Geophysical Signatures and Active Tectonics at the South-Central Chilean Margin

Abstract

ing the past 24 million years. The high and broad central Andes (3–36° S, mean elevation 4 km, up to 800 km wide) are separated from the low and narrow Patagonian Andes (38–46° S, mean elevation < 1 km, ~300 km wide) by the transitional segment of the south-central or northern Pata-gonian Andes (36–42° S; Hervé 1994). This transitional seg-ment has some apparent peculiarities and is the focus of the investigations presented and discussed here (Fig. 8.1). Driven by the convergence between the Nazca and South American Plates, compressional deformation oc-curs in the central Andes. Since the Miocene, shortening and crustal thickening has taken place north of 37° S (Isacks 1988; Allmendinger et al. 1997; Jordan et al. 1997; Giambiagi et al. 2003) and decreases southwards (Ramos 1989; Vietor and Echtler, Chap. 18 of this volume) with non-partitioned deformation (Hoffmann-Rothe et al., Chap. 6 of this volume). In contrast, shortening has not occurred since the late Miocene south of 37° S and strong strain-partitioning is evident. Only the active fore-arc wedge shows high rates of uplift (~2–3 km Myr –1 ; Melnick et al. 2005) and accretion (Bangs and Cande 1997). Since the Pliocene, active deformation has manifested itself at the Liquiñe-Ofqui fault zone (Fig. 8.1; Hervé 1976), a dex-tral, strike-slip system that accommodates oblique sub-duction and the ridge-push forces of the Chile Rise (Thomson 2002; Cembrano et al. 2002). These geological characteristics of the transitional margin segment and its unknown structure and activity at depth, down to the plate interface and, especially, within the seismogenic coupling zone, motivated our investiga-tions that began in 1999. These investigations included geological field work (e.g. Melnick et al. 2005; Vietor and Echtler, Chap. 18 of this volume), active seismic experi-ments (SO-161 Shipboard Scientific Party 2002; Lüth et al. 2003a,b; Krawczyk and the SPOC Team 2003; Rauch 2005) and passive seismological observations and tomography (Yuan et al. 2000, 2006; Bohm et al. 2002; Bohm 2004), as well as three-dimensional (3-D) modeling of gravity (Ta-šárová 2004; Tašárová et al. 2006; Hackney et al., Chap. 16 of this volume) and magnetotelluric data (Brasse et al. 2006), all aimed at constraining the tectonophysical char-acteristics of this part of the Andean active margin. This paper attempts to summarize and integrate these geo-Abstract. The ISSA 2000 (Integrated Seismological experiment in the Southern Andes) and SPOC 2001 (Subduction Processes Off Chile) onshore and offshore projects surveyed the Chilean margin be-tween 36 and 40° S. This area includes the location of the 1960 earth-quake (M w = 9.5) that ruptured the margin from ~38° S southwards for ~1 000 km. Together with gravity and magnetotelluric compo-nents, the active-passive seismic experiments between 36 and 40° S provide the first, complete, high-resolution coverage of the entire seismogenic plate interface. The observed offshore mode of sediment subduction corre-sponds well with the landward extension of the reflection seismic profile at 38°S (westernmost portion of line SPOC-South), which shows material transported downwards in a subduction channel. From the slow uplift of the Coastal Cordillera, we conclude that basal accretion of parts of this material controls the seismic architecture and growth of the south Chilean crust. There is almost no seismicity observed along the entire, approximately 130 km-wide, seismogenic coupling zone. Furthermore, the study area is characterized by a 25–35 km-thick crust beneath the Longitudinal Valley, with high-conductivity zones at 20–40 km depth that correspond to large fault zones. Below the volcanic arc, the crust is generally 35–45 km thick, with a maximum thickness of 55 km at ~36°S. The slab steepens southwards along the margin (13°–21°), and a wedge-shaped body at the plate interface can be either inter-preted as hydrated mantle with 20–30% serpentinization or, when divided, as mafic crustal material in the upper part and serpenti-nized mantle in the lower part. The lower plate could suffer slab rollback while the upper-plate kinematic segmentation exhibits fore-arc extension, possibly combined with corner-flow and active lower-plate retreat. At the top of the active subduction channel, underplating, fore-arc uplift and serpentinization are key processes at the south-central Chilean margin.