Revised Pacific‐Antarctic plate motions and geophysics of the Menard Fracture Zone

Abstract

A reconnaissance survey of multibeam bathymetry and magnetic anomaly data of the Menard Fracture Zone allows for significant refinement of plate motion history of the South Pacific over the last 44 million years. The right‐stepping Menard Fracture Zone developed at the northern end of the Pacific‐Antarctic Ridge within a propagating rift system that generated the Hudson microplate and formed the conjugate Henry and Hudson Troughs as a response to a major plate reorganization ∼45 million years ago. Two splays, originally about 30 to 35 km apart, narrowed gradually to a corridor of 5 to 10 km width, while lineation azimuths experienced an 8° counterclockwise reorientation owing to changes in spreading direction between chrons C13o and C6C (33 to 24 million years ago). We use the improved Pacific‐Antarctic plate motions to analyze the development of the southwest end of the Pacific‐Antarctic Ridge. Owing to a 45° counterclockwise reorientation between chrons C27 and C20 (61 to 44 million years ago) this section of the ridge became a long transform fault connected to the Macquarie Triple Junction. Following a clockwise change starting around chron C13o (33 million years ago), the transform fault opened. A counterclockwise change starting around chron C10y (28 millions years ago) again led to a long transform fault between chrons C6C and C5y (24 to 10 million years ago). A second period of clockwise reorientation starting around chron C5y (10 million years ago) put the transform fault into extension, forming an array of 15 en echelon transform faults and short linking spreading centers.

Quantifying the forces needed for the rapid change of Pacific plate motion at 6Ma

“…Although these calculations are not entirely independent from Croon et al (2008) since they use the East-West Antarctica plate circuit, they can be used for a first order test of consistency. From Doubrovine and Tarduno's (2008) finite rotations we obtain the angular velocity vector for the change in motion in a North America reference frame.…”

“…A detailed set of rotation parameters for the PacificAntarctic plates since 40 Ma has been obtained using magnetic data from GPSnavigated cruises in the sector between New Zealand, Chile, and Antarctica, largely concentrated on several major fracture zones (Croon et al, 2008). They are displayed in Fig.…”

Reconstructing plate-motion changes in the presence of finite-rotations noise

“…Over the last decades great efforts have been made to reconstruct Earth's plate motions at progressively finer temporal resolution from observations of the ocean-floor spreading 7 . In particular, the kinematics of the Eurasia/North America (EU/NA), India/Somalia (IN/SO) and Pacific/Antarctica (PA/AN) spreading systems since mid-/late Cenozoic have been recently reconstructed at unprecedented temporal resolution [8][9][10] from finite rotations of Earth's ocean-floor. These map the imprints of past geomagnetic inversions, therefore providing the cumulative displacement between two adjacent spreading plates-in the form of total angle opened and average geographical position of the rotation axis from some time in the geological past, through to the present day.…”

“…This yields small noise-tosignal ratios when computing Euler vectors directly by differentiation, with respect to time, of stage rotations obtained from two consecutive finite-rotation matrices. However, as finite rotations are reconstructed at progressively higher temporal resolution ( < 1 Myr) [8][9][10] , data noise becomes a major challenge to overcome. To date there is no solution to this problem, and the standard practise has been to smooth temporal series of single measurements [8][9][10] by averaging over 2 to 5 Myr-long time intervals, in the hope of isolating true plate motions and their temporal changes.…”

“…However, as finite rotations are reconstructed at progressively higher temporal resolution ( < 1 Myr) [8][9][10] , data noise becomes a major challenge to overcome. To date there is no solution to this problem, and the standard practise has been to smooth temporal series of single measurements [8][9][10] by averaging over 2 to 5 Myr-long time intervals, in the hope of isolating true plate motions and their temporal changes. Doing so, however, downgrades the native resolution of data-sets and, more importantly, does not provide unique kinematic reconstructions, because different averaging time-windows would yield different results.…”