The Central Asia collision zone: numerical modelling of the lithospheric structure and the present-day kinematics

Tunini, Lavinia

2016-A
Descargar PDF

Resumen

La zona de colisión de Asia Central: modelado numérico de la estructura litosférica y cinemática actual

RESUMEN

Asia Central está dominada por dos importantes orógenos, el orógeno del Zagros y el sistema Himalaya-Tibet, resultantes de de la colisión de las placas Arábiga e India con el margen meridional de la placa Eurasiática.

Esta Tesis se focaliza en: 1) la caracterización del manto litosférico a través de un metódo de modelización geofísico-petrológico integrado y 2) el estudio del efecto de la estructura litosférica y de la reología en la deformación neotectónica relacionada con la convergencia de Arabia y de India respecto a Eurasia utilizando una metodología basada en la aproximación de lámina delgada (thin-sheet).

En el caso del orógeno del Zagros, los resultados revelan que el manto litosférico se adelgaza debajo de Irán Central, del Alborz y parcialmente debajo de la cordillera del Zagros. En el caso del sistema Himalaya-Tibet, los resultados indican una litosfera engrosada en el sector occidental, debajo de la cordillera Himalaya, Meseta del Tibet, Kunlun Shan y Tian Shan, y un adelgazamiento debajo de las cuencas de Tarim y de Junggar. En el sector oriental los resultados confirman que la Meseta del Tibet está suportada por una litosfera más adelgazada y caliente en el norte que en el sur. Ha sido necesario introducir variaciones laterales de composición mantélica, relacionadas con procesos del manto litosférico superior, en todos los perfiles modelados evidenciando la presencia de diferentes dominios litosféricos.

El estudio de la deformación neotectónica ha revelado el rol clave de la reología en la reproducción del campo de esfuerzos y de velocidades en Asia Central, sugiriendo una litosfera menos rígida en la Meseta del Tibet que en la meseta de Irán. En conjunto, la deformación es más rápida en la zona de colisión India-Eurasia que en la zona de colisión Arabia-Eurasia. Finalmente, la presencia de un manto adelgazado en el noreste del Tibet y la consecuente disminución de viscosidad debida al aumento de temperatura explicarían la presencia de fallas extensionales en la Meseta del Tibet y reconciliarían el modelo con los datos de flujo de calor elevado y bajas velocidades sísmicas registrados en la región.

Esta tesis ha sido financiada por el proyecto ATIZA (CGL2009-09662-BTE) y la beca FPI asociada.

Abstract

Summary

The Central Asia region is dominated by one of the largest areas of distributed deformation on Earth, which spans eastern Turkey, northern Middle East, central and south-eastern Asia, covering the central and eastern sectors of the Alpine-Himalayan mountain belt. It is composed by the Zagros orogen in the western sector and the Himalaya-Tibetan orogen in the eastern sector, which are the results of the subduction of the Tethys oceanic lithosphere towards the NNE and the subsequent collisions between Arabia and India plates with the Eurasia plate during the Cenozoic. The strong and resistant Archean-to-Proterozoic shields of Arabia and India plates collided with the complex mosaic structure of the Eurasian ancient margin, which was formed by different Gondwana-derived continental blocks accreted by Late-Mesozoic time. The collisions resulted in tectonic escapes toward lateral regions (in Anatolia and south-eastern Asia), oblique convergence in the Zagros fold-and-thrust belt, the formation of the Makran accretionary wedge, convergence in the Hindukush, shortening in the Himalaya, Karakorum and Tibetan Plateau, and the development of two indentations at the edge of the Indian sub-continent. In addition, the Zagros and Himalaya-Tibetan orogens are excellent examples of diffused deformation, with wide deforming areas in the continent interiors, and the development of other mountain belts further north with respect to the Arabia-Eurasia and India-Eurasia suture zones, such as Caucasus, Alborz, Kopet Dagh, Pamir and Tian Shan mountain belts.

The lithosphere structure plays an important role on controlling the surface deformation and its propagation to the continental interiors. The compositional and strength heterogeneities within the lithosphere directly affect to the tectonic behaviour of the region and, hence, to the evolution of the orogenic systems. This thesis focalizes on the characterization of the present-day lithospheric structure of the Zagros and the Himalayan-Tibetan orogens and the role of the lithospheric structure and rheology in the accommodation of the deformation related to the Arabia and India convergence against Eurasia.

By combining geophysical and petrological information, the crust and upper mantle of the Zagros and the Himalaya-Tibetan orogens have been characterized from the thermal, compositional and seismological point of view. Four 2-D lithospheric profiles (two crossing the Zagros orogen and other two crossing the Himalaya-Tibetan orogen) have been modelled down to 400 km depth, in which the resulting crust and upper mantle structure are constrained by available data on elevation, Bouguer anomaly, geoid height, surface heat flow and seismic data including tomography models. In the Zagros orogen, the results on the crustal thickness show minimum values beneath the Arabia platform and Central Iran (42-43 km), and maximum values beneath the Sanandaj Sirjan Zone (55-63 km), in agreement with seismic data. Major discrepancies in Moho depth from those derived from seismic data are locally found in the Sanandaj Sirjan Zone (central Zagros) and Alborz Mountains where more moderate crustal thicknesses are modelled. Results on the lithosphere thickness indicate that the Arabian lithosphere is ~220 km thick along both profiles, whereas Eurasian lithosphere is

up to ~90 km thinner, especially below the Central Iran and Alborz Mountains. The lithosphere-asthenosphere boundary (LAB) shows different geometries between the two transects. In the northern profile (northern Zagros), the LAB rises sharply below the Sanandaj Sirjan Zone in a narrow region of ~90 km, whereas in the southern profile (central Zagros), rising occurs in wider region, from the Zagros Fold-and-Thrust Belt to the Sanandaj Sirjan Zone. The best fit of seismic velocities (Vp, Vs) and densities requires lateral changes in the lithospheric mantle composition. Our results are compatible with Proterozoic peridotitic mantle compositions beneath the Arabian Platform, Mesopotamian Foreland Basin and the accreted terrains of Eurasia plate, and with a more depleted Phanerozoic harzburgitic-type mantle composition below the Zagros Fold-And-Thrust Belt and Imbricated Zone.

In the Himalaya-Tibetan orogen, the results show a Moho depth of ~40 km beneath the western Himalayan foreland basin, progressively deepening north-eastwards to ~90 km below the Kunlun Shan. Tarim Basin and Tian Shan show a nearly flat crust-mantle boundary at 50-65 km depth. The lithosphere-asthenosphere boundary lies at 260-290 km depth below the western Himalaya and Tibetan Plateau, Tian Shan and Altai Range, and it shallows to ~230 km depth below the southern Tarim Basin and to ~170 km below the Junggar region. The north-eastern Tibetan Plateau is underlined by a thinner lithosphere (LAB depth at ~120 km) with respect to its southern sector (LAB depth at ~280 km), confirming the results of previous 2D-geophysical integrated models carried out in this region. The modelled lithospheric mantle composition is generally compatible with a lherzolitic mantle-type, slightly changing to a more undepleted composition in the deep lithosphere beneath the Tarim Basin due to metasomatism. However, the mantle beneath Tian Shan, Junggar region and Altai Range is characterized by a FeO-MgO-rich composition, likely related to subduction slab-derived fluids, and the north-eastern Tibetan Plateau is highly depleted in MgO and enriched in FeO, Al2O3 and CaO, as retrieved by xenolith samples. Our results of the geophysical-petrological study finally suggest that the Himalaya-Tibetan orogen is supported by a thick buoyant lithospheric mantle in the western profile and by a lithospheric mantle thinning in the north-eastern sector of the Tibetan Plateau along the eastern profile.

The combination of the present-day lithospheric structure of the Zagros and the Himalaya-Tibetan orogens with plate kinematics, geodetic observations and stress data allowed investigating the neotectonic deformation related to the collision of the Arabia and India plates against Eurasia. A geodynamic modelling technique based on the thin-sheet approximation has been used for this purpose. Through considering the crustal and lithospheric mantle structure of the Central Asia, the topography, the surface heat flow and rheological behaviour for both crust and upper mantle depending on temperature, this method allowed inferring the surface velocity field, stress directions, tectonic regime and strain distribution by imposing velocity conditions at the model boundaries.

The results allow obtaining a first order approximation of the velocity field and of the stress directions in the whole Central Asia, reproducing the counter-clockwise rotation of Arabia and Iran, the westward escape of Anatolia, and the eastward extrusion of the northern Tibetan Plateau by only imposing the convergence of Arabia and India plates respect to the

fix Eurasia. The simulation of observed extensional tectonics within the Tibetan Plateau requires, instead, a weaker lithosphere, which can be provided i) by a change in the rheological parameters or ii) by reducing the lithosphere thickness in the NE-Tibet. Furthermore the temperature increase generated by the lithospheric thinning in the NE-Tibet would permit to reconcile the model with the high heat flow values and the low mantle seismic velocities observed in this area.

Índice

CONTENTS

Summary IV

Part I: Introduction and Geological framework 1

Chapter 1: General Introduction 3

1.1 Background and motivation 3

1.2 Objectives 6

Chapter 2: Geological setting 9

2.1 Central Asia 9

2.2 The Arabia-Eurasia collision zone 10

2.3 The India-Eurasia collision zone 14

2.4 The Arabia-India inter-collision zone 17

Part II: Present-day lithospheric structure 21

Introduction 23

Chapter 3: Method: The integrated geophysical-petrological modelling 24

3.1 Mantle temperature distribution 25

3.2 Mantle thermal conductivity 26

3.3 Densities 28

3.4 Potential fields 29

3.5 Mantle seismic velocities 30

3.6 Elevation 30

3.7 Sub-lithospheric anomalies 31

3.8 Mantle characterization 31

Chapter 4: The Zagros orogen 38

4.1 Data 40

4.1.1 Regional geophysical data 40

4.1.2 Crustal structure and depth of the Moho 42

4.1.3 Depth of the lithosphere-asthenosphere boundary 44

4.1.4 Mantle seismic velocities 44

4.1.5 Lithospheric mantle composition 46

4.2 Results 48

4.2.1 Crustal structure 48

4.2.2 Lithospheric mantle structure 49

4.2.1 Changing the lithospheric mantle composition 58

4.3 Discussion 60

4.3.1 Geophysical-petrological versus pure-thermal approaches 60

4.3.2 Crustal geometry 61

4.2.1 LAB geometry and compatibility with tomography models 62

4.4 Concluding remarks 63

Chapter 5: The Himalaya-Tibetan orogen 65

5.1 Data 67

5.1.1 Regional geophysical data 67

5.1.2 Previous studies on the crustal and lithospheric mantle structure 69

5.1.3 Upper mantle P-wave tomography 73

5.2 Results and discussion 74

5.2.1 Crustal structure 74

5.2.2 Lithospheric mantle structure 78

5.2.3 Mantle seismic velocities 81

5.2.4 Lithospheric structure variations along the strike of the

Himalaya-Tibetan orogen 83

5.3 Concluding remarks 89

Part III: Neotectonic modelling of Central Asia 91

Introduction 93

Chapter 6: Method and model construction 96

6.1 Model domain and faults 97

6.2 Model inputs. Lithosphere and thermal structure 100

III

6.3 Plate motion and boundary conditions 102

6.4 Model constraints 104

Chapter 7: Results 109

7.1 Reference model 109

7.2 Change in the rheological parameters 115

7.3 Change of the lithospheric mantle thickness in NE-Tibet 119

7.4 Changing the velocity conditions in the south-eastern boundary 123

Chapter 8: Discussion and concluding remarks 129

8.1 Discussion 129

8.2 Concluding remarks 133

Part IV: General conclusions 135

Chapter 9: General conclusions 137

List of figures and tables 143

References 151

Conclusiones

PART IV:

GENERAL CONCLUSIONS 136 Part IV: General conclusions Chapter 9: General conclusions

137

Chapter 9: General conclusions

The Central Asia collision zone has been studied in this Thesis by using two different numerical methodologies: 1) the geophysical-petrological approach which allowed the characterization of the present-day lithosphere structure beneath the Arabia-Eurasia and the India-Eurasia collision zones, and 2) the thin viscous sheet approach which allowed investigating the large scale neotectonic deformation in the whole Central Asia region.

In the integrated geophysical-petrological approach, geological, geophysical and petrological data are combined within an internally consistent thermodynamic-geophysical framework. This technique has proven to be a valuable tool for the characterization of the lithosphere structure from the thermal, compositional and seismological point of view. Since density in the lithospheric mantle is a function of P-T conditions and the chemical composition, the resulting lithospheric structure allows taking into account mineral phase changes and lateral compositional heterogeneities within the lithospheric mantle of the Zagros and the Himalaya-Tibetan orogens.

The four 2D lithospheric models are in agreement with geophysical observables, geological data and seismic tomography studies. The results show that the present-day lithosphere mantle structure of the Arabia-Eurasia and India-Eurasia collision zones are laterally-varying along the strike of the Zagros and the Himalaya-Tibetan orogens, not just in terms of crust and lithospheric mantle thickness, but also in mantle density, temperature and composition. The study also allowed distinguishing different lithospheric domains within the Zagros and the Himalaya-Tibetan orogens. The 2D-models obtained along the four selected profiles show that the ranges of the Central Asia (Zagros, Alborz, Himalaya, Tian Shan, Qilian Shan) are characterized by the presence of crustal roots, but not necessarily underlain by lithospheric mantle roots.

In the Zagros orogen, the resulting crustal thickness is smaller beneath the Arabia platform and Central Iran, and greater beneath the Sanandaj Sirjan Zone and the Alborz range. The lithospheric mantle thickness is greater beneath the Arabian Platform, the Mesopotamian Foreland Basin and the frontal sector of the Zagros range than beneath the continent interiors. In the northern profile (northern Zagros) the LAB rises sharply below the Sanandaj Sirjan Zone in a ~90 km narrow region, whereas in the southern profile (central Zagros) the thinning is smoother and affects a wider region, from the Zagros Fold-and-Thrust Belt to the Central Iran. The transition from the Arabian to the Eurasian lithospheric domain is located beneath the Zagros range, and it is marked by a change in the mantle velocity anomaly and in the lithospheric mantle composition.

In the Himalaya-Tibetan orogen, the crustal and lithospheric mantle thickness increase from the Indo-Gangetic plain to the Tibetan Plateau, but in different ways from east to west. In the western sector, the thickening is gradually and it reaches the maximum below the Part IV: General conclusions Chapter 9: General conclusions

138

northern edge of the Plateau and the western Kunlun Shan. In the eastern sector, the thickening generates sharp steps at both Moho and LAB discontinuities beneath the Himalaya Range, but it affects only the southern part of the plateau. The lithospheric mantle thins abruptly beneath the Qiangtang and the Songpan Ganzi terrains, even without comparable thinning in the crustal thickness. Beyond the Tibetan Plateau, the lithospheric mantle thins below the Tarim and Junggar basins, although with smaller extent compared to the thinning in the north-eastern Tibet. The transition from the Indian to the Eurasian lithospheric domain is located within Tibet, but with conspicuous differences in the amount of the northward underthrusting from east to west. In the western sector, the Indian lithospheric mantle underlies the whole Tibetan Plateau up to the boundary with the Tarim Basin, while in the eastern sector the underthrusting is restricted to the north up to the Bangong-Nujiang Suture. Different Eurasian domains have been also identified beneath the Tarim Basin and the Altaids region (i.e. Tian Shan, Junggar and Altai range) by means of different lithospheric mantle compositions.

The location of the lateral changes in the lithospheric mantle thickness roughly coincides with the position of the lateral variations in the lithospheric mantle compositions (Figure 9.1). Although the non-uniqueness of the compositional space remains an intrinsic problem since a wide range of compositions can explain multiple geophysical data, the chosen compositions fit the seismic velocity anomalies of both P- and S-waves and are compatible with available xenolith data and with the tectonothermal age. A generic lherzolitic mantle composition is dominant along the four profiles, being suitable for the lithospheric mantle beneath the Mesopotamian Foreland Basin and Persian Gulf, the Indo-Gangetic plain, the Himalaya Range and the western Tibetan Plateau, the Qaidam Basin, Qilian Shan, North China Block as well as the accreted terrains on the Eurasian side of the Arabia-Eurasia collision zone (Urumieh Dokhtar Magmatic Arc and Central Iran). However, the results show a change to a more fertile mantle composition beneath the Tarim Basin, and a relative enrichment in FeO and MgO beneath the northern Eurasian domains of the Himalaya-Tibetan orogen (Tian Shan, Junggar and Altai Range). Furthermore, the frontal parts of the Zagros range (Zagros Fold-and-thrust belt and Imbricated Zone) are compatible with a Phanerozoic harzburgitic-type mantle composition, and a refertilized dunitic lithospheric mantle is proposed for the thin lithospheric mantle of the north-eastern Tibetan Plateau.

In terms of major element composition, the transition from the Arabian-Indian to the Eurasia lithospheric mantle portion is recorded by a MgO depletion and Al2O3 as well as CaO enrichments in the Eurasian plate lithospheric mantle. The subductions of the Arabian and Indian plates beneath Iran and eastern Tibet (up to the Bangong-Nujiang Suture) respectively, are most probably responsible for metasomatic processes in the upper (Eurasian) plate induced by Al-Ca-rich slab released fluids. However, whether additional processes represented by partial or extensive melting, slab breakoff or delamination, have contributed or not in producing the solid assemblage defined by the NCFMAS chemical composition is hard to discern from this study. Part IV: General conclusions Chapter 9: General conclusions

139

Figure 9.1. Scheme to summarize the results on the lateral compositional variations within the lithospheric mantle obtained beneath the Arabia-Eurasia (A-A’ and B-B’ profiles) and India-Eurasia (C-C’ and D-D’ profiles) collision zones. Composition colour for the asthenosphere domain (i.e. Primitive Upper mantle) has been omitted for clarity. Part IV: General conclusions Chapter 9: General conclusions

140

In fact, mantle composition in terms of major elements should be supported by trace element geochemistry and isotopic data to distinguish between different mantle processes. In this way it would be possible to decipher the long-term chemical evolution of the subcontinental lithospheric mantle and, hence, to evaluate the mechanism responsible for the present-day lithospheric mantle composition and thickness.

The neotectonic modelling permitted to match the results on the present-day lithospheric structure of the Zagros and the Himalaya-Tibetan orogens with the present-day kinematics, GPS observations and fault activities related to the collision of the Arabia and India plate with Eurasia. The variations in the lithospheric structure obtained in the Zagros and the Himalaya-Tibetan orogens have significantly affected the lithospheric strength and, hence, the distribution of the surface deformation. The SHELLS program considers a vertically-integrated viscosity in a laterally-varying thermal and lithosphere structure and, therefore, it allows showing clearly the relative strengths of the different parts of the lithosphere. The neotectonic modelling based on this thin-shell approach has proven to be a valuable tool to investigate the effect of the lithospheric structure, rheology, boundary conditions, and friction coefficient values on the predicted surface velocities, deformation patterns, stress orientations and tectonic regime in Central Asia.

Although the results obtained using the SHELLS code do not permit leaning toward a single model to fit all the kinematic features of such a complex geodynamic setting, the study shows that a model considering a “typical” rheology (e.g. the reference model), allows obtaining a first order approximation of the velocity field and of the stress directions in the whole Central Asia. The model reproduces the counter-clockwise rotation of Arabia and Iran, the westward escape of Anatolia, and the eastward extrusion of the northern Tibetan Plateau by only imposing the convergence of Arabia and India plates respect to the fix Eurasia. The simulation of observed extensional tectonics within the Tibetan Plateau requires, instead, a weaker lithosphere, which can be provided i) by a change in the rheological parameters or ii) by reducing the lithosphere thickness in the NE-Tibet. The results show that the temperature increase generated by the lithospheric thinning in the NE-Tibet would permit to reconcile the model with the high heat flow values and the low mantle seismic velocities observed in this area. Furthermore, the thinning affects to the strain distribution within the Tibetan Plateau, showing that the northern-eastern sector is faster deforming than the southern and western parts. The anisotropic distribution of the deformation within the plateau, as well as the lateral strength variations throughout the Himalaya-Tibetan orogen, characterized by the presence of the rigid Tarim and Sichuan basins, can produce changes in the mass distribution and in the transmission of the deformation into the continent interior, with further consequences in the topography evolution of this region. In the Zagros orogen, the modelling results show lower strain rates compared to the Himalaya-Tibetan orogen, with the deformation concentrated in the ranges surrounding the rigid Central Iran block. The lithosphere weakness generated by the thinning of the lithospheric mantle is compensated by the presence of this rigid block, resulting in a lithosphere overall more viscous in the Arabia-Eurasia collision zone than in the India-Eurasia collision zone. This characteristic combined with the slower advancement of the Arabia plate against Eurasia result in a lower deforming orogen in which the tectonic Part IV: General conclusions Chapter 9: General conclusions

141

convergence of the Arabia plate is mainly accommodated at the sides of the Zagros range, i.e. toward the west in the Anatolia region and toward the east, in the Arabia-India inter-collision zone.

With this approach, it has not been possible to simulate the tight clockwise flow of crustal material around the eastern Himalaya syntaxis observed by GPS measurements. It is probably describing a surficial feature derived by the squeezing of the continental block sandwiched between the strong NE-ward advancing of the India plate and the rigid Sichuan Basin. An improvement on the program would be to implement the possibility to vary laterally the rheological parameters and the friction coefficient between the different faults inside the model. It would be more realistic, permitting to fit the observed geological slip-rates.

In summary, this Thesis has provided new insights on the present-day lithosphere structure of the Zagros and the Himalaya-Tibetan orogens, consistently with tomography images and integrated geophysical models, and new insights on the accommodation of the deformation related to the collision of the Arabia and India with Eurasia. I hope that the results obtained will help to further understand the close relations between lithospheric mantle structure, upper mantle processes and the tectonic behaviour in the Central Asia collision zone.