Click for next page ( 74


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 73
7 ACTIVE MARGINS: GROUP 1 DYNAMICS OF SHORT-TERM DEFORMATION AT ACTIVE MARGINS THE SINGLE MOST IMPORTANT SCIENTIFIC OBJECTIVE At the workshop, the title of this working group was "Mechanics of plate motion." During its discussions, the group decided to emphasize deformation along the convergent margin megathrust and in the accretionary wedge, and the vertical and horizontal distribution of forces and displacements at active continental margins. Consequently, the working group established that its single most important scientific objective is to understand the dynamics of short-term deformation at convergent and transcurrent margins e Within this broad objective, the group identified the three major areas of investigation discussed below. THREE MAJOR AREAS FOR INVESTIGATION What Controls Aseismic Deformation and the-Recurrence Characteristics and Locations of Catastrophic Earthquakes? Relative plate motion at active continental margins is accommodated by both seismic and aseismic processes. Some pi ate boundary segments have no historical record of great earthquakes, ant] in some cases the record covers a sufficiently ~ ong period of time that it appears likely that the subduction zone never s] ips seismically. Other segments rupture at regular or irregu~ ar intervals with events of varying size and frequency. For still others, it is not at all clear what fraction of relative motion occurs seismic ~ y or aseismically. The factors that control the mode of subduction slip and the space-and-time characteristics of earthquake recurrence at subduction zones are as yet poorly known. Progress in understanding these processes is important both for assessing long-term seismic risk and in making progress towards short-term prediction and hazard mitigation. 73

OCR for page 73
The principal factors that are believed to control the recurrence rate, size, and location of catastrophic earthquakes are as foil OWSe 1. The relative velocities of the adjacent plates. This is most important in the recurrence rate since it controls the time needed to build up the stress to a level that will overcome the static frictional resistance between the plates. 2. Physical properties of earth materials. The temperature, rock type, porosity, and Theological behavior of material in the active margin determines how the earth responds to plate displacements and whether energy is released catastrophically or aseismically. For this reason, knowledge of the physical properties is of the utmost importance. 3. Area of contact between the adjacent plates. This parameter controls the maximum possible size of the earthquake, or the amount of energy released. It depends on the local plate geometry. 4. The temperature and pore fluid pressure in the active margin. These are among the critical parameters controlling whether the materials respond in a ductile or brittle fashion. They cannot be measured in.situ and must be inferred from other physical properties, such as seismic velocity, heat flow measurements, and electrical conductivity. 5. Deep geologic structure. The distributions of plate boundary asperities, mantle inhomogeneities, faults, -and-other defor~national structures must be better understood if we are to explain the response of the active margin to the applied stresses. They are often complicated and affect the local stress and strain fields in ways that make it difficult to interpret measurements of stress and strain. However, these features must be understood if the earthquake processes are to be understood. It is well known that convergent margins exhibit different types and levels of seismicity. Strategically, it may be important to design integrated geophysical and geological investigations both at margins that are characterized by great earthquakes and at margins that lack such events. Thus, it will be necessary to include studies both at margins that produce long (> 200 km) ruptures and at those that have no historical record of great earthquakes (rupture 1 ength < 100 km) . Strongly focused integrated studies of several active margins, including structural, seismic reflection, geodetic, bathymetric, and seismic network data -are needed to identify those f eatures that are d iagnostic of se i smic and ase i smic accommodation of red ative pi ate motion. 74

OCR for page 73
What Controls Accretion, Nonaccretion, and Eros ion of the Upper P1 ate at Convergent Margins? There is a remarkable range of tectonic styles at the front of convergent margins, ranging frown accretionary to erosional Margins of the Lesser Antilles and Cascadia, for instance, grow by accretion, whereas the Middl e America margin varies locally from accretionary to erosional, and the northern Japanese margin is erosional . The causes of this variability are not obvious, but they bear on the fundamental problem of how mass is added to and remover] from continents. It is important to understand the structure of accretionary wedges at two different seal es. At a regional scale, the shape of the wedge is important because of its dependence on the rel ative strengths of the material s within and below the wedge e Determination of the smaller-scale internal structures in the wedge is important because of the information they contain on the stress, strain rate, temperature, ant} fluid conditions under which the deformation has taken place. Orogenic-wedge model s imply that accretion and erosion at convergent margins are influenced by temporal and spatial variable ity in the frictional coupling between the downgoing and overriding plates. It has been suggested that an important mechanism for accretion or erosion of sediments is the upward or downward Lisps acement of the master clecoll ement. If this is the case, it is unclear what alters the material strengths in such a way as- to make it mechanical ly preferred for the basal thrust to shift position. Frictional coupling is a poorly constrained] function of the rheology of the accretionary deposits, and it is particularly sensitive to variations in pore fig uid pressure. Unfortunately, the Theological behavior of accretionary materials is poorly known under the combination of temperatures, pressures, porosities, and pore-f~uid pressures prevalent in most accretionary wedges. The margins most likely to permit critical assessment of the relationship between the mechanics of the accreted materials and the structural configuration of the decollement are those in which the effects of variability in the sediment input can be isolated. Thus, margins with considerable 1 atera~ variation in both sediment intake and decollement structure, such as Peru-Chile and the Lesser Antilles, would provide promising targets. Margins undergoing active erosion (e.g., the eroding portions of Peru-Chile and northern Japan) may provide drilling targets at reachable depths, permitting direct measurements of both mechanical properties and stress. The geometry of the descending oceanic plate has been suggested as another primary control on the tectonics of the frontal part of the overriding plate, although no clear causal 75

OCR for page 73
frontal part of the overriding plate, although no clear causal relationship has been demonstrated. The geometry of the Benioff zone appears to be related to the age of the downgoing lithosphere, the relative velocities of the two plates' and the tectonic history of the upper plate. Topography on the converging ocean crust, when inserted into the subduction zone, disturbs the tectonics of that region. Even when the basement red ief is insufficient to erode the upper plate, it may wed ~ perturb the stress field by an amount suf f icient to alter the conf iguration of the decollement ~ The subduction of a 3500-m-high seamount beneath northern Honshu produces ups ift of the upper plate above the leading se amount flank. Another seamount, further into the subduction zone, has produced the coil apse of the margin down its trailing f lank ~ In less than a million years ~ the impact and subduction ~ f the seamount caused a change f ram accretion to eras ion and back to accretion again. This example if lustrates a process which must af feet convergent margins at many scales . Swath mapping (two-dimensional bathymetric surveys) and high-reso~ution se ismic surreys now provide powerful tools for the study of collisions at subduction zones. What Determines the Partitioning of Deformation In P1 ate-Boundary Zones? The concept that the lithosphere acts as a high-strength stress guide is fundamental to plate tectonics. The concept works well in plate interiors, where the crust may move with the (nearly unifo ~) velocity of the underlying high-strength mantle lithosphere. However, plate boundaries are not simple, localized zones of shear. This is particulary true at continental margins, where deformation can be distributed over zones ranging up to ~1, 000 km in width. The mechanics of this departure from simple plate behavior represents a fundamental aspect of the geodynamics of plate boundaries. A starting point for understanding strain partitioning at plate boundaries is to recognise the general variation of Reformational properties within the lithosphere. There are alternating strong and weak layers that have prompted the analogy of a peanut butter and jelly sandwich. In the shallower half of the crust, rocks support substantial shear stress. At these depths, rocks ultimately deform by brittle failure and can produce earthquakes. Relating earthquakes to the general strain release at plate boundaries has considerable societal as well as scientific value. Unlike the rocks nearer the surface, the lower crust is very weak, deforming continuously by any of several creeping processes. The uppermost mantle is at a lower fraction of its men ting temperature. It is again structurally strong, and 76

OCR for page 73
in many areas earthquakes occur within it. The details of the distributions of deformation modes and shear stresses depend on a number of factors. Low temperature and low confining pressures favor brittle failure, while high temperatures promote creep. High deviatoric stress' pore fluids and vocatives also have weakening effects. The upper mantle part of the lithosphere is nor-~ally the structurally strongest layer in the lithospheric "sandwich." For this reason it usually constitutes the primary stress guide of the lithospheric plates. The mushy nature of plate boundaries in continental margins may result from the weakening of this normally strong member of the lithospheric sandwich due to high temperature. The western United States provides an example of distributed plate-boundary deformation. In this case, strain is broadly distributed across a region extending from the continental margin to the Rocky Mountains. The maximum displacement in central and southern California over recent geologic time is on the San Andreas fault, but substantial deformation, on both geological and seismological timescales, occurs throughout the region. The pattern of deformation involves more than the simple shear expected as the result of transform motion between the adjacent rigid plate interiors. For example, strike-slip motion on the San Andreas fault contrasts sharply with the nearly orthogonal horizontal contraction indicated by many structures in the Transverse Range, including the Big Bend of the San Andreas. Many of the largest historic earthquakes in California, in fact, have been associated with these contractional features. It has been suggested that the Big Bend and associated convergence are the result of a distributed zone of weakness in the area along the San Andreas fault. This weakness might be related to the large total slip of the fault. Forces due to the str~ke-slip relative motion of the plate interiors would thus be transmitted through the lithosphere and converted to convergent motion simply as a kinematic consequence of the weak zone being curved. An alternative hypothesis is that all the major faults in the upper crust, as well as the underlying lithospheric mantle this region, are relatively weak and incapable of transmitting large shear tractions over large distances. In this case, the substantial forces needed to drive convergence and build the actively growing Transverse Ranges must be provided locally, perhaps by stresses associated with convective flow in the underlying mantle. Partitioning of deformation is also observed at some obliquely convergent margins, such as the the western Aleutian and northern Sumatra margins. In such settings, contractional 77

OCR for page 73
structures apparently accommodate the component of convergence orthogonal to the plate boundary, and the motions along transcurrent faults accommodate the boundary-parallel part of the relative plate motion. Space-based geodetic techniques (e. g., Gl obal Positioning System) will be crucial in determining (on time scat es of years to decades) the pattern of deformation across continental margins. Seismic tomography is a rapidly developing tool used to determine the structure of the crust and upper mantle. Determination of variations in the state of stress using borehole measurements or focal mechanisms is important for discriminating between tectonic hypotheses, as is the accurate resolution of the structure:of the lithosphere. The comparison of geologic strain with geodetic strain is important in determining seismogenic potential of structures. Relating seismic activity at depth to geologic structure would shed light on the nature of the seismic rupture process. Finally, numerical modeling is essential for the quantitative testing of hypotheses against data. NEEDED STUDIES Necessary studies include those required to determine the three-dimensional geometry of the downgoing plate and to isolate the factors that control whether plate-boundary deformation is seismic or aseismic. The studies will require the acquisition of structural, seismic reflection, geodetic, bathymetric, and seismic network data. Geometry of the Downgoing Plate It has been suggested that the geometry of the descending oceanic plate strongly influences the deformation of the frontal part of the overriding plate. This geometry is difficult to resolve accurately in the upper-50 to 75 km, and requires integration of seismic tomography, gravity studies, anc} hypocentral determinations made us ing local seismic networks . A maj or impediment to seismic studies at convergent margins has been the fact that seismic networks have been larger y restricted to the continent. Even when stations can be pa aced on islands, the islands are commonly wider y dispersed and are commonly 1 ocated atop structures with locally anomalous seismic velocities which complicate the interpretation of data. For these reasons, a combination of both OBS (ocean-bottom seismometer) and land-based stations will be needed in future seismic networks at convergent margins. Seismic tomography provides a three- dimensional view of the subducting ocean crust down to mantle depths. This reconnaissance information can be augmented with high-energy seismic reflection images of the crust and Moho. Such studies could be carried out at any sufficiently 78

OCR for page 73
well-studied margin. Examples providing contrast would include the Andean, Marianas, Lesser Antilles, and Alaskan margins . Structures, Geodetics, and Seismicity of Subduction Margins It is well known that convergent margins exhibit dif ferent types and levels of seismicity ~ Strategically, it may be important to design integrated geophysical and geological investigations both at margins that are characterized by great earthquakes and at margins characterized by smaller magnitude events. Some possible candidates are: 1. Great earthquakes ( > 200 km ruptures) - Chile (virtually all its coast from Arica to Chonos Archipelago) Central Aleutians - Kamchatka - Alaska 2. No historical great earthquakes (> 100 km ruptures) - Izu-Bonin-Marianas (no events M > 7~4) - Ryukyu - Tonga-Kermadec (may have experienced a few earthquakes, ~200 km ruptures)- Strongly focused integrated studies of several active margins are needed to identify features diagnostic of seismic and aseismic accommodation of relative plate motion. Observational programs aimed at delineating structural features and determining contemporary patterns of crustal deformation and seismicity will provide correlations and constraints. The age, geometry, structure, and deformation styles of accretionary wedges are expected to play an important role in determining the seismic-or aseismic character of convergence. High-resolution seismic imaging at specific convergent margins will contribute to the mechanical understanding of wedges, and it can complement the results of laboratory studies designed to determine the constitutive properties of wedge materials and to define their failure modes under a range of pressure and temperature conditions. ~~ - New measurements of present-day crustal deformation are needed to determine the spatial and temporal patterns of movement and relate them to seismic and aseismic processes and to the long-term deformation of active margins e Use of Global Positioning System COPS ) geodetic surveying onshore and new ocean-bottom techniques in offshore environments can provide the data needed to define the present geodetic movement pattern from the magmatic arc to the oceanic outer rise and to monitor its temporal variations. Integration of these data with structural information supplied by multichannel seismic, high-resolution 79

OCR for page 73
bathymetric surveys, and small earthquake locations and focal mechanisms obtained from onshore/offshore seismic networks will provide a context for relating present geodetic movements to the recent geologic record. Although much can be learned from dry-land geodetic surveys, the development of seabottom measurement techniques is vital for obtaining a complete picture of active margin deformation. Experience with land-based measurements in Japan, California, and New Zealand indicates considerable breadth (100 to 200 km or more) in the zone of contemporary deformation. A similar range Bill probably be found in other active margin settings. Offshore geodetic measurements are needed to bound the regi on of present-day deformation, to relate it to the seismic or aseismic character of each active margin, and to determine the partitioning between elastic straining, which will ultimately be relieved by earthquakes, and permanent deformation, which will be preserved in the geologic record. Seismic observations have suggested a linkage between subduction zone earthquakes, outer rise seismicity, and lateral migration of great underthrust earthquakes. Geodetic monitoring, both onshore and offshore, can critically constrain this strain migration process. Developments in several research areas are particularly important in understanding short term deformation at active margins. Those singled out here include numerical modeling, enhanced geophysical instrumentation, and rheological studies of geologic materials. Numerical Modeling Numerical modeling is essential for the quantitative testing of hypotheses against data. More sophisticated models are needed, incorporating realistic three-dimensional geometries and coupling of rheological variables. Such models require sophisticated software, as well as access to powerful computers. Modeling of the distribution Of stress and strain at active continental margins can be used to relate observations of deformation to rheological laws obtained in the laboratory. Modeling can also suggest improvements in observational strategies and indicate those laboratory studies that are likely to prove most critical. Geodetic observatories using GPS and surface techniques will provide maps of surface deformation at the centimeter level within the next decade . Breakout studies in existing wells can provide stress directions in many areas. Two-dimensional and three-dimensional finite-element models will be required to interpret these observations. Models should include elastic, fault related, plastic, and nonlinear viscous rheologies. 80

OCR for page 73
Several goals are clear, among them: 1. to distinguish those margins where deformation is relatively monotonic from those where it is cyclic due to the accumulation of elastic strain energy and fault rupture; and 2. to provide a better understanding of the cyclic accumulation and release of strain associated with great earthquakes. Enhanced Geophysical Instrumentation - Certain specific new measurement capabilities will be needed. They include: the ability to make extensive geodetic measurements on very short notice; improved abilities to conduct underwater geodetic and seismic network studies; and improved techniques for stress measurements and deep seismic reflection. Some of these future needs are outlined below. Submarine and Rapidly Deployed Geodetics Recent developments in space-based geodetic techniques Berg., GPS) have opened the possibility of accurate geodetic observations on a scale far larger than has been heretofore feasible. These techniques will be crucial in determining, on timescales of years to decades, the general pattern of deformation across continental margins. Rapid deployment of instruments to the near-field region after earthquakes would allow the observations of temporal variations in postseismic strain that are crucial for constraining the Theological properties of fault zones, the lower crust' and the upper mantle. Extension of these techniques to allow accurate positioning on the seafloor appears technically feasible and is extremely important for the obvious reason that most deformation at active margins spans the border between land and sea. Ocean-Bottom Seismometers Seismic tomography is a rapidly developing too] that is used to determine the structure of the crust and upper mantle. For example, in the Transverse Ranges of southern California, regional tomographic studies have revealed a curtain of high velocity material extending to a depth of 250 km. This feature has been interpreted as the convective downwelling of the cold, dense base of the thermal lithosphere. The tomographic image loses resolution in the offshore region because of the lack of seismic stations there. This is a good example of an important tectonic problem that could be addressed much more effectively with the deployment of OBS to fill in gaps in coverage. In this and other tectonically active regions along continental margins, 81

OCR for page 73
the Kept oyment of dense seismic arrays should be a high scientific priority. Measurements of Stress anti Deep Structures Determination of the in situ state of stress using seismic focal mechanisms and at selected locations using boreho~ e measurements is important for discriminating among tectonic hypotheses. Accurate description (in three dimensions) of the geologic structures is also crucial for understanding the mechanics of tectonic processes, both to determine crustal kinematics on geologic timescales and to identify the structural units that are important mechanically. Modern structural geology, utilizing data from seismic reflection surveys and boreholes, provides an additional approach to this problem. Rheological Studies Because orogenic-wedge models suggest that accretion and erosion at convergent margins are inf luenced by variate! ity in coupling along the zone of major deformation, studies to define the rheological behavior of accretionary materials and to assess the hydrology of pore fluids are needed. However, the mechanical properties of sedimentary rocks are much less well known under accretionary wedge conditions than within the high pressure-low porosity conditions generally studied in rock mechanics or the low pressure-high porosity conditions studied in soil mechanics. Laboratory studies are needed of the conditions representative of those associated with the processes of subduction erosion and underplaying (0 to 200 MPa and 10 to 500C), in order to determine the material behavior associated with these processes. Laboratory samples, however, cannot include larger-scale (>10 cm) features , such as veins and fractures, which are ubiquitous in accretionary complexes and which may control the strength of these materials in situ. Although the frictional coupling between lithospheric plates is difficult to quantify from direct observations, the integration of downhole geotechnical (stress-strain, pore pressure) determinations, laboratory results, and numerical modeling could provide realistic constraints on material behavior. Porosity, intergranular and fracture permeabilities, and pore pressures within the prism are parameters that are needed to understand fluid flow. Although we need to determine in situ values for these parameters, such measurements are difficult and expensive. Therefore, hydrologic modeling, constrained by a few high-quality field measurements, may prove to be the most efficient means to improve our understanding of fluid flow and the water budget within the prism. 82

OCR for page 73
A variety of microscale Reformational processes operates in the crust and mantle. They are responsible for the wide range of structures seen in both active and passive continental margins. Many of these processes can be studied by experiments in the laboratory, where it is possible to obtain quantitative information on rheology under controlled Reformational conditions. This in turn makes possible a better understanding of the mechanics of crustal-scale deformation because it permits estimates of stresses if strain rates are known, and vice versa. Some of the major rock mechanics issues related to the tectonics of active plate margins are discussed in Chapter 10. ~3

OCR for page 73