Seismic structure in southern Peru: evidence for a smooth contortion between flat and normal subduction of the Nazca Plate

Abstract

S U M M A R YRapid changes in slab geometry are typically associated with fragmentation of the subducted plate; however, continuous curvature of the slab is also possible. The transition from flat to normal subduction in southern Peru is one such geometrical change. The morphology of the subducted Nazca Plate along this transition is explored using intraslab earthquakes recorded by temporary regional seismic arrays. Observations of a gradual increase in slab dip coupled with a lack of any gaps or vertical offsets in the intraslab seismicity suggest warping of the slab. Concentrations of focal mechanisms at orientations which are indicative of slab bending are also observed along the change in slab geometry. The presence of a thin ultra-slow velocity layer (USL) atop the horizontal Nazca slab is identified and located. The lateral extent of this USL is coincident with the margin of the projected linear continuation of the subducting Nazca Ridge, implying a causal relationship wherein increased hydration of the ridge results in the formation of the USL downdip. Waveform modelling of the 2-D structure in southern Peru using a finite-difference algorithm provides constraints on the velocity and geometry of the slab's seismic structure and confirms the absence of any tears in the slab. The seismicity and structural evidence suggests smooth contortion of the Nazca Plate along the transition from flat to normal subduction. The slab is estimated to have experienced 10 per cent strain in the along-strike direction across this transition.

Effects of change in slab geometry on the mantle flow and slab fabric in Southern Peru

“…To the south, we observe a progressive change in the dip of the slab from flat to steep. Our observations are consistent with a continuously contorted slab as reported in previous studies [ Dougherty and Clayton , ; Phillips and Clayton , ; Ma and Clayton , ]. At ~17°S we observe high shear wave velocities aligned with the 30° dipping plane of seismicity (Figure f).…”

“…While the combination of several factors such as trench retreat, suction, and ridge subduction acted together to form the flat slab, the removal of the ridge, as it moved too far south over time, caused the flat slab to fail to the north [ Antonijevic et al , ] (Figure ). In contrast, along the southern edge of the flat slab the descending plate is still continuous but sharply contorted [ Dougherty and Clayton , ; Phillips and Clayton , ; Ma and Clayton , ].…”

Flat slab deformation caused by interplate suction force

“…This is the case with the SE events because the slab is dipping toward the SE direction near the transition from flat to normal subduction (Figure ). The dip angle is evident in the cross sections of seismicity parallel to the trench and is estimated to increase from ~10° at ~250 km from the trench to ~20° at ~450 km from the trench (Figure S4 of Dougherty and Clayton []).…”

“…Instead of remaining flat or shallowly dipping down, the slab rises afterward and closely follows the shape of the Moho. In the contact region between the Western and Eastern Cordillera, the Moho warps upward and the flat slab follows this trend (Figure ) along with the seismicity (Figure S1 of Dougherty and Clayton []). The seismicity seems to form two clusters centered at 225 km and 325 km distance (Figure S4), where the largest lateral deformation of the slab occurs.…”

“…Distinguishing between the structure of flat and normal subduction zones is important for understanding the mechanisms as well as consequences of the flat subduction. With the recent deployments of seismic arrays, several studies have emerged on the structure in southern Peru where the slab transitions from normal subduction in the southeast to flat subduction in the northwest [Dougherty and Clayton, 2014;Ma and Clayton, 2014;Phillips and Clayton, 2014;Phillips et al, 2012]. The flat subduction zone structure to the north is being investigated by several studies [Antonijevic et al, 2015;Bishop et al, 2013;Eakin et al, 2015].…”