Structure and geometry of the Aksay restraining double bend along the Altyn Tagh Fault, northern Tibet, imaged using magnetotelluric method

Abstract

Large restraining bends along active strike‐slip faults locally enhance the accumulation of clamping tectonic normal stresses that may limit the size of major earthquakes. In such settings, uncertain fault geometry at depth limits understanding of how effectively a bend arrests earthquake ruptures. Here we demonstrate fault imaging within a major restraining bend along the Altyn Tagh Fault of western China using the magnetotelluric (MT) method. The new MT data were collected along two profiles across the Aksay restraining double bend, which is bounded by two subparallel strands of the Altyn Tagh Fault: Northern (NATF) and Southern (SATF). Both two‐dimensional (2‐D) and three‐dimensional (3‐D) inversion models show that the Aksay bend may be the center of a positive flower structure, imaged as a high‐resistivity body extending to an ~40 km depth and bounded by subvertical resistivity discontinuities corresponding to the NATF and SATF. In the western section of the Aksay bend, both the NATF and SATF show similar low‐resistivity structure, whereas in the eastern part of the bend, the low‐resistivity anomaly below the SATF is wider and more prominent than that below the NATF. This observation indicates that the SATF shear zone may be wider and host more fluid than the NATF, lending structural support to the contention that fault slip at depth is asymmetrically focused on the SATF, even though surface slip is focused on the NATF. A south dipping, low‐resistivity interface branching upward from the SATF toward the NATF indicates a fault link between these strands at depth.

Crustal Deformation of the Altyn Tagh Fault Based on GPS

“…(2) DEFNODE has numerical problems if the fault dip is set to exactly 90°, consequently, the dip angle for the Altyn Tagh fault is set to 89°, in accordance with the fault structure from magnetotelluric data (e.g., Xiao et al, 2017). The dip angle of the southern Tien Shan fault is set to 30°, as determined by a grid search (between 5°and 85°) to minimize the misfit between the observed and modeled GPS velocities.…”

Lateral Rheology Differences in the Lithosphere and Dynamics as Revealed by Magnetotelluric Imaging at the Northern Tibetan Plateau

“…The slip rate of the KLF is much larger than that of the Haiyuan fault in the Qilian Shan area (Duvall & Clark, ) but close to that of the ATF in the section west of the Qaidam Basin (Zhang et al, ). The crust of the Altyn Tagh range also generally has high resistivity, with discontinuities below major faults (Zhang et al, ; Xiao et al, , ). The Altyn Tagh range was interpreted to be controlled by a positive flower structure (Xiao et al, , ), so we believe that the Kunlun Shan range south of the Qaidam Basin is also controlled a by flower structure in response to the northward subduction of the Indian continent below the Tibetan Plateau.…”

“…The crust of the Altyn Tagh range also generally has high resistivity, with discontinuities below major faults (Zhang et al, ; Xiao et al, , ). The Altyn Tagh range was interpreted to be controlled by a positive flower structure (Xiao et al, , ), so we believe that the Kunlun Shan range south of the Qaidam Basin is also controlled a by flower structure in response to the northward subduction of the Indian continent below the Tibetan Plateau. The branched faults in the Kunlun Shan range (including the Qiman Tagh area) are steeply converging to the KLF.…”

“…The observations were collected with ADU07 systems, which were manufactured by Metronix of Germany. The datarecording and data-processing procedures for the natural electric and magnetic fields were similar to those in our previous work (Xiao et al, 2015(Xiao et al, , 2017. A modern robust method that was introduced by Varentsov et al (2003) was used to process low-frequency data.…”

“…The transverse magnetic (TM mode) and transverse electric (TE mode) data were inverted with the nonlinear conjugate gradient (NLCG) code from Rodi and Mackie (2001). The preparation of data and initial models for 2-D inversions was similar to that in our previous work (Xiao et al, 2015(Xiao et al, , 2017. That is, ρ + fitting (Parker & Booker, 1996) was used to eliminate inconsistent points in either apparent resistivity or phase data, and the initial models were constructed with a 100-Ωm uniform half-space and incorporated topography.…”