T41F-1275 0800h
The New Geophysics in a Crack Critical-System
Seismic shear-wave splitting indicates stress-aligned fluid-saturated cracks, in almost all in situ rocks, that are so closely spaced they are critical systems. Evolution of fluid-saturated cracks in changing conditions can be calculated by anisotropic poro-elasticity (APE): the deformation mechanism is fluid-movement along pressure-gradients between neighboring grain-boundary cracks, flat pores, and pore throats at different orientations to the stress field. Difficult to confirm directly, because of the inaccessibility of deep-rocks, there are at least 20 different phenomena that are approximately matched by APE-modeling. Using earthquake swarms as shear-wave source, systematic changes in shear-wave splitting are seen, with hindsight, before about a dozen earthquakes, and the time and magnitude of a M5 event in SW Iceland was successfully stress-forecast. In the most direct confirmation, modeling fluid-injections into hydrocarbon reservoirs, the response of high-pressure and low-pressure CO2-injections were matched almost exactly by inserting the actual injection pressures into APE. This sensitivity is confirmed at the first borehole Stress-Monitoring-Site (SMS) in a transform zone in Iceland. With spectacular 0.02ms sensitivity, 10% variations in shear-wave splitting are observed, in shear-wave propagation over 300m at 500m-depth, correlating with small-scale seismicity at 70km with a total energy equivalent to less than one M$\sim$3.5 event. This high resolution is achieved by using the highly-repeatable Downhole Orbital Vibrator (DOV) as a source of shear waves. The physical implications are that microcracks are so closely spaced that they verge on fracture criticality and failure. Hence rock is so weak to shear-stress that stress accumulation occurs over enormous volumes: 100s of thousands to billions of cubic km before the largest earthquakes. This means that the approach to criticality and failure occurs very widely (witness the large distances that changes in shear-wave splitting are observed). The eventual failure typically occurs along pre-existing faults where there is some form of stress-relaxation before the impending earthquake. The implications are that SMSs can routinely monitor stress-accumulation before earthquakes.
http://www.glg.ed.ac.uk/~scrampin/opinion/
T41F-1276 0800h
Evolution of Fault Systems and its Associated Geomorphic Structures: Strike-Slip and Dip-Slip Fault Model Test and Field Survey
Sandbox experiments were performed to investigate evolution of fault systems and its associated geomorphic structures caused by strike-slip and dip-slip motion on basement faults. A 600 cm long, 250 cm wide, and 60 cm high sandbox and a 200 cm long, 40 cm wide, 25 cm high sandbox were used in a strike-slip fault model test. Computerized X-ray tomography applied to the sandbox experiments made it possible to analyze the kinematic evolution, as well as the three-dimensional geometry, of the faults. The deformation of the sandpack surface was analyzed by use of a laser method 3D scanner, which is a three-dimensional noncontact surface profiling instrument. In the dip-slip fault test, a 332.5 cm long, 200 cm high, and 40 cm wide sandbox was used. The fault type, fault dip, fault displacement, thickness and density of sandpack and grain size of the sand were varied for different experiments. Field survey of active faults in Japan and California were also made to investigate the evolution of fault systems and its associated geomorphic structures. A comparison of the experimental results with natural cases of active faults reveals the following: (1) In the left-lateral strike-slip fault experiments, the deformation of the sandpack with increasing basement displacement is observed as follows. a) In three dimensions, the right-stepping shears that have a "cirque" / "shell" / "ship body" shape develop on both sides of the basement fault. The shears on one side of the basement fault join those on the other side, resulting in helicoidal shaped shear surfaces. Shears reach the surface of the sand near or above the basement fault and en echelon Riedel shears are observed at the surface of the sand. The region between two Riedels is always an up-squeezed block. b) Lower-angle shears generally branch off from the first Riedel shears. c) Pressure ridges develop within the zone defined by the right-stepping helicoidal shaped lower-angle shears. d) Grabens develop between the pressure ridges. e) Y-shears offset the pressure ridges. f) With displacement concentrated on the central throughgoing fault zone, a liner trough developed directly above the basement fault. R1 shears and P foliation are observed in the liner trough. Such evolution of the shears and its associated structures in the fault model tests agrees well with that of strike-slip fault systems and its associated geomorphic structures. (2) Low-angle and high-angle reverse faults commonly migrate basinward and rangeward with time, respectively. With increasing normal fault displacement in basement, normal fault develops within range after reverse fault has formed along range front. (3) In the fault model tests, the horizontal distance of surface rupture from the basement fault normalized by the height of sandpack (W/H) does not depend on the height of sandpack and grain size of sand. The values of W/H from the fault tests agree well with those of earthquake faults in alluvium.
T41F-1277 0800h
Detecting Faults and Shear Zones With GPS and Other Survey Data
I have found a method that seems to let us determine the orientation and nature of shear zones and, in some cases, faults with GPS and other survey data. I will describe the idea in two dimensions. 1) Let us define an ideal fault trace as a line at the ground surface across which there is a discontinuity in displacement. The component of the deformation gradient, $\frac{\partial s}{\partial N}$, is infinite. In this abstract, {\it s} is current position of points parallel to the fault or shear zone and {\it N} is original position of points normal to the fault or shear zone. 2) Let us define an ideal shear zone at the ground surface as a belt across which either: (a) the first derivative, $\frac{\partial }{\partial N}$, normal to the shear zone, in the deformation-gradient component $\frac{\partial s}{\partial N}$, that is, $\frac{\partial^2 s}{\partial N^2}$, is infinite or (b) the second derivative of the component, $\frac{\partial^3 s}{\partial N^3}$, is infinite. 3) A very narrow shear zone can be considered a fault if the component $\frac{\partial s}{\partial N}$ is very large. In practice, our survey is in terms of arbitrary coordinate systems---the initial $(X,Y)$ and current $(x,y)$ systems---rather than in terms of natural coordinate systems of the fault or shear zone---$(S,N)$, $(s,n)$ and $\alpha$, where $\alpha$ is the clockwise angle between {\it X} and {\it S}. One computes the deformation gradient tensor, $F_X_Y$, with the survey data and then determines the natural coordinates by maximizing the quantity $\frac{\partial s}{\partial N}$ as a function of $\alpha$. Then one uses quantities such as $\frac{\partial^2 s}{\partial N^2}$ and $\frac{\partial^2 s}{\partial \alpha \partial N}$ to recognize shear and displacement discontinuities.
T41F-1278 0800h
Stable and Critical Noncohesive Coulomb Wedges: Exact Elastic Solutions
The theory of critically tapered Coulomb wedge has been successfully applied to model active fold-and-thrust belts or submarine accretionary prisms. Brittle mountain building is episodic in nature, controlled by changes in basal friction, erosion and sedimentation, and hydrogeology. Sediment accretion may be modulated by great subduction earthquakes. Between deformation episodes and/or during transition between compressional and extensional tectonics, the Coulomb wedges are stable (i.e., supercritical), to which the critical taper theory does not apply. In this work, we provide an exact elastic solution for stable wedges based on Airy stress functions. The stress equilibrium equation and definition of basal friction and basal and internal pore fluid pressure ratios are exactly the same as those used for Dahlen's [1984] exact solution for critical noncohesive Coulomb wedges, but internal friction $\mu$ becomes irrelevant. Given elastic - perfectly Coulomb-plastic rheology, for stresses in a wedge on the verge of Coulomb failure there must co-exist a critical taper solution involving $\mu$ and a unique equivalent elastic solution not involving $\mu$. Our elastic solution precisely reduces to Dahlen's critical taper solution for critical conditions. For stable conditions, normal stress perpendicular to the surface slope $\sigma_z$ and shear stress $\tau_{xz}$ are identical with those in a critical taper, but the slope-parallel normal stress is different. The elastic solution is also generally applicable to purely elastic wedges and useful for modeling geodetic observations. A stable noncohesive Coulomb wedge differs from a general elastic wedge in that its upper and lower surfaces stay at zero curvature during loading. Dahlen, F.A. (1984), Noncohesive critical Coulomb wedges: An exact solution, JGR, 89, 10,125-10,133.
T41F-1279 0800h
Pattern and Development of Deformation Bands Near Strike-Slip Fault Junctions
Deformation bands (DBs) are an important class of strain localization in porous granular materials. Here we report observations of cataclastic DBs formed in Navajo Sandstone from the Sheets Gulch area of the San Rafael swell of southeastern Utah. Two mutually cross-cutting orientations of DBs located between the junctions of four faults (of similar orientations to the DBs) consistently lack slip surfaces, unlike the bounding faults that have them. The DB array defines an intermediate principal strain axis rotated 20° to the southeast from bedding normal, consistent with early strike-slip deformation banding, faulting, and bedding rotation within the thrust fault stepover beneath the Waterpocket monocline. Subequal and synchronous development of both orientations (S68°W, 63° and S18°W, 62°) of DBs in the area suggest that neither set is older and dominant, which is inconsistent with DB formation at a strike-slip stepover. Further, the bounding faults define a nearly orthogonal network instead of an echelon array. Instead, the DB array may represent an exposure of broad-scale strain-hardening between larger-displacement strike-slip faults consistent with block rotations and macroscopic (regional-scale) flow as the monocline grew.
T41F-1280 0800h
Growth and Displacement-Length Scaling of Grabens: Examples of Fault Restriction and {\it D-L} Saturation on Mars
A fault in a layered sequence grows with a constant {\it D-L} ratio until restricted from further down-dip propagation by a change in the mechanical properties of the sequence. The transition from a "small" to "large" fault is marked by a change from a peaked to a flat-topped displacement profile. We measure the throw distribution on faults in an exceptionally well exposed population of grabens in northeast Tharsis on Mars. We find a transition from peaked to flat-topped profiles as a function of fault length. The graben-bounding faults with flat-topped profiles exhibit a relatively constant displacement ({\it D-L} saturation), indicating these faults are restricted by a mechanical boundary within the Martian crust. Faults less than 25 km in length exhibit peaked (triangular) displacement profiles and {\it D-L} ratios consistent with those of other measured normal faults on Mars ($\sim$1$\times$10$^{-3}$). Graben-bounding faults greater than 60 km clearly exhibit flat-topped displacement distributions, with a relatively constant throw of $\sim$45m (there is a lack of data on faults between 25 and 60 km). Restriction likely controls the nearly constant graben spacing. The effects of interaction and fault linkage can be seen within the displacement profiles. Many of the faults have asymmetric displacement distributions, with the point of maximum displacement skewed towards the central segments of the grabens of Tantalus Fossae (to the south). Some of the grabens with skewed profiles are in close proximity and appear to be in the process of linking with other grabens along-strike. Reductions in the amount of displacement can be correlated with faults that have linked but have not yet accumulated enough displacement to erase all evidence of this linkage. Changes in graben character, such as an increase in width, correlate with an increase in displacement. Additionally, our preliminary data suggest that as one follows the segmented grabens to the south (away from their distal edges), the stress perturbations caused by the faults overcome the strength of the mechanical barrier and displacements increase until the graben-bounding faults are again restricted by a deeper mechanical boundary.
T41F-1281 0800h
Relationship between surface ruptures and subjacent source faults
Concepts of setting source fault parameters from surface rupture due to active faults are discussed for predictions of strong ground motion. Estimation of ground motion plays important rule for the prevention of earthquake hazards. From recent developments in waveform inversion analysis of fault rupture processes through large earthquakes, it is found that strong ground motion is strongly affected by fault geometry and slip heterogeneity. At the prediction of strong ground motions for scenario earthquakes by active faults, the initial parameters of source faults, such as fault length, direction and dip are thus necessary to be determined. Among the parameters used for strong ground motion simulation, fault length and dip are derived from information related to surface fault ruptures. Because these parameters take significant effect on the simulation, much care is required for geological investigation to construct source fault model. On the other hand, the development of study on surface faulting (rupture), recurrent interval and probability of occurrence are estimated from assessment of surface rupture. Surface fault morphology provides fault length, its segmentation, displacement, direction and dip. However, relationship between the parameters of the surface rapture and those of the subjacent source fault is still a problem. We first studied correlation between surface faults and subjacent source faults, such as basic geological estimation of the source fault derived from surface fault ruptures. We collected information of surface ruptures (length, dip, displacement and morphology) and results of waveform inversions (source fault length, dip and distribution of asperities) in twelve large earthquakes. They are being carefully investigated from the view point of geology and seismology. The relation among the intended parameters shall be presented in our poster.
T41F-1282 0800h
GPR Surveying for Paleochannels at the Green Valley Fault
We conducted a ground-penetrating radar (GPR) survey on the Green Valley Fault, located on the NE margin of San Francisco Bay, in an effort locate paleochannels to better constrain the slip rate of the fault. We began this study with a detailed topographic survey to determine the most likely paths of paleochannels, and situated the GPR lines so as to cross them. The GPR survey consisted of nine lines recorded using 50 MHz antennas, providing a total footprint of 100 $\times$ 180 m. The GPR data indicate possible paleochannels 5-10 m wide, at 1-5 m depth. The top surface of GPR basement, Sonoma Volcanics, is the most continuous reflector in the data. It is clearly imaged on several of the GPR lines and can be easily correlated on adjacent lines. It outcrops at the surface near the western boundary of the survey area and dips eastward to reach depths of 20 m at the eastern boundary of the survey area. We also recorded a 70 m long seismic refraction line near the middle of the survey area using a 24-channel seismograph and a sledgehammer source, and derived a three-layer model from the data. At the location of the seismic line, the top layer is 3-5 m thick with $V_P \approx$ 300 m/s, the second layer extends to 20 m depth, with $V_P \approx$ 1700 m/s, and the deepest layer is Sonoma Volcanics, which extends beneath 20 m depth, with $V_P \approx$ 6000 m/s. The geophysical response of paleochannels is derived from the material contrast of sandy channel deposits vs silts and clays. We used a hand auger to sample to 5.3 m depth near the center of the survey area, and encountered mostly fine-grained sediments, clayey and sandy silts. We did not intersect any obvious channel sands. A modern stream that passes within a few 100s of m of the survey site provides an analog useful for estimating potential channel size and sediment types of paleochannels. The modern channel is approximately 5 m wide and contains a bedload of sand and gravel with cobbles and occasional small boulders. Sandy point bars are present. We expect similar coarse-grained sediments associated with paleochannels to be present in the survey area, and to be detectable with GPR and seismic methods. Additional seismic lines and auger boreholes should confirm whether or not channel-like features of the GPR data are truly paleochannels.
T41F-1283 0800h
Possible Geometry and Implications for Potential Blind Thrusts Beneath the Marin County - Mt. Tamalpais Region
The cause of high elevations in the Marin County - Mt. Tamalpais region is enigmatic. Because the region is adjacent to a segment of the San Andreas fault with essentially no transpression, fault-normal shortening is unlikely to be the cause of the uplift. Rather it has been proposed that the uplift may be associated with blind thrust(s) in the area (Furlong and Kirby 2004, this meeting). If such structures do indeed exist, then their extent and orientation would be a key component in assessing earthquake potential for the area. This study examines the specific geometries of possible blind thrusts that could produce the Mt. Tamalpais uplift, and the seismogenic implications. The study area, bounded by the Point Reyes - Tomales Bay segment of the San Andreas (on the west) and the Hayward/Rodgers Creek fault system (on the east), increases in elevation significantly from north to south. Mt. Tamalpais sits at the southernmost extent of the uplifted area with the highest elevations at 780 meters above sea level. We have utilized 2-D and 3-D elastic deformation models to evaluate the orientation and spatial extent of possible blind thrusts that could generate the uplift. The combination of potential fault area and rate of deformation (shortening across the structure) allows us to place limits on the possible size of earthquakes that could occur on such blind faults. Preliminary analyses indicate that structures in the Marin County - Mt. Tamalpais could host up to a M = 6 event with a recurrence time of 1000 years.
T41F-1284 0800h
Structural Inversion of the Palos Verdes Fault, Southern California, and its Implications for Seismic Hazards Assessment
The Palos Verdes Fault (PVF) defines the western margin of the Los Angeles basin, and is regarded as a likely source of moderate to large earthquakes that would affect the coastal metropolitan regions of southern California. In most hazard compilations, the PVF is generally considered to be a vertical, predominantly right-lateral, strike-slip fault system that extends continuously from the Santa Monica thrust southward across Santa Monica Bay, crossing the Palos Verdes Peninsula and continuing southeast across the Inner Borderlands to the area of Coronado Banks. A restraining bend where the fault dips steeply to the southwest generates uplift and folding of the Palos Verdes Peninsula. However, previous studies documenting the activity, slip rate, and slip sense of the PVF have used shallow subsurface excavations and high-frequency seismic data, which have generally limited observations to the upper kilometer of the crust. We use an extensive grid of petroleum industry seismic reflection data and well logs to define the three-dimensional subsurface geometry of the PVF in the region south of the Palos Verdes Peninsula. Our seismic data cover the complete offshore extent of the fault, from Santa Monica Bay to the Coronado Banks, and provide direct constraints on the fault geometry extending down to about 5km depth. We use the shapes of folded strata imaged in the seismic data and penetrated by wells to invert for permissible geometries of the fault as it extends to the base of the seismogenic crust. Our data and structural analyses indicate that the PVF developed by Pliocene inversion of a Miocene normal fault system. The fault has a significant component of reverse slip and southwesterly dip at depth along its extent. Oblique displacement on the fault appears to be partitioned at shallow levels into nearly pure right-lateral strike slip on near-vertical faults and contractional folding above gently to moderately dipping blind-thrust fault splays. These observations are used to define a realistic 3D geometry of the PVF, to define the sizes, shapes and spatial relationships of fault segments that may rupture in earthquakes, and to extend shallow slip and slip rate estimates from previous studies to depth along the fault. This fault model will provide improved forecasts of the possible size and frequency of large earthquakes on the PVF, and will provide more accurate geometric fault representations that can be used to predict strong ground motions resulting from these events.
T41F-1285 0800h
Fluid overpressures on the San Andreas Fault following the passage of the Mendocino Triple Junction
Fluid pressures significantly greater than hydrostatic have been hypothesized to account for the weak nature of many large plate-boundary faults. However, on the San Andreas Fault, the hypothesized subsurface processes which could create, sustain, and potentially localize such pressures over millions of years are not well understood. In this study, we use two-dimensional finite element models of coupled fluid flow and heat transport perpendicular to the fault to evaluate hypothesized mechanisms for generating elevated pore pressure. The models account for transient changes in crustal geotherm and thickness of the seismogenic crust in response to the passage of the Mendocino Triple Junction. Theoretical curves of whole-rock fluid content as functions of pressure and temperature allow us to calculate fluid sources due to metamorphic dehydration within the Franciscan m\'{e}lange as a function of depth and thermal history. Average fluid sources in the seismogenic crust range from 10$^{-18}$ to 10$^{-16}$ s$^{-1}$ over the 15 Myr spanned by our models. We consider a variety of permeability distributions within the models, including a range of homogenous permeability and depth-dependent permeability. We also consider heterogeneous permeability distributions reflecting fault properties and geologic features such as serpentine sills. Our results show that over 15 Myr, thermal expansion of pore fluids due to initial burial, followed by additional heating during exhumation, can create significant overpressures. In addition, models which include fluid sources from metamorphic dehydration of the Franciscan m\'{e}lange result in pore pressures approaching a significant fraction of lithostatic. Generally, all model results show overpressures extending several kilometers to each side of the fault. Due to the continual nature of many of these processes, overpressures are sustained for millions of years without the need for complex and/or extremely low-permeability seals. Models which include geologic structures such as serpentine sills and fault core and damage zones allow only limited localization of overpressures within the fault zone, thus offering a possible mechanism for regional crustal weakness, but not for localized fault weakness.
T41F-1286 0800h
Spatial Localization of Moment Deficits in Southern California
The balance between interseismic elastic strain accumulation and coseismic release defines the extent to which a fault system exhibits a surplus or deficit of large earthquakes. We calculate the regional moment accumulation rate in Southern California based on a slip rate catalog estimated from a block model of interseismic deformation constrained by GPS measurements. The scalar moment accumulation rate, $17.8 \pm 4.6 \times 10^{18}$ Nm/yr, is approximately $50%$ larger than the average moment release rate over the last 200 years. Differences between the accumulated and released elastic displacement fields are consistent with moment deficits that are localized in three regions: the southern San Andreas and San Jacinto faults, offshore faults and the Los Angeles and Ventura basins, and the Eastern California shear zone. The moment budget could be balanced by coseismic events with a composite magnitude of M_W=8.
T41F-1287 0800h
Drainage Basin Evolution Along the San Jacinto Fault, Southern California
Uplifted, unconsolidated sediments, progressively exposed to incision, provide a means to evaluate drainage development and tectonic activity in the San Timoteo badland region, southern California. Morphometric analysis of these developing drainages along the San Jacinto fault zone indicates that slip is stepping away from the topographic escarpment and is accommodated to the southeast by the interaction of the Casa Loma fault and an unnamed, intrabasinal fault. Motion along the right-lateral San Jacinto fault through a restraining bend resulted in the uplift of these sediments adjacent to a topographic bedrock high. Further displacement has juxtaposed these uplifted sediments with an extensional graben. This has resulted in increased relief, thus enhancing incision and drainage formation. Morphometric properties of the drainages (density, frequency, bifurcation ratio, first-order stream slope), drainage basins (area, length, relief, relief ratio, surface to planimetric ratios, denudation) were measured for 14 drainage basins flowing to the southwest. In addition, hypsometric integrals were measured for selected basins, and SL Indices were determined for the axial streams in these drainages. The drainage density, first-order stream length, and stream branching character of these developing drainage basins all suggest that the model of drainage evolution of initiation, elaboration and reduction is applicable to this setting. The stream branching character of the SJ drainage basins also indicates progressive development with distance from the topographic high. First-order stream slope, and the pattern of SL Index values both suggest that there is an increasing relative uplift to the southeast, and defines the region affected by subsidence of the structural pull-apart basin. A decrease in denudation rate corresponds to increasing fault complexity, and suggests that southeast of this, fault slip is occurring primarily between the Casa Loma and an unnamed intrabasinal fault, and has thus shifted away from the Claremont fault. Bifurcation ratios, first-order stream slopes, and SL indices respond to tectonic uplift, while drainage density and frequency, bifurcation ratios between first and second order stream segments, and first order stream length provide an indication of drainage development as a function of time.
T41F-1288 0800h
Evidence for Quaternary Slip on a Low Angle Normal Fault: Searles Valley, CA
Low angle normal faults have been documented in extensional terranes worldwide, however conventional models of fault mechanics preclude slip on planes dipping less than 30 degrees. The global catalogue of earthquake focal mechanisms reveals very few occurrences of seismicity (active slip) on low angle structures, lending support to mechanical arguments against active slip on shallowly dipping planes. Recent field studies of low-angle normal faults in the Baja California and Death Valley regions of western North America, however, suggest that active slip on these structures may be more common than typically thought. Here we investigate the relationship between high angle alluvial scarps in Searles Valley and a low-angle detachment fault in order to determine if displacement on the detachment is active. The network of young and recent fault scarps along the eastern margin Searles Valley can be broadly separated into two primary segments with overlapping fault tips that form a range-scale relay zone in the vicinity of Sand Canyon. South of this relay, the active trace of the fault is marked by a series of graben developed within Late Pleistocene - Holocene alluvial fans. Within the bedrock of the Slate Range, and immediately along-strike of the graben system, is a west dipping, low-angle fault system (Sand Canyon `thrust' - Smith et al., 1968). This fault is continuous within the range for some 20 km and links with a west-dipping normal fault near Manly Pass, and is thus thought to have accommodated west directed normal-sense displacement during Plio-Quaternary time (Andrew and Walker, 2002). Mapping and structural observations at the intersection of the active fault system and the Sand Canyon fault reveals that high-angle scarps cutting Pleistocene alluvium root into a low-angle (10-15°), west-dipping gouge zone. Faults do not significantly displace the detachment surface, and thus, scarp-forming displacement must have been accommodated by slip on the detachment itself. We combine high-precision differential GPS surveys of fault displacement with existing chronology gleaned from a range of Late Pleistocene - Holocene lacustrine deposits related to Searles Lake to develop estimates of fault slip rate along the length of the fault system. Slip rates provide insight into both the local question of how slip is partitioned across the Sand Canyon relay zone, as well as the more regional question of how deformation within the Eastern California Shear Zone is accommodated.
T41F-1289 0800h
More Evidence for Young Tectonism Along the Saline River Fault Zone, Southern Mississippi Embayment
Within the southern Mississippi embayment the 315-striking Saline River fault zone (SRFZ) was recently recognized from linear river segments, aligned epicenters, fault and fold exposures, sand blows, seismic reflection profiles, and trench investigations. In our latest work, shallow S-wave reflection profiles were acquired across the SRFZ within a sand blow field in Holocene alluvium at the southeastern limit of the mapped SRFZ. Along the central part of the SRFZ our field surveys of linear river segments subparallel to the SRFZ reveal that these segments are associated with faulting, fracturing, and folding of Eocene, late Pleistocene, and Holocene strata. In addition, we used logs of coal exploration holes to map shallow Eocene structure across the central SRFZ. In the SE of our study region, two S-wave reflection profiles (parallel and overlapping) were acquired along a 0.6 km N-S transect of the fault zone in NE Ashley County, AR. They reveal the SRFZ has a flower structure geometry with both positive and negative elements. The Eocene/Quaternary contact is clearly seen at 0.3 s (30 to 40 m depth). Several faults cut through the Quaternary section, and Quaternary and Eocene strata are folded harmonically. A fault in the overlap zone can be shown to strike 289 deg. New luminescence (IRSL) dates constrain a significant sand blow episode near this locality to between 7690 and 5725 yrs BP and a later significant episode to younger than 6845 yrs BP, consistent with previous results of middle to late Holocene sand venting here. In the central portion of the SRFZ, previous investigations of a linear river segment revealed a faulted anticline deforming alluvium younger than 720 years old. Our latest river surveys reveal that other linear river segments are associated with fractures and folds in Eocene substrate and young alluvium. Some oblique structures suggest a subsidiary strike-slip array. Shallow subsurface structure in Eocene strata mapped from coal exploration logs across the central portion of the SRFZ in Cleveland County, AR suggest a restraining-bend anticline/horst near the site of a moderate 1911 earthquake and the young river-bank anticline. This positive structure strikes NE, at a high angle to the SRFZ strike. Growing geologic, geomorphic, and geophysical evidence documents young tectonism in the southern Mississippi embayment, specifically in association with the SRFZ. Characterization of the SRFZ and any similar fault zones is crucial to understanding the seismotectonics and seismic hazard of the southern midcontinent.
T41F-1290 0800h
Temporal change in permeability of the Nojima fault zone after the 1995 Hyogoken-Nanbu earthquake
Three boreholes were drilled near the Nojima fault, which the 1995 Hyogoken-Nanbu earthquake occurred on. In order to research the properties and the fault healing process of the fault, the water injection experiments were conducted in every three years. In this report, we researched the permeability of the fault as a measure of the fractured state of the fault. The pore water pressure changes in rock due to the water injection experiments were observed as the discharge changes from the borehole or the groundwater level changes at the borehole. Using the numerical calculation, the permeability of the fault fracture zone was estimated at each experiment. The permeability has been decreasing as time passed and the fact is thought to show the fault healing process of the Nojima fault after the 1995 Hyogoken-Nanbu earthquake.
T41F-1291 0800h
Late Cenozoic and active transpression along the Dead Sea fault in northwestern Syria
The left-lateral Dead Sea fault (DSF) constitutes the boundary between the Arabian and African plates as they converge with Eurasia. In northwestern Syria, the DSF emerges from the 200-km-long "Lebanese" restraining bend with a single fault trace that bifurcates at the Ghab Valley. Despite locally transtensional features like the Ghab Valley, neotectonic activity along the northern DSF in northwestern Syria demonstrates that oblique plate motions result in an overall transpressive tectonic regime. Constraints on recent tectonism are provided by neotectonic mapping, trenching of Holocene sediments, and analyses of a 20-meter pixel digital elevation model constructed using InSAR. Evidence of Neogene and Quaternary displacement on the northern DSF includes truncation and offset of a large, early Pliocene volcano. Preliminary estimates of the left-lateral slip rate south of the Ghab Valley are 4 - 7 mm / yr. In the Ghab Valley, hanging valleys, beheaded drainages, and displaced late Quaternary lava flows demonstrate that plate motion is distributed among several active fault branches. Furthermore, warping and tilting of a late Miocene - early Pliocene paleo-surface, as well as morphometric analyses, suggest that tectonic uplift of the Syrian Coastal Range has been coincident with recent left-lateral faulting on the adjacent DSF. Uplift is asymmetrically distributed in that it is almost exclusively located in the western block of the DSF. The region of uplift is greatest adjacent to the Ghab Valley. This may reflect a contribution from isostatic uplift as a result of the local transtension. We suggest that a convergent component of plate motion is responsible for uplift of the entire Syrian Coastal Range. Our hypothesis is consistent with regional plate tectonic models that predict 10° - 25° obliquity between the Arabian-African plate motion and the strike of the northern Dead Sea fault.
T41F-1292 0800h
A Long-Term Slip-Rate Study Along The North Anatolian Fault, Eksik, Turkey Using Cosmogenic $^{36}$Cl
"Are fault loading and strain release rates constant in time and space?" This is a fundamental but as-yet unanswered question in active tectonics. In order to assess the constancy of fault loading rates, it is necessary to determine slip rates at a variety of locations along the fault and at a variety of time scales. GPS (global positioning system) is a common tool for obtaining a snapshot of the rate of strain accumulated on a fault. Herein we describe our determination of the slip rate on a millennial scale from the North Anatolian fault near the village of Eksik, Turkey, where, during the summer of 2004, we mapped a series of dextrally-offset fluvial terraces. The Eksik site is an ideal location at which to determine a long-term fault slip-rate because of the relative simplicity of the fault system at this location and the availability of datable materials at the site. At Eksik, the surface trace of the North Anatolian fault trends east-west, and is crossed by near-perpendicular, south-flowing drainages. At present, these drainages are actively incising a sequence of fill terraces comprising predominantly limestone cobbles. We mapped a total of three terrace surfaces (T1-T3), differentiated by elevation, surface morphology, and lithology. Detailed geomorphologic mapping and aerial photo analysis of the offset terraces along the fault, as well as a highly detailed total-station topographic survey, allow us to constrain the minimum offset of terrace T3 to 43 $\pm$ 3 m. In order to date these terrace surfaces, we collected limestone samples for $^{36}$Cl cosmogenic nuclide analysis. We collected both surface samples and a profile of sub-surface samples to a depth of 1 m. We have dated 10 of the surface samples from terrace T3, both north and south of the fault. Terrace T3 is removed from sources of fluvial and colluvial sediment input, and therefore surface samples should accurately characterize the deposition age. Nine out of the ten samples returned remarkably consistent ages of 1,700-2,400 years before present. A combination of these age determinations with our estimate of total offset yields a fault slip-rate of $\sim$22 mm/yr. This value is in close agreement with recent GPS rates and with the few other geomorphic slip-rate determinations on the North Anatolian fault. Paleoseismologic data from a nearby high-resolution site (Sugai et.,al 1998) suggest that the 43 m of slip we measured accumulated during 5 earthquakes. Thus, the slip rate averaged over these few events approximately equals the current rate of elastic strain accumulation along the North Anatolian fault.
T41F-1293 0800h
Neotectonics and Evolution of the Yenicaga Basin, Bolu - Turkey
The Yenicaga Basin, located along the North Anatolian Fault System, is interpreted to be a fault-wedge basin with the North Anatolian Fault System's master strand, the Gerede Fault, cutting across the basin itself. The basin and its surroundings contain both paleotectonic rock units and neotectonic rock units. Paleotectonic units, which are deposited or formed during prior tectonic regimes, comprise several formations. The most important of these is the Upper Miocene-Lower Pliocene Eskipazar formation which plays an important role in the understanding of the evolutionary history of the basin. The main Neotectonic unit deposited under control of the present tectonic regime is the Plio-Quaternary Betemurlu formation. The Betemurlu formation unconformably overlies the paleotectonic Eskipazar formation throughout the study area and the unconformity separating these two units corresponds to the time interval during which the paleotectonic stress regime changed into the neotectonic stress regime. Thus, the onset age of the strike-slip neotectonic regime in the study area is Late Pliocene (~ 2.6 Ma). Common basin-margin-bounding faults of the Yenicaga Basin are the Asagi Kuldan fault, the Aksu fault, the Izmirli fault set, the Saraycali fault, the Degirmen fault set and the Hamzabey fault set. These fault systems display well-preserved fault scarps in places. Morphological expressions of these faults and their geometrical relationships with the local stress regime indicate that these faults are mainly strike-slip and oblique-slip faults. Morphotectonic expressions of the faults exposed within the study area indicate that these faults remain active. Most of the population centers within the study area are located on water-saturated, loose basin fill near the active faults. Hence, these population centers are open to future earthquake hazards.
T41F-1294 0800h
Sources of the AD 551, 1202 and 1759 earthquakes (Lebanon and Syria)
The sources of three large (M$\sim$7.5) Near East earthquakes - in July, AD 551, May 1202 and Nov. 1759 - remain controversial, because their mesoseismal areas overlap, straddling the three sub-parallel active faults of the Lebanese restraining bend. Paleoseismic trenching in the Yammo\^{u}neh basin yields unambiguous evidence both for slip on the Yammo\^{u}neh fault in the 12th-13th centuries AD, and for the lack of a posterior event. Only two seismic events are visible on both walls, in the uppermost 80 cm of the trench. Based on the calibrated ages of 14C samples, the latest ground-breaking earthquake occurred between AD 1008 and 1345. The only possible candidate for this event is the 1202 earthquake, since macroseismic damage for other large Near East events was clearly located either well south (AD 1033) or well north (AD 1157 and 1170) of the Beqaa. The penultimate event in the trench has a 14C-calibrated date between AD 324 - 537 and AD 802 - 1001. Such dates cannot be used to rule out that the AD 551 event took place on the Yammo\^{u}neh fault. However, the 551 event is famous for having ruined most of the seaports on the Lebanese coast and having caused a large tsunami which wiped out Beirut. The recent discovery (SHALIMAR cruise, 2003) of fresh seismic scarps related to oblique thrusting on the seafloor offshore Beirut makes it more likely that rupture of one segment of the Mount Lebanon thrust system caused the AD 551 earthquake and tsunami. Thus, we propose that each of the three earthquakes discussed originated on a distinct fault: AD 551 on the Mount Lebanon thrust, 1202 on the Yammo\^{u}neh fault, and 1759 on the R\^{a}cha\"{i}ya-Serghaya fault. Our conclusion regarding the last two events is further supported by a comparison of the freshest visible seismic scarps, which indicates more recent slip on the R\^{a}cha\"{i}ya-Serghaya system than on the Yammo\^{u}neh fault. Regarding the latter, the trenching results suggest that a recurring 1202-like, M$\sim$7.5 event might be due this century, as part of a sequence similar to that of AD 1033 - 1202, whose beginning might already have been heralded by the 1995, Mw$\sim$7.3 Aqaba earthquake. Clearly, a thorough re-assessment of seismic hazard in Lebanon, and on the entire Levant fault, is overdue.
T41F-1295 0800h
Interdisciplinary Approach to the Study of Some Cover-Collapse Sinkholes Located in the Acque Albule Basin,Italy. Preliminary Attempt to Find Relationship Between Collapsed Areas, Seismicity, Data From Satellite Techniques and Field Observations
The Acque Albule travertine basin is located in the Lazio region, Central Italy, on the margin of the Tyrrhenian sea. Geologically the basin has been interpreted as a pull-apart tectonic system driven by right strike-slip faults N-S trending that gave origin to subsidence zones bordered by normal faults. Sinkholes phenomena affecting this area have been reported since the beginning of the last century. In the period June-July, 2001, a low magnitude seismic sequence occurred in the northern Acque Albule basin, in a portion of Guidonia and Tivoli towns. Even though the events had a low magnitude (less than 3.0), local phenomena represented by rumbles and strong ground shaking, frightened inhabitants. Six months before (January, 25th 2001) in Marcellina village, situated few kilometres far from Guidonia, a big sinkhole (60 m of diameter) appeared without fore signals. Recent collapses (June-July 2004) have also affected built-up areas so that the entire basin is now monitored and object of multidisciplinary studies aimed to understand the complex seismotectonic and morphotectonic setting of the area. The analysis of the possible correlations between sinkhole collapses and local shallow seimicity has been integrated by data collected in the field and data acquired by satellite platforms. In particular the Permanent Scatterers technique has been applied to perform accurate measurements of ground displacement velocities. The analysis of spatial data retrieved in the period 1992-2000 has allowed the identification of several highly subsident areas directly related to the presence of sinkholes.
T41F-1296 0800h
Active Stress Field Analysis: Example from Central and Southern Apennines (Italy)
We present two examples of active stress field analysis conducted by integrating geological, geophysical and statistical data. The areas selected are Citt\'{a} di Castello-Sansepolcro in central-northern Apennines and the Val d'Agri-Melandro-Pergola in southern Apennines. Although low seismicity is recorded since the instrumental era (1980 to present), the areas have been repeatedly struck by moderate to large earthquakes in the historical time (since 461 B.C.). This suggesting that they may contain earthquake sources but are in a quiescent period. Our studies are focused on the characterization of the active stress field in these regions and on the definition of the spatio-temporal distribution of the earthquakes. Then, our data contribute to improve the knowledge of the seismogenic behavior of the areas and provide useful information for seismic hazard evaluation. In order to detect the pattern of the active stress field, we analyzed -the earthquakes recorded by national and local seismic networks in the period 2001-2002, with particular attention to the sequences; - ten deep wells (down to 5.5 max depth) for borehole breakouts analysis. At last, we have used two non-parametric statistical procedures (Tanner and Wong, 1984; Faenza et al., 2003) to characterize the spatio-temporal distribution of large historical earthquakes and to account tectonics-physics parameters that can influence the spatio-temporal variability. The results show that the areas are characterized by: i. tectonic structures favorably oriented with the active stress field oriented N44-a$\pm$18-a in the southern Apennines and N50-a$\pm$17-a in the central Apennines; ii. stress regime with a mainly extensional kinematics; iii. cluster distribution of seismicity; iv. the probability that an earthquake with the M$-Y$5.5 will occur in the next 10 years is about 40%.
T41F-1297 0800h
Experimental Analysis of Hybrid Fracture in Berea Sandstone
Previous triaxial extension experiments investigating the transition from extension fracture to shear fracture in low porosity, polycrystalline Carrara marble demonstrate abrupt changes in strength and a continuous transition in fracture orientation and morphology with increasing confining pressure, Pc. New tests on Berea sandstone investigate the same transition in a porous aggregate. Notch cut cylinders (30 mm neck diameter) of Berea sandstone (18% porosity, 0.15 mm average grain size, 80% quartz, 20% feldspar, and trace rutile and kaolinite) were extended in a triaxial apparatus from 0 to 160 MPa confining pressure at a rate of 20 $\mu$m/s. Stress at fracture is characterized by the least compressive principal stress, S3, and maximum compressive principal stress, S1 (S1=Pc). An abrupt change in fracture strength at Pc=50 MPa corresponds to a change from pure macroscopic extension fracture to mixed-mode opening and shear (hybrid) fracture. Within the extension fracture regime, S3 at failure becomes slightly more tensile with an increase in Pc, unlike the constant tensile strength observed for marble. Within the hybrid and shear fracture regimes, S3 at failure becomes more compressive with an increase in Pc. The angle between the fracture surface and S1 increases continuously with Pc, consistent with the marble results. In both rock types, hybrid fractures appear as linked, stepped extension fractures; the length of extensional segments decreases with increasing pressure. The abrupt change in failure strength at the transition from extension to hybrid modes in both rock types likely reflects the increase in mean stress that suppresses the propagation of extension fractures, and the interaction between closely-spaced stepped cracks. In the extension fracture regime, the different dependence of fracture strength on Pc for sandstone and marble may reflect differences in grain scale deformation mechanisms.
T41F-1298 0800h
Grain-scale Numerical Models of Mechanical Particle Size Reduction
When unconsolidated granular material is stressed (e.g. during sediment burial, or shear of gouge-filled faults) it can deform by a combination of particle rearrangement and grain fracture or crushing. The nature of these two mechanisms govern the particle size evolution, spatial distribution of grains and stresses, chemo-mechanical processes and strength characteristics of the material (e.g. rate of diagenesis, yield strength, frictional properties). We use discrete element numerical models to explore the relationships of grain-scale heterogeneity, grain breakage and rearrangement to macroscopic material strength, and intergranular porosity. We then compare our results to observations from laboratory deformation experiments on unconsolidated sand and simulated fault gouge. Our models consist of 70-1500 circular grains packed between rough rigid walls. We apply normal and shear stresses to induce slow deformation rates. The forces on each grain are calculated from elastic, viscous and frictional interaction laws at the grain contacts. When a grain experiences a critical distribution of contact forces, we break it into a set of seven new, smaller grains. We have examined two criteria for grain failure: (1) tensile stresses in the grain interior, which typically occur when a grain is subjected to two opposing contact forces, and (2) large stress concentrations at grain contacts, which are more likely in grains subjected to many contact forces.. We deformed granular systems under both hydrostatic and simple shear loading. During shear, the rate of grain breakage increases to a maximum at small strains and then decreases with continued shear. If failure criterion (1) is used, breakage occurs preferentially in small grains because surviving large grains have more contacting neighbors. With failure criterion (2), breakage is more uniformly spread across all particle sizes. Systems deformed under hydrostatic compression show compaction and grain breakage rates that vary non-linearly with the imposed stress, in good agreement with laboratory deformation tests on unconsolidated sand. Further, the numerical simulations allow for separate analysis of strains induced by grain-scale elastic compression, particle rearrangement and grain crushing.
T41F-1299 0800h
Relations Between Microstructure and Strength in Carrara Marble During Semi-brittle Flow
Deformation in metasediments during orogenic events is often localized within limestone and marble formations. Such faults and shear zones in carbonate rocks are likely to be important in determining overall rock strength, in the partitioning of strain, and in setting the style of deformation. Fortunately, there is an extensive data base bearing on the mechanical properties of carbonate rocks in general, and of Carrara marble in particular. This body of work suggests that at laboratory strain rates, and temperatures from room temperature to 800K, deformation occurs by a mixture of mechanical processes including microfracturing, mechanical twinning, and dislocation creep. Unfortunately, many aspects of the rheology are still poorly understood, and a thorough correlation between the micromechanical processes and rock strength has not been established. We completed a suite of conventional triaxial mechanical tests over a broad range of conditions: 300 to 900K, confining pressures of 50 to 300 MPa, and strain rates of 10$^{-3}$ to 10$^{-6}$ s$^{-1}$. The tests extend and overlap conditions in previously published work and provide a matrix of samples that can be used to examine the relationship between microstructure and rheology. Because the material is continually hardening as it deforms, it seems clear that some microstructural variable(s) important in determining mechanical state are changing; candidates include crack density, twin density and spacing, grain size, or some aspects of dislocation structure. Both the overall strength and the rate of hardening of the rock increase with decreasing temperature and increasing strain rate. Such characteristics suggest that a rate-dependent process like dislocation creep is important. However, the overall strength and the hardening rate are also pressure dependent, at least within the temperature range from 300 to 700K. Thus, strain must also involve local dilatancy or microfracturing. Interactions among the three mechanisms occur and are probably mechanically important. In addition to the well-known correlation between peak strength and dislocation density, detailed transmission electron microscopy suggests that dislocation patterning occurs at all conditions and that interactions between mechanical twins and dislocations are an important part of the evolution of the microstructure. Qualitatively, micro fracture density decreases with increasing temperature, but the correlation between specific surface area and confining pressure or strain rate is not as strong.
T41F-1300 0800h
Prediction of fault-related damage zones in porous granular rock using strain energy density criteria
In granular geologic materials such as porous sandstone, fault-related damage zones are formed by deformation bands, which are tabular discontinuities characterized by pore space collapse and shear. We present results of a study in which separate strain energy density-based criteria are used to successfully predict the tendencies for the nucleation and for the propagation of deformation bands in a classic outcrop of fault-related damage zones within the brittlely-deformed Jurassic Wingate sandstone exposed in the Laramide-aged Uncompahgre Uplift, in western Colorado. The separate distributions of volumetric and distortional strain energy density are calculated for the geometry and stress state of the causative Laramide-aged thrust fault displacements from boundary element calculations of the attendant slip-induced local stresses. Volumetric strain energy density predicts the tendency for deformation band nucleation, the growth stage at which the deformation bands are defined by pore space collapse. Deformation band propagation, where shear occurs along the band, is predicted by distortional strain energy density. The relative magnitudes of elevated volumetric and distortional strain energy density are correlated with deformation band intensity (i.e. the mapped fracture intensity). Within a damage zone, enhanced deformation band nucleation tendencies are predicted and observed to occur within the upper hanging wall and ahead of the causative thrust fault, as well as along the frictionally-slipping base of the Wingate. Additionally, enhanced deformation band propagation tendencies are predicted ahead of and slightly within the footwall of the thrust. Here, propagation would occur along deformation bands that nucleated at an earlier stage of fault growth. The predicted tendencies for deformation band propagation are consistent with the observed distributions of compressive mode II deformation band stepover structures, which occur solely between propagating deformation bands. Further, deformation band intensity for both nucleation and propagation tendencies is predicted and observed to increase toward the fault. These model predictions are consistent with independent observations of fault-related deformation band damage zone architecture from other paradigmatic outcrops in southern Utah and Nevada. By implication, specific locations within a damage zone that have the greatest reductions in fluid conductivity due to deformation band growth can be identified. We show that the tendency for fault growth and interaction within porous granular rock can be systematically predicted based on an understanding of in-situ stress state, fault and/or fold geometry, and rock strength and deformability at the time of deformation. This method is not limited to the prediction of deformation bands, but can also be used to predict the distribution of other types of fractures in other rock types, given that the appropriate critical strain energy density values are determined through laboratory testing for each fracture and rock type.
http://equinox.unr.edu/homepage/chriso/
T41F-1301 0800h
The Effects of Composition and Finite Strain on Rock Rheology: Constraints From Analysis of Natural Deformation
We present preliminary results of 3-D finite strain analysis across a strain gradient exposed in polymictic lapilli tuffs deformed by the Gem Lake shear zone in the eastern Sierra Nevada, California. The polymictic nature of the tuff allows us to directly examine how different materials behave during the same deformation. The presence of a finite strain gradient across the shear zone allows us to test the hypothesis that material rheology remains unchanged as finite strain increases. The polymictic lapilli tuffs in the field area were deformed by the Gem Lake shear zone at greenschist facies conditions. Field mapping and 3-D finite strain analysis demonstrates that these polymictic lapilli tuffs are more deformed in the center of the shear zone than at its periphery. Four clast types make up the lapilli in the tuff. In order from least to most deformed they are (1) very fine-grained quartz lapilli; (2) very fine-grained intermediate composition (feldspar with subsidiary quartz and oxides) lapilli lacking mica; (3) very fine-grained intermediate composition micaceous lapilli, and (4) highly altered mica-rich lapilli. Examining relative finite strain ratios between different clast types, we find that, as whole-rock finite strain increases, finite strain recorded by individual clast types does not increase proportionately. For example, the ratio of finite strain recorded by mica-rich lapilli relative to that of quartz lapilli increases as whole-rock finite strain increases. The clast type competence hierarchy outlined above appears to be valid across the strain gradient, but the range of competence displayed by the various clast types increases as whole-rock finite strain increases. Thin section analysis suggests that, as whole-rock finite strain increases, fabric development increases more quickly in lithologies recording high amounts of finite strain than in lithologies recording lesser amounts of finite strain. Consequently, we ascribe the relative weakening of intermediate and mica-rich lithologies across the strain gradient to fabric development in those lithologies. These preliminary results suggest that both composition and finite strain (fabric development) affect rheology.
T41F-1302 0800h
Grain-size Reduction of Feldspar in Middle Greenschist-facies Shear Zone
Deformation behavior of feldspar grains in granitic rocks is important in assessing crustal strength especially when they constitute a load-bearing framework. Grain-size reduction of feldspar can be achieved by fracturing and albite neocrystallization at greenschist facies condition, and by myrmekite formation, reaction and dissolution together with rotation recrystallization at epidote-amphibolite and higher-grade condition. Change in dominant deformation mechanism can occur with grain-size reduction. We report that grain-size reduction of feldspar is induced by a combination of fracturing, strain-related myrmekite formation and albite replacement in granitic mylonite from the middle greenschist facies Yecheon shear zone, South Korea. In granitic mylonite (K-feldspar 40%; plagioclase 34%; quartz 23%; muscovite 2%), feldspar grains deform by fracturing, mechanical twining and kinking while quartz grains deform by dislocation glide with recovery process mainly by subgrain formation and rotation recrystallization. High strained margins of K-feldspar porphyroclasts are characterized by intact or fractured myrmekite with fracture filling by muscovite. Along intragranular shear fractures of K-feldspar porphyroclasts, replacement (or neocrystallization) of fine-grained albite occurs whereas their intragranular extension fractures and pressure shadow are filled by fine-grained quartz and K-feldspar with minor amount of chlorite and muscovite. On the other hand, high strained margins and intragranular shear fractures of plagioclase porphyroclasts are replaced by muscovite and quartz while their intragranular extension fractures and pressure shadow are filled by fine-grained K-feldspar, muscovite and quartz. In the highly deformed portion of the shear zone, fine-grained plagioclase (46%), quartz (32%) and muscovite (20%) form monomineralic layers although plagioclase-rich layers contain minor amount of muscovite and quartz. If grain-size reduction of feldspar in brittle-ductile transition regime is accomplished by strain-related myrmekite formation and replacement in addition to fracturing, the crustal strength is presumably lower than assuming feldspar fracturing only.
T41F-1303 0800h
A New Methodology to Probe the Failure and Compactive Yield Behavior of Porous Rocks Under Undrained Conditions
The poromechanical behavior of rocks under undrained condition is of fundamental importance in many geologic and geotechnical problems. In a conventional undrained experiment the saturated sample is deformed under constant confining pressure while fluid movement in or out of the pore volume is inhibited. Maintaining a constant pore volume requires a pore pressure system that contributes negligible volume and compressibility to the overall mechanical response, which is difficult to achieve in a relatively compact rock. To circumvent this difficulty associated with the conventional experiment, we have developed a modified undrained methodology in which the rock sample is deformed while the pore pressure is maintained constant and the confining pressure continually adjusted to maintain the pore volume constant. Triaxial compression experiments were conducted on water saturated samples of Adamswiller and Diemelstadt sandstone at a constant pore pressure of 10 MPa, confining pressure ranging from of 50-275 MPa, and a nominal axial strain rate of 1.9 x 10-5 s$^{-1}$. In a typical test several stress contours for one rock sample at constant porosities ranging from 12-20% can be mapped out. If the elastic volumetric strain is negligible, then such a stress contour corresponds to a plastic yield envelope associated with zero plastic volumetric strain. Mechanical data for drained tests provide constraints on the compactive yield caps at different values of plastic volumetric strain, which are in accord with stress contours derived from modified undrained tests at differential stresses below a critical threshold. Acoustic emission (AE) measurements show that this stress threshold is marked by an upsurge in AE activity that signals appreciable damage and plastic yield.
T41F-1304 0800h
Deformation Structures in Poorly Consolidated Clastic Sediments of Varying Grain Size
The structure and texture of deformation bands were analyzed in poorly consolidated sandstones, conglomerates, and mudstones of Miocene to Pleistocene age in northern California to assess depositional and diagenetic controls on their mode of deformation. Poorly consolidated arkosic sandstone in the footwall of an active reverse fault contains three sets of deformation bands: two inclined sets of reverse shear bands and a third set of subhorizontal bands that have no apparent shear offset and that were characterized as dilation bands. Within shear bands, elongate grains are rotated parallel to the orientation of the macroscopic band. Grain breakage appears largely restricted to softer lithic grains. Bands are preferentially cemented with clay and iron oxide minerals due to the infiltration of fine grains. In interbedded sandy conglomerate layers, shear bands are more diffuse compared to those in sandstone. Elongate pebbles are preferentially rotated parallel to the shear band direction and slip surfaces are localized within the clayey matrix. In mudstone, shear bands appear as darker bands that are about 1-3 mm wide forming anastomosing arrays in lenticular pattern. These bands are bound on one side by striated slip surfaces. These observations indicate that deformation bands are a fundamental mode of deformation in poorly lithified sediments involving grain translation, reduction or increase in porosity, and grain rotation, with only minor grain fracturing. Together with the diagenetic attributes, they should act as low permeability elements at present time.
T41F-1305 0800h
Compaction Localization in Notched Samples of Bentheim and Berea Sandstones: Mechanical Experiments, Microstructural Observations and Stress Analysis
Compaction band formation has been documented in field and laboratory studies as a localized failure mode in porous sandstones. To simulate the development of compaction bands induced by local stress concentration, we introduced a V-shaped circumferential notch in cylindrical samples of Bentheim and Berea sandstones and conducted triaxial compression tests at a range of confining pressures from 150 to 350 MPa. Our mechanical and microstructural data indicate that the stress concentration caused a compaction band to initiate at the notch tip as a cluster of fractured. The critical differential stress for its initiation decreased with increasing confining pressure, similar to the yield stress for an unnotched sample that maps out a cap with negative slope in the stress space. At the initiation stage, the asymptotic stress field in the vicinity of the notch can be evaluated using linear elastic fracture mechanics. Near the notch the stress path is approximately linear with a slope dependent only on the Poisson's ratio. The localized mean stress is significantly enhanced relative to the remote loading. Plastic yield initiates when this stress path intersects the yield cap determined for unnotched samples. From the asymptotic stress analysis, we infer that the initiation of localized yield, as indicated by a surge in acoustic emission in a notched sample, typically involves a damage zone extending from the tip to one or two grains, in agreement with our microstructural observations. Beyond initiation, compaction localization was observed to propagate differently for the two sandstones. While the bands propagated as "anti-cracks" in a direction subperpendicular to the maximum principal stress in Bentheim sandstone, they developed as conjugate bands at an angle of \sim$55\deg$ in Berea sandstone. This difference is possibly related to the different modes of localization observed in unnotched samples: Bentheim sandstone develops discrete compaction bands with episodic stress drops, whereas Berea sandstone develops diffuse compaction bands that are significantly thicker accompanied by strain hardening.
T41F-1306 0800h
Interpreting the Frictional Behavior of the Smectite Clay Montmorillonite
Montmorillonite has been the most widely tested phyllosilicate mineral in soil and rock mechanics friction investigations, but many of the published data are contradictory, with reported values of the coefficient of friction (= shear stress/effective normal stress) ranging from 0.06 to 0.78. We report some new laboratory and petrographic data which illustrate that this wide variation is in part a function of the large difference in coefficient of friction between thoroughly dried (0.7 or greater) and water-saturated (0.3 or lower) montmorillonite. Dry montmorillonite gouge is subject to standard frictional processes such as abrasion, wear, and fracture during shear. In contrast, shear of water-saturated montmorillonite gouge is concentrated in thin films of water that are adsorbed onto the (001) surfaces of the platy grains. Our recent studies suggest that the water-saturated shear strength of sheet-structure minerals increases with the strength of the bonding of the polar water molecules to the (001) surfaces, and the relative weakness of water-saturated montmorillonite may be largely owing to its small layer charge. Values of the coefficient of friction for montmorillonite that are considered to represent water-saturated, equilibrated conditions increase from 0.06 at effective normal stresses below 1 MPa to 0.30 at 300 MPa. This correlation is attributed to decreasing thickness of the surface water films with increasing effective normal stress. Similar stress dependence of frictional strength can be demonstrated for the serpentine minerals, muscovite, biotite, phlogopite, chlorite, kaolinite, and talc, and it is considered to be characteristic of sheet silicates. This behavior contrasts with that of most other silicate minerals, for which the coefficient of friction exhibits little pressure sensitivity below 200 MPa effective normal stress and then decreases at higher stresses (Byerlee's law). Most of the published strength data for montmorillonite fall outside the range of values for water-saturated, equilibrated samples. Of these, the samples that are overly strong for a given set of experimental conditions may have been only partially saturated. Those samples that are weaker than the water-saturated, equilibrated samples at a given effective normal stress appear to result from inadequate drainage and consequent build-up of internal pore pressure. The velocity dependence of montmorillonite strength has not been extensively investigated, but water-saturated montmorillonite gouge is velocity-strengthening over the range of conditions tested to date whereas dry and partially saturated montmorillonite gouge may be velocity weakening at some velocities. These results highlight the hazards of interpreting fault-zone behavior based on experiments that do not approximate natural conditions.
T41F-1307 0800h
Borehole Breakouts and Compaction Bands in a Medium Porosity, Quartz-rich Sandstone
We conducted laboratory-drilling experiments in Coconino sandstone in which vertical boreholes were drilled into prismatic samples subjected to true-triaxial far-field stresses (\sigma$_{H}$ $>$ \sigma$_{v}$ $>$ \sigma$_{h}$) simulating field conditions. This early Permian aeolian sandstone comes from Arizona, and was selected because of its medium porosity (16-20%) and its rounded quartz grains (96%, with minor amounts of feldspar). Its fine-grains (0.1 mm) are well sorted and bonded through weak grain-contact suturing and some quartz overgrowth. Boreholes drilled under sufficiently high far-field stresses developed narrow and tabular fracture-like breakouts (constant width of 10 grain diameters) extending along the \sigma$_{h}$-springline and perpendicular to \sigma$_{H}$ direction. This behavior is consistent with other sandstones previously tested, which possessed similar composition and texture but had porosities in excess of 20%. However, breakout length in the Coconino was dependent on porosity. The higher porosity variety (20%) produced much longer breakouts than the 16% samples. SEM analysis of the damage zone ahead of the breakout tip revealed a narrow compaction band, having the same width as the breakout, localized along the \sigma$_{h}$-springline where compressive stress concentration is the highest. Within the band, grains only a few millimeters away from the breakout tip are already mostly intact, but debonded, with the sutured contact replaced by visible microcracks. The existence of a compaction zone is reflected by the reduction in the 2D porosity measured within the band. The initial failure mode leading to the localized compaction appears to be non-dilatant, since it involves no extensile microcracking. The zone of cracked and crushed grains immediately ahead of the breakout tip is a secondary failure zone within the compaction band as a result of excessive stress concentration there. Grain fragments and debonded grains are removed from the compaction band by the circulating drilling fluid, resulting in a long tabular fracture-like breakout. Our conclusion, based on testing a large number of sandstones, is that the failure mode leading to fracture-like breakouts is likely to occur in sandstones regardless of their porosity (at least within the tested 15-28%) and grain size (0.1-0.5 mm), as long as grains are rounded and overwhelmingly quartz, and bonding is through weak sutures. Such grain contacts are readily broken under sufficient normal stress owing to low fracture toughness along sutured surfaces. Debonded grains are then compacted through their relative movement into adjacent pores, creating a compaction band and facilitating fracture-like breakout formation.