Obliquity limit of megathrust earthquakes

Technical Summary: Obliquity limit of megathrust earthquakes

Problem Statement
The largest earthquakes on Earth occur on subduction-zone megathrusts, where tectonic plates often converge obliquely to the trench. This oblique convergence generates a trench-parallel component of relative plate motion that must be accommodated either by slip on the megathrust itself or by deformation (specifically strike-slip faulting) within the overriding plate. While early models proposed a critical convergence obliquity threshold above which strain is fully partitioned into upper-plate strike-slip faults, recent geodetic analyses indicate that upper-plate deformation occurs even at low obliquities, and the relationship between obliquity and forearc deformation varies significantly between margins. A quantitative framework linking convergence obliquity, subduction geometry, megathrust slip direction, and upper-plate deformation has been lacking. This gap hinders accurate estimations of total strain accumulation, seismic coupling, and earthquake recurrence intervals, particularly because common simplifying assumptions (that megathrust slip is either fully trench-normal or fully trench-parallel) misrepresent the actual strain accumulation.

Methodology
The authors combine three global datasets to quantify convergence and megathrust slip obliquities: the Global Centroid Moment Tensor (GCMT) catalog (1976–2020) for earthquake focal mechanisms, the Slab2 model for subduction interface geometry, and the Global Strain Rate Model (GSRM) for plate motions. For each interface thrust earthquake, they compute the convergence obliquity ($\beta$) and the earthquake slip obliquity ($\alpha$) relative to the local down-dip azimuth of the subducting slab.

To explain the observed systematics, the authors develop a mechanical work-minimization model. The model compares the tectonic work required by two end-member deformation configurations:

1. Fully Partitioned: Trench-parallel strain is released entirely by upper-plate strike-slip faulting, resulting in trench-normal megathrust slip ($\alpha = 0$).
2. Non-Partitioned: All oblique strain is released by megathrust slip ($\alpha = \beta$).

The model assumes that earthquakes release accumulated elastic strain in a manner that minimizes the total tectonic work on the coupled megathrust-upper plate fault system. By defining an effective shear resistance ratio ($R$) between the megathrust and upper-plate strike-slip faults, the authors derive an analytic expression for a dip-dependent obliquity limit ($\beta_{rb}$). They further develop a continuous formulation to identify an optimal slip obliquity ($\alpha_{opt}$) that minimizes total work.

To apply this framework, the authors estimate the population-level effective shear resistance ratio ($R^*$) for 14 global subduction zones. They use a grid search to find the $R^*$ value that places the largest fraction of observed "partitioned" events (where $\alpha < \beta$) into a model-consistent regime. They also validate these findings against independent stress proxies derived from earthquake b-values.

Key Results

• Global Systematics: Globally, most megathrust earthquakes slip less obliquely than the local plate convergence ($\alpha < \beta$), but they rarely slip in a direction normal to the trench. Segment-averaged slip obliquities typically range from 10 to 30 degrees, deviating from the common assumption of pure orthogonal slip.
• Obliquity Limit: The data supports a dip-dependent limit on megathrust slip obliquity. The analytic model predicts a "Partitioned" regime (a forbidden zone for megathrust slip) where the convergence obliquity exceeds a limit determined by the slab dip and the resistance ratio $R$.
• Regional Variability: Analysis of four representative margins (Sumatra, Aleutians, Philippines, and Kermadec) reveals distinct optimal resistance ratios ($R^*$) for each region: Sumatra (1.74), Aleutians (1.41), Philippines (1.35), and Kermadec (3.75). These values reflect local variations in megathrust strength and upper-plate rock types. However, in all cases, the observed slip obliquities align with the dip-dependent optimal slip direction predicted by the work-minimization model.
• Strain Partitioning: Across the 14 margins evaluated, approximately 68% of the 10,383 analyzed megathrust earthquakes (and over 80% of $M_w \ge 7$ events) slip less obliquely than local plate convergence, indicating that a substantial fraction of trench-parallel strain is accommodated by upper-plate deformation rather than megathrust slip.

Significance and Claims
The paper claims to provide a mechanical basis for estimating strain accumulation on megathrusts undergoing oblique convergence. By revising the classical threshold view of strain partitioning, the authors argue that the transition between partitioned and non-partitioned deformation depends not only on convergence obliquity but also critically on slab dip and the effective resistance ratio between fault systems.

The authors assert that their work-minimization framework explains why megathrust slip directions are generally intermediate between the fully partitioned and non-partitioned end members. They claim that assuming either extreme (pure orthogonal slip or full plate-parallel slip) introduces significant biases in estimating strain accumulation, seismic coupling, and recurrence intervals. For example, in the Cascadia subduction zone, assuming an optimal slip direction rather than full convergence or pure orthogonal convergence could alter recurrence time estimates by tens of percent.

Furthermore, the paper suggests that this dip-dependent mechanism provides a physical explanation for long-term forearc translation and terrane movement over geologic time, linking earthquake-scale strain partitioning to large-scale tectonic transport without requiring extreme convergence obliquity. The agreement between observed slip obliquities and the simple energetic model suggests that slab dip is a previously underappreciated controlling parameter in oblique subduction tectonics.