INGEIS   05370
Unidad Ejecutora - UE
congresos y reuniones científicas
Upper mantle and crustal electrical conductivity structure at the southern edge of the Nazca flat slab, west-central Argentina
Beijing , China
Workshop; IAGA WG 1.2 on Electromagnetic Induction in the Earth; 2008
The Nazca Plate near 31°S subducts eastward under central Chile. Instead of dipping uniformly into the mantle, it flattens at a depth of about 100 km and does not roll over to plunge steeply until it is 800 km from the coast under Argentina. Above this flat slab, volcanism has been quiescent for about 9 MA. South of about 33°S the Nazca slab returns to a more typical dipping geometry and there is active volcanism in the Andes. Although there is almost no mantle seismicity associated with the transition between the flat and dipping portions of the slab, seismologists are in general agreement that the slab is warped and thus continuous. The strongest argument for excluding a slab tear is that seismic focal mechanisms unambiguously due to tearing have never been observed. Booker, et al. (2004) present clear evidence for a vertical current connection spatially coincident with the plunging the Nazca flat slab. To explain this observation, they discuss deep slab conductivity due to hydrous fluids, partial mantle melt due fluids escaping from the deep portion of the slab and triggering a plume of partial melt from the transition zone by the penetration of the slab. In 2004, we collected a NW-SE long-period MT profile perpendicular to the expected strike of the slab in the transition from flat to dipping. We were surprised to observe large induction vectors parallel to the expected depth contours of a warped slab. This implied a conductor parallel to our profile and perpendicular to what we expected from the strike of the expected mantle structure and also perpendicular to the Sierras Pampeanas, the dominant crustal structure. We immediately altered our field plan to include a second profile approximately perpendicular to the first. The induction vectors along this second profile reverse just south of our first profile but have a very asymmetric variation. To the north, the vector amplitudes approach 0.5 in a broad region. To the south of the reversal, the vectors are much weaker. The data from the second (NNE-SSW) profile are marginally 2D and minimum structure inversion of the TM mode plus the vertical induction vector reveals a local lower crustal conductor that may be the expression of what South American geologists call the Rio Quinto Lineament (RQL). This puzzling lineament is not coincident with proposed crustal sutures, but seems to have a major influence on late Tertiary regional tectonics: the thick-skinned deformation of the Sierras Pampeanas is prominent to its north and weak or absent to its south. In the mantle under the RQL we see a resistive structure that dips steeply southward to at least 200 km. This feature separates regions of very different mantle.   To the north, there is a moderately conductive zone between 100 and 200 km over a resistive mantle that extends to the top of the transistion zone at 400 km. This is compatible with what Booker, et al. found for mantle beneath the Nazca flat slab. To the south, the mantle is quite conductive at all depths constrained by the data. It is as if high conductivity in the transition zone wells up to the top of the mantle in the south of our profile. This too is not a complete surprise, because we are crossing the latitude of the transition between quiescent and active Andean volcanism. The deep, near vertical resistive structure under the middle of our profile is a major puzzle. It is a very robust feature. Any attempt to allow a current path across this structure at depths shallower than 200 km drastically alters the data misfit. A provocative suggestion is that it is the signature of a tear in the Nazca slab. Thus instead of a continuous warp, the slab to the south is dipping and has an asthenospheric wedge, while the slab to the north is flat. The mantle we sense north of the resistor is beneath a flat slab. The mantle we sense south of the resistor is above a dipping slab. It is clear that we cannot understand this region with one 2D inversion. Two crossed profiles are not ideal for a 3D inversion. Since 2006, we have been acquiring an array of MT sites extending from clearly south of the Andean volcanic transition to our long EW MT profile across the broadest extent of the flat slab. The nominal site separation is 50 km or less. The final four sites will be occupied in August. We expect that these data will allow us to extract a 3D image of the back arc mantle across the volcanic transition and shed light on whether the Nazca slab is warped or torn.