Seismicity and state of stress in the central and southern Peruvian flat slab

Abstract

We have determined the Wadati-Benioff Zone seismicity and state of stress of the subducting Nazca slab beneath central and southern Peru using data from three recently deployed local seismic networks. Our relocated hypocenters are consistent with a flat slab geometry that is shallowest near the Nazca Ridge, and changes from steep to normal without tearing to the south. These locations also indicate numerous abrupt along-strike changes in seismicity, most notably an absence of seismicity along the projected location of subducting Nazca Ridge. This stands in stark contrast to the very high seismicity observed along the Juan Fernandez ridge beneath central Chile where, a similar flat slab geometry is observed. We interpret this as indicative of an absence of water in the mantle beneath the overthickened crust of the Nazca Ridge. This may provide important new constraints on the conditions required to produce intermediate depth seismicity. Our focal mechanisms and stress tensor inversions indicate dominantly down-dip extension, consistent with slab pull, with minor variations that are likely due to the variable slab geometry and stress from adjacent regions. We observe significantly greater variability in the P-axis orientations and maximum compressive stress directions. The along strike change in the orientation of maximum compressive stress are likely related to slab bending and unbending south of the Nazca Ridge.

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

“…We observe the strongest anisotropy along the southern edge of the flat slab, where the slab is heavily contorted due to a rapid change in slab geometry from flat to steep [ Ma and Clayton , ; Phillips et al , ]. The orientation of fast axes along the southernmost edge of the flat slab/transition zone at intermediate periods (58 s and 66 s, Figures c and d) is consistent with the orientation of T axes (the maximum extension direction) inferred from focal mechanism solutions [ Kumar et al , ]. The pattern shifts inland at longer periods and aligns with slab contours (Figures e and f and supporting information Figures S6c and S6d), consistent with source‐side splitting measurements [ Eakin et al , ] (Figure c).…”

“…Based on the peak sensitivities of these periods (Figure ), the observed pattern likely illuminates the subducting plate progressing to the northeast. At long periods (91–125 s) we observe a roughly N‐S alignment of fast directions along ~71°W longitude, from ~16.5°S to ~12°S, which strikes parallel to the contours of the contorted slab at greater depths [ Kumar et al , ; Scire et al , ] (dashed red area in Figure f and supporting information Figure S6c). Beneath the inboard easternmost corner of the flat slab, at ~72°W, ~13°S, we observe a trench‐parallel alignment of fast directions at long periods (91 s, Figure f, 100 s, supporting information Figure S6c) with ~2% peak‐to‐peak anisotropy associated with pronounced low Rayleigh wave phase velocities.…”

Lithospheric structure beneath the northern Central Andean Plateau from the joint inversion of ambient noise and earthquake‐generated surface waves

“…Green (circles) earthquake locations are from Kumar et al . [] and blue (circles) earthquake locations are from Instituto Geofísico del Perú. Morphotectonic provinces from Figure are shown as blue dashed lines with the locations of three 1‐D profiles (Figures a–c) shown as yellow circles and the location of five 2‐D cross‐section slices (Figure ) shown as black lines.…”

“…By including a priori information into our starting model, such as a well‐constrained Moho interface and a crustal velocity structure, we image a relativity coherent slab anomaly across most of our model space that is consistent with the predicted location based on slab earthquakes. Below ~80 km, the continuous slab anomaly (>2%) follows the Slab 1.0 contours [ Hayes et al ., ] and the earthquake locations relativity well [ Kumar et al ., ]. The apparent topography on the top of the slab (Figures b and c) and depression (Figure a) in the slab anomaly (relative to the Slab 1.0 contour) are almost certainly artifacts resulting from model smoothing integrating higher velocities above the slab topography examples (Figures b and 6c) and low velocities in the depressed slab anomaly example (Figure a).…”

Overriding plate, mantle wedge, slab, and subslab contributions to seismic anisotropy beneath the northern Central Andean Plateau

“…Event locations from the catalogs of Kumar et al . [] and the Instituto Geofisico del Perú [see Eakin et al ., ] were used in this study. CAUGHT station locations are shown with red stars; thin red lines indicate slab contours from the model of Scire et al .…”