Figure 1- Map showing the Bouguer anomaly across the Central Andes. Prominent features are the NS trending negative anomalies which Götze et all, 1994 interpreted as the absence of heavy mantle material which is displaced by an overthickened crustal root.

Figure 2 - Map showing the free air anomaly for the Central Andes.

Figure 3 - Distribution of all the seismicity recorded and located by NEIC-USGS between 1990 and 2000. In general, the seismicity is shallow near the trench and increases its depth eastward. However, shallow seismicity can be observed in the back-arc beneath the Bolivian fold and thrust belt and thick skin deformation zone below the Sierras Pampeana. Certain gaps of intermediate and deep seismicity can be observed around 18-20°S, 27°S and 30°S. However, this observation may be biased due to the lack of regional seismic networks.

Figure 4 - Distribution of the seismicity for 6 EW cross sections distributed between 22°S and 24.5°S. The lines corresponds to the top of the slab as infered in Graeber & Asch, 1999 . The bottom plot shows all the lines together showing the change of the geometry of the downgoing Nazca slab.

Figure 5 - Plot showing the station distribution of the BANJO/SEDA deployment (Beck, Zandt) over an EW trending profile located at 19-20°S. The upper part shows the topography and the location of the stations. The lower plot shows the depth of the interpreted Moho discontinuity based on receiver function analysis. The dotted line shows the predicted depth of the Moho caused by an isostatic model. In general, a good agreement can be observed between the receiver function analysis and the isostatic model. However important discrepancies can be observed beneath the altiplano and surrounding cordilleras, suggesting that the seismic Moho must not reflect necesarily a density transition, but may be related to change in rheology/phase of the envolved material. Moho depths of up to 75 km can be observed below the Western Cordillera 
(Swenson et al, 2000)

Figure 6 - The red line corresponds to the one dimensional P-wave velocity structure below the Altiplano infered from body wave inversion of regional earthquakes (Swenson et al, 2000). The blue and green lines correspond to seismic velocity gradients infered from Christensen & Mooney, 1995 for a felsic and mafic crust assuming a low, average and high heat flow. Clearly, the middle and lower crust present a strong felsic signaturedown to a depth of ~55-60km. Compared to a normal crust, these velocities are anomalously low suggesting the presence of high temperatures or partial melt (or both). The consequences of this result will be analysed together with the results of the tomography in the next section.

The following tomographic studies were done by the PISCO'94, BANJO/SEDA'95 and ANCORP'96 experiments. Here, we will limit ourself to a brief description of the results leaving the interpretation for the last section of this page where we compiled most of the available information to get a general overview of the main features of the Central Andes.

Figure 11 - Interpretation of a seismic profile at 21°-22°S (Graeber & Ash, 1999).

This figure compiles most of the results obtained by the studies mentioned above. The most prominent features can be summarized as:

• Beneath the Brazilian shield (the eastern edge of the figure), the continental crust (yellow region) is roughly 38-40 km thick.In the Subandean region it becomes gradually thicker reaching its maximum thickness of ~70km bellow the main bulge of the Andes, which corresponds to the Eastern Cardillera (EC), Altiplano and Western Cordilleras (WC). In the Precordillera the thickness diminishes and disapears completely at the trench being replaced by the underlying oceanic crust (not shown in the figure). Bellow the Subandes and Precordillera, Moho depths infered from receiver function analysis, seismic profiling (reflectors) and isostatic modeling are in excellent agreement with each other. Although a geometric asymetry can be observed between both sides of the orogen which is caused by the underlaying configuration of the subductiong Nazca plate and Brazilian shield. However, beneath both Cordilleras and the Altiplano the picture becomes much more complicated. Essentially, a transition zone of up to 10 km can be observed between isostatic and seismic Moho depths (pink region), reflecting a change in rheology, phase and/or presence of partial melt .

• The wedge constrained by a strong relfector arround 60 km and the downgoing slab below the Precordillera-Western Cordillera is caracterized high P-wave velocities, small S-wave velocities and therefore high Vp/Vs ratio. The attenuation presents no anomaly. The strong reflector is probably caused by interface between hydrated lower-crust and serpentinized mantle. Additionally, partial melt may take place as well.

• Further east, a low attenuation (high Q-values), high Vp and Vs velocity and low Vp/Vs ratio zone is observed (B). The origin of this anomaly is somewhat controversial.

• The continental lithosphere shown as a blue shaded region underthrusts? the overriding crust up 66°S. This can be infered from anomalous high velocties and low attenuation (cold material). 

• Anomaly A is characterized by low P-wave velocties and Vp/Vs ratio and high attenuation. Partial melt may take place. Boundary is controled by 1200°C isotherm. May be caused by partial melt (temperature driven) or eclogitization and consecuent delamination of the overlaying continental litosphere.

• Anomaly B may be a tip of the continental lithosphere which got trapped between the crust and slab and is detached due to the potential delamination of the lower crust/upper mantle. The exact interface between these three layers is somewhat unclear (denoted by a ?). 

• Overall, hot mantle material seems to flow into the corner defined by the continental crust and slab producing partial melt defining the present day volcanic arc within the Western Cordillera.