D heat conduction equation and 2-D faulting model, with the values of thermal diffusivity evaluated within a representative temperature range, the thermal and pore fluid pressure history at depths 1110.37–1111.34 m in hole A is constructed. Results show that the peak temperature at the center of the heated layer could have been higher than 1100°C during faulting, and temperature rise decreased with increasing distance and time. The possible relationships between the temperatures and chemical reactions of clay minerals are discussed in detail. In the heated layer, pseudotachylites have been formed and quartz plasticity might also have been operative during faulting. Outside this slip zone, the temperature rise was low and thus clay minerals were stable during faulting. The evaluated pore fluid pressure is 22.5 MPa. This suggests the existence of a suprahydrostatic state in the fault zone during the earthquake. Indirect evidence of the existence of fluids in the fault zone during faulting is described. Additionally, rock physics is applied to interpret the observations.

Temperature estimates of coseismic heating in clay-rich fault gouges, the Chelungpu fault zones, Taiwan

To investigate the coseismic frictional temperature in seismogenic fault zones, we examine the characteristics of clays in the Chelungpu-fault zones with isothermal heating experiments, scanning electron microscope coupled to an energy dispersive spectrometer (SEM/EDX), and thermogravimetry analysis (TGA). In the TCDP case (Taiwan Chelungpu fault Drilling Project), three fault zones of the Chelungpu-fault system were identified at the depth of 1111 m, 1153 m, and 1222 m (described as FZ1111, FZ1153, and FZ1222 hereafter), respectively. The clay mineral assemblages of FZ1111 show evidence of melting, and the temperature in a ~ 2 cm band within the black gouge zone is estimated to be from 900 °C to 1100 °C by comparing the SEM images of in situ natural samples with those of heated materials, and the finding of no recrystallization of kaolinite-amorphous aluminosilicates-spinel in the fault samples. The clay mineral assemblages of FZ1153 suggested that kaolinite has been broken down by the thermal decomposition/dehydroxylation but chlorite has not. The clay characteristics and results of SEM/EDX and TGA constrain the faulting temperature from 500 °C to 900 °C, with a spatial distribution up to ~ 1.3 m. The clay characteristics of FZ1222 indicated that clays were changed by experiencing high temperature acid fluids, instead of thermal decomposition/dehydroxylation processes, and that the temperature is localized in ~ 2 cm and ranges from 350 °C to 500 °C, the lowest temperature among three fault zones. The estimates of temperature ranges, and thermal anomaly intervals among three fault zones provide important information and constraints on the physical and chemical processes, coseismic dynamic weakening mechanism, and earthquake energy budget in the future.

Clay mineral anomalies in the fault zone of the Chelungpu Fault, Taiwan, and their implications

The Taiwan Chelungpu‐fault Drilling Project (TCDP) Hole‐A recovered continuous core samples across the rupture zone of the 1999 Chi‐Chi earthquake (Mw7.6). Studying in‐situ chemical properties sequentially from fresh‐fault‐zone materials of the Chelungpu fault provides insight into possible faulting mechanisms. Distinct anomalies of mineral assemblages at the 1111‐m fault zone of TCDP Hole‐A are found to be: (1) A decrease in clay content within the primary slip zone (PSZ); and (2) A significant decline of illite, disappearance of chlorite and kaolinite, and spike in smectite within the PSZ. Meanwhile, features relating to melting or amorphous material in the PSZ have been observed by SEM and TEM. The results suggest that the PSZ might have experienced generation of glassy materials such as pseudotachylyte by the expense of clay minerals due to strong shear heating, then prompt alteration of pseudotachylyte into smectite. Characteristics of clay minerals and images obtained from electronic microscopes in the PSZ thus imply that pseudotachylyte possibly developed during the 1999 Chi‐Chi earthquake, but quickly altered into smectite. This particular phenomenon may explain why pseudotachylyte is rarely found in exhumed hydrated fault zones.