Paleomagnetic constraints on deformation models for uppermost oceanic crust exposed at the Hess Deep Rift: Implications for axial processes at the East Pacific Rise

Abstract

[1] Studies of oceanic crust exposed in tectonic windows and in ophiolites have revealed the importance of normal faulting and attendant tilting of upper crustal rock units in the accretion process at oceanic spreading centers. We present paleomagnetic remanence data from 45 fully oriented samples from dikes, gabbros and a small number of basaltic lavas from fast spread crust exposed along the Hess Deep Rift. Over $25 km along this escarpment, dikes and dike-subparallel fault zones dip consistently away from the East Pacific Rise (EPR) while lava flows dip toward the ridge. Underlying gabbro is less deformed but contains widely spaced, low-angle fractures, tentatively interpreted as shear zones. As expected from the crustal age ($1.07-1.48 Ma), most remanence data indicate reversed polarity magnetization and are compatible with the expected range of secular variation at the site. Overly steep and directionally scattered gabbro remanence and observed low-angle shear structures within this unit are tentatively interpreted as the manifestation of three-dimensional strain along anastomosing shear zones. Although some remanence directions are incompatible with any plausible deformation history, and thus likely reflect orientation errors, the overall data set is consistent with a model involving sequential rotations on (1) outward dipping, EPR-parallel ($N-S) normal faults and (2) Hess Deep Rift-parallel ($E-W) normal faults Average rotations for these sequential events are 22°to the east (defined by the mean dike attitude) and 10°to the south (estimated by bathymetry), respectively. This model best explains the remanence data, observed dikes and lava orientations, presence of dike-parallel fault zones, and the observation of steep, little deformed dikes cutting both east dipping dikes and faults. The data support a structural model for spreading at the EPR in which outcrop-scale faulting and rotation is linked to subaxial subsidence and to consequent development of dominantly outward facing normal faults close to the spreading axis. Because these faults form within the neovolcanic zone, they are subject to burial and are expected to have subdued to little surface expression.

Paleomagnetic constraints on deformation of superfast-spread oceanic crust exposed at Pito Deep Rift

“…Observations of rare vertical dikes crosscutting this assemblage of inward-tilted lavas and sheeted dikes suggest the total accumulated thickness of lava and significant tilting of uppermost crustal units all occurred in a narrow zone (∼2 km) beneath the ridge axis. A similar uppermost crustal structure occurs in the Equatorial Pacific in crust formed at a spreading rate of 135 km Ma −1 [Lonsdale, 1988] at Hess Deep Rift [Karson et al, 2002a;Varga et al, 2004]. The lack of volcanic constructional relief (<∼200 m), and the structure of lava flows that generally increase in dip toward the ridge axis with depth imply significant (≥400 m) creation of accommodation space created by subaxial subsidence that occurs at or within ∼2 km of the ridge axis [Karson, 2002;Karson et al, 2002aKarson et al, , 2002b.…”

“…[16] A total of 62 basalt dike and 5 gabbro samples collected for paleomagnetic analysis by Alvin and ROV Jason II were fully oriented in situ using the Geocompass [Hurst et al, 1994a;Varga et al, 2004]. The strike and dip of two or more surfaces for each individual block were determined when the device was held flush against a rock surface.…”

“…Cataclastic fault zones primarily focused between individual dikes or separating panels of subparallel dikes, suggest an inseparable kinematic relationship of the damage zones and the tilting of dikes Hayman and Karson, 2009; A. Chutas, Structures in upper oceanic crust: Perspectives from Pito Deep and Iceland, unpublished Masters thesis, 122 pp., Duke University, 2007]. The mechanical anisotropy of the uppermost crust imparted by subparallel dikes and faults provides planes of weakness that preferentially slip allowing tilting to occur, similar to block rotations in other extensional settings including ophiolites and other tectonic windows [Varga, 1991;Varga et al, 1999Varga et al, , 2004. Observations of rare vertical dikes crosscutting this assemblage of inward-tilted lavas and sheeted dikes suggest the total accumulated thickness of lava and significant tilting of uppermost crustal units all occurred in a narrow zone (∼2 km) beneath the ridge axis.…”

“…[3] Tectonic windows into the oceanic crust provide the opportunity to directly investigate structural, magnetic, and compositional aspects of extensive exposures of the upper oceanic crust in situ from which to infer spreading processes [Francheteau et al, 1990;Karson, 1998;Tivey et al, 1998;Karson et al, 2002aKarson et al, , 2002bVarga et al, 2004;Larson et al, 2005;Pollock et al, 2005;Hayman and Karson, 2007]. Investigations of tectonic windows and drill cores confirm a generalized layered sequence of rock types for crust formed at intermediate to fast spreading rates that consists of basaltic lavas overlying a sheeted dike complex underlain by massive gabbro, and reveal similar structural relationships [Karson, 2002].…”

Crustal faults exposed in the Pito Deep Rift: Conduits for hydrothermal fluids on the southeast Pacific Rise

“…Most workers envision that faults nucleate within the ridge axis, dip toward the ridge axis, and then are spread to the axial flanks where they are reactivated as abyssal hill faults [Macdonald et al, 1996]. If most subaxial and abyssal-hill related strain is taken up by ''inward dipping'' faults [Carbotte and Macdonald, 1994], then the dikes in the SDC should generally dip toward the ridge either through hanging wall rotation or through a ''bookshelf'' rotation of the faults and surrounding dikes [Varga et al, 2004]. However, in all drill cores and tectonic windows, dikes dip predominantly to the east, away from the spreading axis [Karson, 2002;Chutas, 2007;Tominaga et al, 2009].…”