From Thrust to Strike-Slip: A Tentative Model for the Multi-Stage Evolution of the Sumatran Fault

Obliquity of subduction plays a key role in shaping the tectonic morphology and lithospheric structure of the overriding plate. Globally, 90% of subduction zones are oblique. The Sumatra subduction zone, formed by the oblique subduction of the Indo-Australia Plate beneath the Eurasian Plate, is the most active one that has the strongest deformation and generates numerous destructive megathrust and intracontinental earthquakes. Instead of a widespread forearc basin often developed in orthogonal subduction, a 1900-km-long trench-parallel dextral strike-slip fault—the Sumatran Fault—has greatly modified the overriding Eurasian Plate to accommodate the highly oblique relative plate motion. The Sumatran Fault is a sinusoidal fault with over 20 seismological segments which restrict most earthquake ruptures, while it can also be divided into 3 tectonic segments: the northern, central and southern segments. But it remains unclear how and when the fault initiated or interconnected.

We conducted a systematic study on low-temperature thermochronology of the Sumatran Fault. 17 samples have been collected for zircon and apatite (U-Th)/He (ZHe and AHe) dating. Three age peaks have been recognized, ~90 Ma, ~40 Ma, and 15-10 Ma, and the youngest peak exists in both ZHe and AHe data. One sample from the central segment yields the youngest AHe age at 2.5±0.11 Ma, consistent with previous AHe dating on the Sumatran Fault. To further constrain the tectonic cooling events of the Sumatran Fault, we collected samples along a fault-perpendicular cross-section and revealed age-distance correlation of AHe and ZHe data. Away from the fault, both ZHe and AHe ages increase from 84.2 Ma to 95.7 Ma, and 13 Ma to 19.3 Ma, respectively.

Our new data unravel three rapid cooling events. The first one at ~90 Ma is consistent with the emplacement of the Woyla unit, which thrust this intraoceanic arc northeastward onto the Sumatra basement during the closure of the Neo-Tethys. The second one at ~40 Ma resulted from the Wharton ridge subduction and related tectonic compression on the Sumatra Island. The last one of the Miocene age probably represents the initiation of ongoing compression in the forearc basins. The detailed analysis on the age-distance profile shows part of the Sumatran Fault may have been formed before Late Cretaceous, and it was reactivated or interconnected through rapid movement since late Miocene (~10 Ma).

Finally, we can build a tentative tectonic model for Late Mesozoic-Cenozoic evolution of the Sumatran Fault. The Late Cretaceous closure of the Neo-Tethys Ocean leads to collision of the Woyla intraoceanic arc and the Sumatra Island. This event forms a thrust-system within the Sumatra Island, leaving a weak zone for the Sumatran Fault. A compression occurred in the Eocene but only caused regional cooling. During Middle Miocene, intensive forearc compression interconnected the potential weak zone in crust and initiated the Sumatran Fault as a thrust fault. Since ~2 Ma, the Sumatran Fault became a strike-slip fault to accommodate the oblique component of the subduction.