Dan M. Davis

Mechanical role of backstops in the growth of forearcs

We use numerical and laboratory modeling to analyze the mechanical role of backstops in the overriding plate at subduction zones. A backstop is defined as a region within a forearc that has significantly greater shear strength than the sediment lying farther trenchward; it can be thought of as the bulldozer behind an accretionary wedge. We calculate the stress and displacement fields within forearcs for various backstop models using the finite element method, and we simulate deformation over a backstop using a small-scale laboratory model. In this way we model the effects of the mechanical properties and geometry of a backstop on forearc structures. We find that the growth of an outer arc high, the development of an inner deformation belt with landward vergence, and the seemingly paradoxical presence of an undeformed forearc basin within an otherwise highly deformed forearc can be explained by assuming the existence of a geometrically simple backstop with geologically reasonable contrasts in mechanical properties compared to the sediments just trenchward of it. The structures produced in our numerical models are quite insensitive to the rheology, boundary conditions, and exact mesh geometry employed. The general types of structures observed in the laboratory models depend only weakly upon the strength and the geometry of the backstop. These results suggest that a detailed picture of an underlying backstop cannot be determined from surface information alone. Backstops in which the contact with the accretionary wedge dips arcward rather than trenchward, however, should produce only slightly different forearc structures, with less development of both the outer arc high and the landward-vergent inner deformation belt. Although natural forearcs are far more complex than our simple models, they exhibit many of the same features, indicating that relatively simple backstop mechanics may be a very important factor in the overall growth of forearcs.

Sandbox model simulation of forearc evolution and noncritical wedges

Convergent plate boundaries exhibit a wide range of behaviors, ranging from sediment accretion and underplating to large-scale erosion of the forearc. We have used sandbox modeling to simulate the growth of thin-skinned contractional wedges over a strong backstop at an accretionary margin. The backstop in our model causes the growth of two distinct wedges with quite different growth histories, analogous to the accretionary wedge and the inner deformation belt (retrowedge) found in at least some forearcs. After starting with an initially steep slope, the surface flattens as the accretionary wedge grows by frontal accretion, until the wedge attains and maintains a minimum possible critical taper. The retrowedge, driven by uplift of the outer arc high, increases its surface slope as it grows until a kinematic steady state is attained. It exhibits a stable taper with slope larger than the minimum critical taper but never larger than the theoretical maximum critical taper when a basal decollement is present above the backstop. In models in which the retrowedge has not developed a decollement, it has a taper with a slope at the angle of repose. When a basal decollement is present above the backstop, the main body of the retrowedge acts as a bulldozer or a backstop, pushing the sediments in front of it, which then form an additional wedge with a small (minimum) critical taper. A broad outer arc high stands above the toe of the backstop, where maximum principal stresses rotate to meet the stress conditions in the two contractional belts on either side of it. As these belts grow in opposite directions, the accretionary wedge accretes large quantities of sediment at its front, but the retrowedge does not. Because the retrowedge can exhibit a non-critical, stable taper in the absence of rapid erosion or accretion, it is misleading to assume that active thin-skinned wedges are necessarily at maximum or minimum critical taper.

Inferences on sediment strength and fault friction from structures at the Aleutian Trench

The structural relations of conjugate thrust faults and the master decollement imaged in a research-level, processed, multichannel seismic reflection record across the eastern Aleutian Trench allow quantitative estimates of the sediment strength and friction across thrust boundaries of the subduction zone. These estimates are derived from geometric relations between faults and from the overall geometry of accreted prisms. They suggest the presence of high pore-fluid pressures that result in low levels of friction along a master decollement and on the main bounding thrust faults in the accreted sediment. Therefore, coupling in the immediate trench area apparently does not contribute significantly to the generation of horizontal compression upslope.

Kinematics of active deformation in the Sulaiman Lobe and Range, Pakistan

The western margin of the Indian plate is highly oblique to the direction of convergence between India and Asia and represents an excellent example of large-scale oblique continent-continent collision. Determining the strain field in western Pakistan and how it relates to the plate motion and plate margin geometry affords an exceptional opportunity for understanding oblique margin processes in general. Through the inversion of regional and teleseismic body waves, we have determined the source parameters of 10 moderate-sized earthquakes that occurred between 1964 and 1985 in and around the Sulaiman Range, Pakistan. The earthquakes are dominantly thrust events with slip vectors that are approximately perpendicular to the lobate Sulaiman mountain front. Slip vector orientations rotate 60°–70° from a N-S to a WNW-ESE direction of compression, consistent with the geometries of the complex, festoon-shaped mountain belts of this margin. We have estimated the spatial variation of the horizontal strain rate and velocity fields within Sulaiman using vertically averaged models that accommodate plate motion constraints within a deforming layer. The most important factors determining the style of strain rotation in the Sulaiman Lobe and Range are the presence of pure strike-slip motion along the Chaman Fault, and the relatively rigid and undeformed Katawaz Basin that is therefore allowed to translate obliquely relative to India. This same conclusion is obtained using either a three-dimensional, frictional, analogue model with significant basal tractions or a thin sheet viscous numerical model without basal tractions. Thrusting in a predominantly NW-SE direction in the Sulaiman Range accommodates 5–14 mm/yr of N-S motion between India-Eurasia and 3–6 mm/yr of E-W shortening. Seismic moment release this century within the India-Eurasia plate boundary zone, west of the western Himalayan Syntaxis, constitutes roughly 40% of the expected total seismic moment release for this time period. Particularly significant moment rate deficits exist within the Sulaiman Range and along the Chaman Fault.

Changing mechanical response during continental collision: Active examples from the foreland thrust belts of Pakistan

Abstract We have used data from teleseismic, seismic reflection and field geologic studies, along with both geomechanical and gravity modeling to contrast the tectonics of four active orogenic wedges in Pakistan: the Kashmir Himalaya, the Salt Range-Potwar Plateau foldbelt, the Sulaiman Range and the Makran accretionary wedge. In Makran, oceanic crust is still being subducted, and a thick pile of sediments is being accreted and underplated. Undercompaction and excess pore pressures can explain the narrow cross-sectional taper and frontal aseismicity of this wedge. Beneath the Sulaiman wedge, continental crust is just starting to be underthrust. Indirect evidence suggests that fine-grained carbonate rocks found in abundance deep in the stratigraphic section may be deforming ductilely at the base of the Sulaiman wedge and provide a zone of ductile detachment. The collision has proceeded to a much more mature stage in the Salt Range-Potwar Plateau foldbelt and the Kashmir Himalaya. Isostatic response to underthrusting of continental crust has kept the sedimentary pile quite thin in both of these wedges, so in that respect the two foldbelts are similar. However, thick Eocambrian salt beneath the Salt Range and Potwar Plateau permits that foldbelt to be much wider in map view, with a thinner cross-sectional taper and a mixture of thrust vergence directions. A major normal fault in basement causes the Salt Range to rise in front of the mildly deformed molasse basin of the southern Potwar Plateau. Much of the diversity among these mountain belts can be understood in terms of differences in the maturity of the collision process in each area, the resulting thickness of the sedimentary pile encountered at the deformation front, and the presence or absence of large contrasts in strength between the various layers of the stratigraphic section and basement relief.

Intraplate stresses and plate-driving forces in the Philippine Sea Plate

We use a spherical shell elastic finite element analysis to characterize intraplate stresses and plate-driving forces in the Philippine Sea Plate. Our finite element mesh is comprised of 489 nodes and 914 elements, and provide a spatial resolution of about 1°. The Philippine Sea Plate is unusual in that it does not have large numbers of seismic or tectonic stress indicators, and there are no midplate borehole stress data which can be used as constraints in this finite element study. However, it does possess a unique boundary configuration which potentially can lend more insight to plate forces and stresses than can similar studies on plates with more seismically active interiors. Because the plates boundary consists primarily of subduction zones with roughly equal proportions of boundary on the overlying and downgoing sides of trench, the resultant plate stress field is very sensitive to subduction-related forces. As a result, borehole stress measurements in certain parts of the plate would be useful in further constraining these forces. We find that trench pull forces can explain a large component of the velocity of the Philippine Sea Plate. However, trench pull has somewhat less overall effect on the stresses across the plate than do the lateral density and topographic variations within the plate (i.e., gravitational or potential energy force), and trench suction forces are probably about an order of magnitude smaller than the net trench pull forces. Collisional forces may dominate the stress fields near Taiwan and the Izu Peninsula, but they are of only secondary importance in determining the motion and deformation of the Philippine Sea Plate. The forces that generate stress orientations most closely matching the observed stress indicators produce a lithospheric stress field across the plate that averages about 15 MPa over a 70 km thick lithosphere, scaling inversely with lithosphere thickness. Values up to four times as great are produced locally.

Extension during active collision in thin-skinned wedges: Insights from laboratory experiments

The occurrence of strike-normal extension in mountain belts undergoing active contraction is not predicted by analytical solutions for the mechanics of purely frictional orogens. To test the idea that ductile rocks at depth might allow extensional deformation to occur in otherwise frictional and contractional orogens, we have conducted a series of analog experiments with purely frictional and layered frictional-ductile rheologies. By precisely measuring the horizontal deformation field, we have quantitatively confirmed the qualitative observation that analog frictional wedges do not extend. This is consistent with published numerical models, and is in contrast to purely viscous models that commonly display coeval extension with contraction. Our modeling also demonstrates that layered frictional-ductile models can, during active convergence, result in discrete extensional normal faulting even when only a relatively thin ductile layer is present. The resulting extension, though relatively small in magnitude compared to the active contraction, can greatly modify the final profile of a convergent orogen and can lead to the formation of a broad plateau between the pro-wedge and the retro-wedge.