5.2 Accretionary Wedges
Sediment at the Destructive Margin
As oceanic lithosphere subducts, the sedimentary cover on the downgoing plate encounters the leading edge of the overriding plate. The fate of this sediment defines two end-member margin types: accretionary margins, where sediment is scraped off and stacked into a growing wedge, and erosive margins, where the overriding plate is tectonically eroded and material is carried into the subduction channel.
Approximately 57% of global convergent margins are erosive (von Huene & Scholl, 1991), yet the accretionary margins host some of the most spectacular geological structures on Earth. The physics of accretionary wedges is elegantly described by critical taper theory, which treats the wedge as a Coulomb material driven to failure by the compression at the subduction interface.
Critical Taper Theory
The critical taper model, developed by Davis, Suppe & Dahlen (1983) and Dahlen (1984), treats the accretionary wedge as analogous to a wedge of snow or sand pushed by a bulldozer blade. The wedge grows until it reaches a critical taper angle, at which the entire wedge is on the verge of Coulomb failure simultaneously. In this critical state, the wedge can slide stably along its basal décollement while maintaining its taper geometry.
The critical taper is defined as the sum of two angles: $\alpha$ (the surface slope) and$\beta$ (the basal dip of the décollement). For a non-cohesive wedge:
Critical taper condition (exact form):
\[ \alpha + \beta = \frac{(1 - \lambda_b)\mu_b}{1 + \frac{2(1-\lambda)\sin\phi}{1 - \sin\phi}} \]
where $\alpha$ = surface slope, $\beta$ = décollement dip,$\mu_b$ = basal friction coefficient, $\lambda_b$ = basal pore fluid pressure ratio, $\lambda$ = internal pore fluid pressure ratio, and$\phi$ = internal friction angle.
In the simplified limit where internal and basal pore fluid pressures are equal ($\lambda_b = \lambda$), the expression reduces to the more compact form:
Simplified critical taper:
\[ \alpha + \beta \approx \frac{(1 - \lambda_b)\mu_b}{1 + C} \]
where $C = 2(1 - \lambda)\sin\phi \, / \, (1 - \sin\phi)$ encapsulates the internal frictional properties of the wedge material. Typical values: $\phi \approx 30°$,$\mu_b \approx 0.4\text{--}0.85$, $\lambda \approx 0.4\text{--}0.95$.
Physical Interpretation
Sub-critical wedge: If the taper is less than critical (e.g., after sediment accretion at the toe), the wedge deforms internally (thrust faulting) to steepen. Super-critical wedge: If the taper exceeds the critical value (e.g., after erosion of the surface), the wedge extends by normal faulting to reduce its slope. Critical wedge: At exactly the critical taper, the wedge slides stably on its base without internal deformation.
Role of Pore Fluid Pressure
Pore fluid pressure is one of the most important controls on wedge mechanics. The pore fluid pressure ratio $\lambda$ is defined as the ratio of pore pressure to lithostatic pressure:
Pore fluid pressure ratio:
\[ \lambda = \frac{P_f}{\rho g z} \]
where $P_f$ is the pore fluid pressure, $\rho$ is the bulk density of the overlying material, $g$ is gravitational acceleration, and $z$ is depth. Hydrostatic: $\lambda \approx 0.4$; lithostatic: $\lambda = 1.0$.
High pore fluid pressure reduces the effective normal stress on both the basal décollement and internal faults. When $\lambda_b \rightarrow 1$, the effective basal friction approaches zero, producing very narrow taper angles. This explains why some wedges (e.g., Barbados) have surface slopes of only 1–2° despite relatively high material friction coefficients.
Sources of overpressure in accretionary wedges include: rapid sediment burial and compaction (porosity reduction faster than fluid escape), clay dehydration reactions (smectite → illite transition at ~60–150°C), and hydrocarbon generation. Drilling at the Barbados accretionary prism (ODP Leg 110) confirmed near-lithostatic pore pressures along the décollement.
Sandbox Analog Experiments
Sandbox models provide powerful experimental validation of critical taper theory. A layer of dry sand (Coulomb material with $\phi \approx 30°$) is pushed against a rigid backstop, naturally forming a wedge with imbricate thrust faults that closely reproduces the geometry of natural accretionary prisms. Key results include:
Self-Similar Growth
The wedge grows by adding new thrust sheets at its toe while maintaining a constant taper angle. The wedge height and length grow proportionally, confirming the scale-invariant prediction of critical taper theory.
Sensitivity to Basal Friction
Experiments with different basal materials (glass, sandpaper, Mylar) produce different taper angles, exactly as predicted. Low basal friction (Mylar) produces narrow tapers; high friction (sandpaper) produces steep tapers.
Modern sandbox experiments incorporate sieved glass beads (for lower cohesion), silicone putty (for viscous layers), and even pressurized air injection (to simulate pore fluid pressure). Particle image velocimetry (PIV) and X-ray CT scanning allow tracking of internal deformation in real time and in 3D.
Type Examples of Accretionary Wedges
Barbados Accretionary Prism
The Atlantic plate subducts beneath the Caribbean plate at ~2 cm/yr. The Barbados prism is ~300 km wide and has been intensively drilled (DSDP/ODP Legs 78A, 110, 156). It has a very low surface slope (~2°) due to near-lithostatic pore pressures on the décollement. The prism accretes hemipelagic sediments and turbidites from the Orinoco submarine fan.
Nankai Trough, Japan
The Philippine Sea Plate subducts beneath southwest Japan at ~4–6 cm/yr. The Nankai accretionary prism is one of the most studied in the world, with extensive seismic reflection, drilling (IODP NanTroSEIZE), and borehole monitoring. The wedge shows classic imbricate thrust geometry with a well-developed out-of-sequence thrust (splay fault) that may be involved in tsunami generation during megathrust earthquakes.
Makran, Pakistan–Iran
The Makran accretionary prism is the widest on Earth (~500 km from trench to backstop), built from the enormous sediment supply of the Indus and Makran river systems. The prism has a very low taper (~1.5°), thick sedimentary input (~7 km at the trench), and is partially emergent above sea level. Its great width is attributed to high sediment supply and relatively low basal friction from the thick, weak, overpressured sedimentary input.
Mélanges, Décollements & Underplating
The structural architecture of accretionary wedges involves several distinctive features:
Décollement
The basal detachment surface along which the wedge slides over the subducting plate. The décollement typically develops within a weak stratigraphic horizon — often a smectite-rich pelagic clay layer that provides low shear strength. The location of the décollement determines how much sediment is accreted versus subducted.
Tectonic Mélanges
Chaotic mixtures of fault-bounded blocks (exotic and native) set in a pervasively sheared matrix. Mélanges form within the subduction channel by intense shearing, mixing, and disruption of previously coherent stratigraphic sequences. Blocks may include basalt, chert, limestone, and high-pressure metamorphic rocks (blueschist, eclogite) in a serpentinite or shale matrix. The Franciscan Complex of California is the type example of a subduction mélange.
Frontal Accretion vs. Underplating
Frontal accretion adds material at the toe of the wedge by scraping sediment off the downgoing plate at the deformation front. Underplating adds material to the base of the wedge by duplexing — horses of sediment (or even oceanic crust) are transferred from the subducting to the overriding plate along the deeper décollement. Underplating causes uplift of the inner wedge and can elevate the wedge above sea level.
Accretion efficiency (fraction of incoming sediment accreted):
\[ \epsilon_{\text{acc}} = \frac{\dot{m}_{\text{accreted}}}{\dot{m}_{\text{input}}} = \frac{h_{\text{accreted}} \cdot v}{h_{\text{input}} \cdot v} \]
Typical values: $\epsilon_{\text{acc}} \approx 0.3\text{--}0.7$ for accretionary margins;$\epsilon_{\text{acc}} \leq 0$ (net erosion) for erosive margins like Tonga, Guatemala, and Peru.
Key Numbers
~57%
Erosive Margins Globally
1–8°
Typical Surface Slope
~500 km
Makran Wedge Width
φ ≈ 30°
Internal Friction Angle