Why there Are no Earthquakes on the Marquesas Fracture Zone

Document Type


Publication Date


Digital Object Identifier (DOI)



Although the plate tectonic paradigm does not predict relative horizontal motion between lithosphere on opposing sides of a fracture zone, the fact that younger, more rapidly subsiding seafloor lies adjacent to older seafloor implies relative vertical motion. The observation that fracture zones are notably aseismic has led to the proposition of high strength along fracture zones, such that the differential subsidence is accommodated by flexure across a locked fault. This model predicts that a ridge develops on the young side of the fracture zone flanked by a foredeep trough on the old side, with parallel warping of the Moho and large associated gravity anomalies. Previous analyses of satellite altimetric passes over Pacific fracture zones have shown that the amplitude and shape of the gravity anomalies frequently do not conform to the predictions of this simple model. One of the most notable departures has been the Marquesas Fracture Zone (MFZ), where only one limited section was determined to be “high strength.” Curiously, the only earthquake to rupture the MFZ in the last 35 years is located in this locked region. Elsewhere, where we would predict vertical slip must be occurring, the fault is aseismic. To better understand the history of vertical motion on this fault, we have analyzed geophysical data obtained during the recent survey EW9106 aboard the R/V Maurice Ewing. Our detailed Hydrosweep, gravity, and seismic data resolve this paradox by showing no evidence for vertical slip along the fault. Shear stresses caused by differential thermal subsidence do not exceed the strength of this fault. Rather, the failure of the altimetry signal to conform to the predictions of the high-strength model along much of the MFZ is caused by changes in the Pacific-Farallon pole of rotation. Reorientation of the plate boundary was accommodated by propagating rifts, intratransform tension or compression, and changes in transform offset that complicated the signal from differential subsidence across a locked fault. For example, a counterclockwise rotation of the transform in the Cretaceous caused overthrusting in the transform and thus compensation of the depth differential by flexural loading of a very young plate. A clockwise rotation of this right-stepping transform fault at about 35 Ma led temporarily to a crenulated plate boundary and later to the development of intra-transform spreading centers. After accounting for these complications, the fracture zone appears capable of sustaining at least 20 MPa of shear stress and remains locked along the entire length of the fault except perhaps locally where it passed over the Tuamotu and Society hot spots.

Was this content written or created while at USF?


Citation / Publisher Attribution

Journal of Geophysical Research, v. 100, issue B12, p. 24431-24447