Simon Lamb

Origin of the high plateau in the central Andes, Bolivia, South America

The Bolivian Altiplano, in the Central Andes of South America, is part of the second largest high plateau on Earth. It is an elongate region of subdued relief, ∼1.2 × 105 km2 and ∼4 km above sea level, bounded by the Eastern Cordillera and volcanic arc (Western Cordillera). Here the crust is up to ∼75 km thick. We describe the Cenozoic geological evolution of this region, using a revised chronostratigraphy and an analysis of the crustal and lithospheric structure. Crustal shortening and magmatic addition and, locally, sedimentation are the main mechanisms of Cenozoic crustal thickening, leading to nearly 4 km of surface uplift since the Paleocene. Addition of mafic melts appears to be a first-order mechanism of Cenozoic crustal growth, contributing ∼40% of the crustal thickening beneath the volcanic arc. Removal of the basal part of the lithosphere may have caused two episodes of widespread arc and behind-arc mafic volcanism, at ∼23 Ma and 0 – ∼5 Ma, contributing to the surface uplift. The Altiplano originated as a sedimentary basin, several hundred kilometers wide, between the proto-Western Cordillera and a narrow zone of uplift (proto-Eastern Cordillera) farther east. The latter zone formed by inversion of the center of a wide lacustrine or marine Cretaceous - Paleocene basin close to sea-level at ∼45 Ma. A thickness of 2–4 km of Paleogene continental elastics accumulated in the proto-Altiplano basin. Subsequently, in the Oligocene, we estimate that this region and the western margin of the Eastern Cordillera were technically shortened ∼22% (∼65 km), resulting in ∼9 km of average crustal thickening. The Altiplano basin was rejuvenated at ∼25 Ma and subsequently flooded with up to 8 km thickness of detritus eroded from the uplifting Eastern and Western Cordilleras. Between ∼25 and 5 Ma, folding and thrusting in the western margin of the Eastern Cordillera migrated westward into the center of the Altiplano basin, essentially terminating deposition, except in local subbasins, and accommodating ∼13% (∼30 km) of shortening and an estimated ∼7 km of average crustal thickening. Subsequently, there has been strike-slip deformation and limited local thrusting (< 5 km of shortening). Geomorphological and geochronological evidence for 1.5–2 km of surface uplift of this region since the Late Miocene suggests ∼14 km of lower crustal thickening beneath an essentially rigid “lid”, and can be explained by ∼100–150 km of underthrusting of the Brazilian shield and adjacent regions beneath the eastern margin of the Central Andes. The present subdued relief in the Altiplano may be a result of ductile flow in the lower crust and sedimentation and erosion in an internal drainage basin.

Active tectonics of the Andes

Nearly 90 mm a−1 of relative plate convergence is absorbed in the Andean plate-boundary zone. The pattern of active tectonics shows remarkable variations in the way in which the plate slip vector is partitioned into displacement and strain and the ways in which compatibility between different segments is solved. Along any traverse across the plate-boundary zone, the sum of relative velocities between points must equal the relative plate motion. We have developed a kinematic synthesis of displacement and strain partitioning in the Andes from 47°S to 5°N relevant for the last 5 Ma based upon: (1) relative plate motion deduced from oceanic circuits giving a roughly constant azimuth between 075 and 080; (2) moment tensor solutions for over 120 crustal earthquakes since 1960; (3) structural studies of deformed Plio-Pleistocene rocks; (4) topographic/geomorphic studies; (5) palaeomagnetic data; and (6) geodetic data. We recognize four neotectonic zones, with subzones and boundary transfer zones, that are partitioned in different ways. These zones are not coincident with the ‘classic’ zones defined by the presence or absence of a volcanic chain or differences in finite displacements and strains and tectonic form; the long-term segmentation and finite evolution of the Andes may not occur in constantly defined segments in space and time. In Segment 1 (47°–39°S), the slip vector is partitioned into roughly orthogonal Benioff Zone slip with large magnitude/large slip-surface earthquakes and both distributed dextral shear giving clockwise rotations of up to 50° and dextral slip in the curved Liquine-Ofqui Fault System giving 5°–10° of anticlockwise fore-arc rotation. In Segment 2 (39°–20°S), the slip vector is partitioned into Benioff Zone slip roughly parallel with the slip vector, Andean crustal shortening and a very small component of dextral slip, including that on the Atacama Fault System. Between 39° and 34°S, a cross-strike dextral transfer, which deflects the Chile trench and the volcanic arc, absorbs the shortening contrast between Segments 1 and 2. In Segment 3 (20°–6°S), the slip vector is partitioned into roughly orthogonal Benioff Zone slip, crustal shortening, sinistral trench-parallel faulting and northeast-southwest extension. Compatibility between Segments 2 and 3 is maintained by the sinistral east-southeast-trending Cochabamba shear zone and north-trending dextral faults. In Segment 4 (6°S to 5°N), the slip vector is partitioned into roughly orthogonal Benioff Zone slip and dextral strike-slip faulting in the fore-arc and volcanic chain.

Cenozoic evolution of the Central Andes in Bolivia and northern Chile

Abstract The Central Andes in Bolivia and northern Chile form part of a wide and obliquely convergent plate-boundary zone where the oceanic Nazca plate is being subducted beneath the continental South American plate. In the latest Cretaceous and Palaeocene, this part of the Central Andes formed a volcanic arc along what is today the forearc region of northern Chile, with a wide zone of subsidence, as much as 400 km wide, at or close to sea level behind the arc. In the Eocene, the central part of the behind-arc basin was inverted to form a zone of uplift (proto-cordillera), about 100 km wide and along what is today the western margin of the Eastern Cordillera of Bolivia. The Altiplano basin and an early foreland basin were initiated at this time, receiving sediment from the Eocene proto-cordillera. Subsequently, the proto-cordillera widened, as the rate of deformation increased and deformation spread westwards into the early Altiplano basin, and also eastwards towards the Brazilian Shield. In the Late Miocene, deformation essentially ceased in the Altiplano and Eastern Cordillera. An intense zone of shortening was initiated in what is today the Subandean Zone on the eastern margin of the Central Andes, deforming the Oligo-Miocene foreland basin. Shortening in the Subandean Zone accommodated both underthrusting of the Brazilian Shield and also bending of the entire mountain belt about a vertical axis. It is suggested that much of the distinctive Cenozoic tectonic evolution of this part of the Andes is related to pre-Andean strength inhomogeneities in the South American lithosphere.

Lithospheric flexure and bending of the Central Andes

Abstract Gravity anomaly and topography data are used to define the effective elastic thickness of the lithosphere, Te, in the bend region of the Central Andes. Values of Te increase from nearly zero, north and south of the bend, to values greater than 50 km at the bend in Bolivia. There is a close correlation between Te and the style and magnitude of the shortening in the Central Andes since the Late Miocene. In the bend region, where the lithosphere is flexurally strong with large Te, foreland deformation is concentrated into a thin-skinned fold-and-thrust belt above a basal decollement which has absorbed more than 100 km of shortening. Further north and south, where the lithosphere is flexurally weak with low Te, foreland deformation is more complex, involves basement and has absorbed less shortening. The along-strike gradients in foreland shortening have accommodated both clockwise and anticlockwise rotations about a vertical axis of almost the entire width of the Bolivian Andes. We speculate that the observed variation in Te is related to the proximity of the Brazilian shield and has been an important factor in controlling the nature and amount of foreland deformation, and hence bending of the Central Andes. The flexural properties of the lithosphere may play an important role in determining the large-scale evolution of mountain belts.

A model for tectonic rotations about a vertical axis

Abstract The rotation of a rigid ellipsoidal inclusion within a highly viscous fluid, orientated so that two of the principal axes remain horizontal, is used as a model for the rotation of crustal inclusions in wide zones of continental deformation. This model is also applicable to other geological problems involving the rotation of inclusions in a matrix. The pattern of behaviour in such a model is shown to be complex. In general the rotation rate of the inclusion is a function of all components of the velocity field of the deforming medium and the horizontal aspect ratio of the inclusion. However, for a given velocity field, this aspect ratio must exceed a critical value before the inclusion can rotate continuously. Inclusions with lower aspect ratios will rotate, for a certain range of orientations, in an opposite direction to the sense of shear in the deforming zone. The possibility of the inclusion changing shape during rotation adds to the complexity of behaviour.

Active deformation in the Bolivian Andes, South America

A method is described for determining the horizontal field of velocity and velocity gradients in the Bolivian Andes in the vicinity of the Bolivian orocline, using the following data: (1) space-geodetic measurements of crustal motions; (2) the style, distribution, and rate of late Miocene-Quaternary folding and faulting; and (3) paleomagnetic measurements of rigid body rotations about a vertical axis. These data were analyzed in a network of 26 triangles which spanned the Bolivian Andes by solving the velocity gradient compatibility equations between adjacent triangles, subject to the input data constraints. This yielded 120–140 linear equations to constrain 108 unknowns, making the problem moderately overdetermined. A series of experiments were carried out with different combinations of the input data to determine the best fit, in the least squares sense, field of Quaternary horizontal velocities and velocity gradients. The azimuths of velocity vectors in the high Andes are subparallel to both the relative plate convergence vector and the symmetry axis of the Bolivian orocline, with magnitudes relative to stable South America in the range 7–17 mm yr−1, accommodating up to 30% of the relative plate convergence. However, the azimuths of velocities along the range front are nearly orthogonal to the range, with a component of distributed range-parallel dextral or sinistral shear within the Eastern Cordillera and Altiplano. The horizontal velocity field requires bending about a vertical axis of both limbs of the Bolivian orocline. The field of horizontal dilatation, assuming homogeneous deformation of the entire crustal thickness, requires vertical velocities of material points up to 2.5 mm/yr in the sub-Andean zone, decreasing to <0.1 mm/yr farther east in the Eastern Cordillera and Altiplano. The direction of maximum gradients of buoyancy stress, determined from the field of gradients of appropriate combinations of velocity gradients, using a thin sheet model of viscous flow, are broadly parallel to the direction of maximum topographic gradients in the Eastern Cordillera, providing support for the idea that the bulk behaviour of the Bolivian lithosphere is fluid-like. A comparison of these gradients with buoyancy stress gradients calculated from crustal thickness contrasts, suggest that the bulk lithospheric viscosity is in the range 0.5–1.0×1022 Pa s; gradients of velocity gradients along the eastern margin of the Bolivian Andes can be interpreted in terms of marked gradients of viscosity, with viscosity increasing toward the Brazilan shield in the cratonic core of South America. Buoyancy stresses calculated from crustal thickness contrasts, calibrated with the crustal thicknesses at the changeovers from thrust to strike-slip or normal faulting, suggest that the average lithosphericscale deviatoric stresses are in the range 7–15 MPa.

Slip in the 2010–2011 Canterbury earthquakes, New Zealand

The 3rd September 2010 Mw 7.1 Darfield and 21st February 2011 Mw 6.3 Christchurch (New Zealand) earthquakes occurred on previously unknown faults. We use InSAR ground displacements, SAR amplitude offsets, field mapping, aerial photographs, satellite optical imagery, a LiDAR DEM and teleseismic body-wave modeling to constrain the pattern of faulting in these earthquakes. The InSAR measurements reveal slip on multiple strike-slip segments and secondary reverse faults associated with the Darfield main shock. Fault orientations are consistent with those expected from the GPS-derived strain field. The InSAR line-of-sight displacement field indicates the main fault rupture is about 45 km long, and is confined largely to the upper 10 km of the crust. Slip on the individual fault segments of up to 8 m at 4 km depth indicate stress drops of 6–10 MPa. In each event, rupture initiated on a reverse fault segment, before continuing onto a strike-slip segment. The non-double couple seismological moment tensors for each event are matched well by the sum of double couple equivalent moment tensors for fault slip determined by InSAR. The slip distributions derived from InSAR observations of both the Darfield and Christchurch events show a 15-km-long gap in fault slip south-west of Christchurch, which may present a continuing seismic hazard if a further unknown fault structure of significant size should exist there.

High-altitude palaeosurfaces in the Bolivian Andes: evidence for late Cenozoic surface uplift

Abstract Low relief palaeosurfaces are widely preserved in the Cordillera Oríental of the Bolivian Andes at altitudes between 2000 m and 4000 m in a region up to 100 km wide and 600 km long. Remnants of these nearly horizontal or gently undulating erosion surfaces, from 5 to 2000 km2 in area, cut across folded Palaeozoic to Cenozoic bedrock. They appear to be erosional pediments formed between 12 and 3 Ma along broad, low gradient, valleys, which are generally undeformed. In detail, the morphology of surface remnants can be used to construct two distinct palaeodrainage basins. Very little of the material eroded during surface cutting appears to have remained within the Cordillera Oriental and must have been carried into a foreland basin, now preserved in the Subandean fold and thrust belt. The palaeovalleys are generally parallel to the dominant north-south structural grain, forming a tortuous low gradient drainage system which flowed into the foreland basin in an east-west distance of < 150 km. Reconstructions of the drainage systems suggest that they have been uplifted c. 2 km since their formation. These estimates are consistent with plausible models of uplift of the Cordillera Oriental as a consequence of intense deformation in the Subandean fold and thrust belt since c. 10 Ma. However, only since 3 Ma have the surfaces been deeply dissected by drainage cutting more directly across the structural grain. The timing of deep dissection may be partly related to a climate change to wetter and colder conditions.

Southern limit of mantle-derived geothermal helium emissions in Tibet: implications for lithospheric structure

The isotopic composition of helium emitted from geothermal springs in the southern Tibetan plateau, reported as Rc/ RA (Rc = air corrected sample 3 He/ 4 He, RA = air 3 He/ 4 He), ranges from 0.013 to 0.38, and defines two principal domains. In southernmost central Tibet, helium isotope ratios are typical of radiogenic helium production in the crust (Rc/RA 6 0.05, crustal helium domain). Further north, there is a resolvable 3 He anomaly consistent with a mantle contribution (R/RA s 0.1, mantle helium domain). The highest values of 0.27^0.38 RA occur at the southern end of the Karakoram fault. The boundary between the two domains lies 50^100 km north of the Indus-Zangpo suture zone. There seems to be no association between the 3 He anomaly and zones of active normal faulting and litho-tectonic crustal units, such as the ultramafic rocks of the Indus-Zangpo suture zone and the Gangdese intrusive belt. Although scavenging of mantle-derived helium, stored in large ultrabasic and basic intrusions in the crust, cannot be ruled out entirely, we argue that the 3 He anomaly most plausibly reflects degassing of volatiles from young (Quaternary) mantlederived melts intruded into the crust. As such, it defines the southern limit of recent mantle melting and mantle melt extraction beneath the Tibetan plateau. The southern limit of the 3 He anomaly coincides with the junction between the Indian and Asian plates, in the region where the Indian lithospheric slab steepens and is subducted beneath Tibet as suggested by seismic studies. Recent mantle melting and melt extraction is confined to the Asian mantle, but the southern limit of the melt zone may have migrated northwards during the last 10 Ma as the Indian lithosphere has progressively underthrust the Himalayas and Tibet. fl 2000 Elsevier Science B.V. All rights reserved.

Behavior of the brittle crust in wide plate boundary zones

In wide and active continental plate boundary zones, ductile flow in the deeper and strong parts of the lithosphere may control crustal deformation. This is likely if average resistive shear stresses on faults in the brittle crust are much less than 108 Pa and the underlying bulk effective viscosity is much greater than 1021 Pa s. In this case, a simple model of distributed deformation, referred to as the floating block model, may be useful. This treats the crust as an array of rotating and translating rigid blocks, which are floating on an underlying continuous flow with a constant rheology. The model is analyzed in detail in this paper because it has the potential to link detailed observations of crustal deformation with the large-scale flow. Crustal blocks are defined by at least two sets of faults. The kinematics of crustal deformation can be described in terms of the motions of these blocks. Both the relative motion on block boundaries (faults) and block tilting about a horizontal axis can be described in terms of the underlying flow and block rotation about a vertical axis. However, rotations about a vertical axis, which are an important component of the crustal deformation, will depend not only on the underlying flow but also on the shape, orientation and arrangement of the crustal blocks. The average rotation rate about a vertical axis, over finite rotations, will be significantly different from that predicted at any instant. Also, the rotation history is considerably complicated if, as is likely, the underlying flow field, or block shape, has changed with time. These aspects of crustal deformation are discussed with reference to real zones of active deformation in New Zealand, Greece and western North America.