From rift system to doubly vergent orogen : An evolutionary model based on a case study of the Eastern Pyrenees and controlling factors from numerical models Arjan Ruben Grool

To cite this version:

Arjan Ruben Grool. From rift system to doubly vergent orogen : An evolutionary model based on a case study of the Eastern Pyrenees and controlling factors from numerical models. Earth Sciences. Université de Lorraine, 2018. English. ￿NNT : 2018LORR0037￿. ￿tel-01836205￿

HAL Id: tel-01836205 https://tel.archives-ouvertes.fr/tel-01836205 Submitted on 12 Jul 2018

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés.

AVERTISSEMENT

Ce document est le fruit d'un long travail approuvé par le jury de soutenance et mis à disposition de l'ensemble de la communauté universitaire élargie.

Il est soumis à la propriété intellectuelle de l'auteur. Ceci implique une obligation de citation et de référencement lors de l’utilisation de ce document.

D'autre part, toute contrefaçon, plagiat, reproduction illicite encourt une poursuite pénale.

Contact : [email protected]

LIENS

Code de la Propriété Intellectuelle. articles L 122. 4 Code de la Propriété Intellectuelle. articles L 335.2- L 335.10 http://www.cfcopies.com/V2/leg/leg_droi.php http://www.culture.gouv.fr/culture/infos-pratiques/droits/protection.htm PYRAMID

hèse

présentée et soutenue publiquement pour l’obtention du grade de DOCTEUR DE L’UNIVERSITÉ DE LORRAINE et L’UNIVERSITÉ DE BERGEN Spécialité GEOSCIENCES, par :

Arjan Ruben GROOL

From rit system to doubly vergent orogen: An evolutionary model based on a case study of the Eastern Pyrenees and controlling factors from numerical models

Du système de rit à l'orogène à double vergence : Un modèle évolutif basé sur l'étude de cas des Pyrénées Orientales et une étude des facteurs de contrôle à partir des modèles numériques

Soutenue le 22 janvier 2018 membres de jury M. SCHMALHOLZ, Stefan Professeur Université de Lausanne, Suisse Rapporteur M. BELLAHSEN, Nicolas Maître de Institut des Sciences de la Terre de Paris, Rapporteur Conférences France Mme ROUBY, Delphine Chercheur CNRS, Géosciences Environnement Examinateur Toulouse, France M. PIK, Raphaël Directeur de CRPG, Université de Lorraine, France Examinateur Recherche Mme FORD, Mary Professeur CRPG, Université de Lorraine, France Directeur M. HUISMANS, Ritske Professeur Université de Bergen, Norvège Directeur M. DE SAINT BLANQUAT, Chercheur CNRS, Géosciences Environnement Invité Michel Toulouse, France M. MASINI, Emmanuel Chercheur Total, Pau, France Invité

Centre de Recherches Pétrographiques et Géochimiques UMR 7358, 15 rue de Notre Dame des Pauvres 54500 Vandoeuvre-les-Nancy, France his could have been a quote by a famous scientist. Unfortunately it was behind a paywall.

i From rit system to doubly vergent orogen: An evolutionary model based on a case study of the Eastern Pyrenees and controlling factors from numerical models

Abstract he doubly vergent nature of some natural orogens is classically understood as two opposing thrust wedges (pro and retro) that comply with critical taper theory. he evidence that retro-wedges and their associated basins behave diferently from their pro-wedge counterparts has been steadily increasing over the past few decades. However, what causes an orogen to become doubly vergent is currently not well understood. Nor is the relationship between the pro- and retro-wedge during the evolution of a doubly vergent orogen. It is the aim of this work to improve our understanding of: 1) how the pro- and retro-wedges relate to each other during the orogenic process, 2) what factors control the evolution of a doubly vergent orogen and 3) a possible link between the pro- and retro-wedge. Answering these questions requires an improved knowledge of the evolution of a doubly vergent orogen. We focussed on the Eastern Pyrenees as a type example of a doubly vergent orogen, due to the large amount of available data. We performed a detailed tectonostratigraphic study of the retro-foreland of the Eastern Pyrenees (European plate), updating the interpretation based on recent insights into its hyperextended rit origins. We link the evolution of the retro-foreland to that of the pro-foreland (Iberian plate) in order to derive insight into the crustal scale dynamics. Based on cross section restoration, reconstructed shortening rates and subsidence analysis, we subdivide the East Pyrenean evolution into four phases. he irst (Late Cretaceous) phase is characterised by closure of an exhumed mantle domain between the European and Iberian rited margins, and simultaneous inversion of a salt-rich, thermally unequilibrated rit system. Shortening was distributed roughly equally between both margins during this early inversion phase. Following inversion, a quiescent phase (Paleocene) was apparently restricted to the retro-foreland. his phase may record the period of transition between inversion and full collision in the Eastern Pyrenees. he main collision phase (Eocene) records the highest shortening rates, which was predominantly accommodated in the pro-wedge. Retro-wedge shortening rates were lower than during the rit inversion phase. During the inal phase (Oligocene) the retro-wedge was apparently inactive and shortening of the pro-wedge slowed. his demonstrates that the relationship between the pro- and retro-wedges changes through time. We used lithosphere-scale thermo-mechanical numerical models to simulate the evolution of a doubly vergent orogen. Our results show a similar evolutionary pattern as observed in the Pyrenees: A roughly symmetrical rit inversion phase is followed by an asymmetric collision phase. Rit inheritance was found to be essential for enabling double vergence. Other factors, such as surface processes and thin-skinned deformation, were found to have a signiicant efect on the crustal structure and strain partitioning between both wedges. A salt décollement layer in the sedimentary cover promotes the formation of a crustal antiformal stack such as observed in the Pyrenees and Alps by forming a wide and low-taper thin-skinned fold-and-thrust belt that forces crustal deformation to focus in the hinterland. Finally, we show that the evolution of the pro- and retro-wedges is inextricably linked: events or conditions on one side of the doubly vergent orogen have an immediate efect on the other side of the orogen. his is clearly demonstrated in our models by constant variations in shortening rates of the pro- and retro-wedge in response to accretion of new pro-wedge thrust sheets. he High Atlas (Morocco) and Pyrenees can be seen as examples of symmetric rit inversion and later asymmetric collision phases, respectively. keywords modelling, retro-wedge, shortening distribution, Eastern Pyrenees, evolution ii Du système de rit à l’orogène à double vergence : Un modèle évolutif basé sur l’étude de cas des Pyrénées Orientales et une étude des facteurs de contrôle à partir des modèles numériques

Résumé

Les orogènes à double vergence sont classiquement déinis comme deux prismes critiques opposés (pro et retro) qui évoluent ensemble. Les études récentes montrent que les rétro-prismes et leurs bassins d’avant- pays associés se comportent diféremment des pro-prismes. Cependant, ni les facteurs qui mènent un orogène à devenir doublement vergent, ni la relation entre le pro- et rétro-prisme ne sont bien compris. Le but de cette étude est d’améliorer notre connaissance 1) de la relation entre le pro- et le rétro-prisme pendant l’orogénèse, 2) des facteurs contrôlant l’évolution d’un orogène à double vergence, et 3) d’un lien dynamique possible entre le pro- et le rétro-prisme. Répondre à ces questions nécessite une connaissance améliorée de l’évolution d’un orogène à double vergence. Nous nous sommes concentrés sur les Pyrénées Orientales, en raison de la grande quantité de données disponibles. Nous avons efectué une étude de terrain tectono-stratigraphique détaillée à l’est du Massif de Saint Barthelemy et dans l’avant-pays autour de Lavelanet (plaque Européenne). Notre interprétation d’une coupe restaurée intègre une coniguration crustale pré-orogenique en tant qu’une marge hyper-amincie. Nous relions l’évolution détaillée du rétro-prisme à celle du pro-prisme (plaque Ibérique), ain de mieux contraindre la dynamique à l’échelle crustale. Nous subdivisons l’évolution des Pyrénées Orientales en quatre phases. La première phase (Crétacé Supérieur) est caractérisée par la fermeture d’un domaine de manteau exhumé entre les plaques et l’inversion synchrone d’un système de rit riche en sel et thermiquement déséquilibré. Le raccourcissement était distribué de façon égale entre les deux marges pendant cette première phase d’inversion. Une phase de quiescence (Paléocène), limitée au rétro-prisme, enregistre la transition entre l’inversion et la phase de collision. La phase de collision principale (Éocène) enregistre le taux de raccourcissement le plus élevé, et était principalement accommodé dans le pro-prisme. Pendant la phase inale (Oligocène) le rétro-prisme était largement inactif et le raccourcissement du pro- prisme a ralenti. Cela démontre que la relation entre le pro- et rétro-prisme change avec le temps. Nous avons utilisé des modèles numériques 2D thermomécaniques à l’échelle lithosphérique pour simuler l’évolution d’un orogène à double vergence s’initie après avec un rit. Nos résultats montrent un modèle évolutif similaire à celui observé dans les Pyrénées Orientales avec une phase d’inversion du rit approximativement symétrique suivie d’une phase de collision asymétrique. L’héritage du rit est essentiel pour permettre le développement d’un orogène à double vergence. Des autres facteurs, comme les processus de surface et la déformation de la couverture, ont un efet signiicatif sur la structure crustale et la répartition du raccourcissement entre les deux prismes. Un niveau de décollement (sel) à la base de la couverture favorise la formation d’un empilement antiformal d’écailles crustales, similaire à la géométrie observée dans la Zone Axiale des Pyrénées, en formant un prisme à faible pente qui force la déformation crustale à se concentrer dans l’arrière-pays. Enin, nous montrons que l’évolution des pro- et rétro-prismes est inextricablement liée : des événements ou des conditions d’un côté de l’orogène ont un efet direct sur l’autre côté de l’orogène. Ceci est clairement démontré dans nos modèles par des variations constantes des taux de raccourcissement du pro- et rétro-prisme en réponse à l’accrétion dans le pro-prisme. Le Haut Atlas (Maroc) et Pyrénées peuvent être respectivement considérés comme des exemples d’inversion de rit symétrique et de phases de collision asymétrique ultérieures. mots clés géomodélisation, rétro-prisme, répartition de raccourcissement, Pyrénées Orientales, évolution Foreword and acknowledgements iii Foreword and acknowledgements

Well, this took longer than expected. he famous quote goes: “I apologise for writing such a long letter, for I had not the time to write a shorter one”. From Pliny the Younger. Or Blaise Pascal. Or Mark Twain. It depends who you ask, apparently. Not everything you hear is true, it seems. Seeing as I truly do not have the time to crat this foreword and acknowledge those who have supported me, I suppose the premise of that quote shall also be disproven by this foreword. Door in het buitenland te wonen ben ik het belang van familie steeds meer op prijs gaan stellen. Mijn bezoeken thuis waren gevoelsmatig te infrequent en te kort. Ik wil jullie bedanken voor de steun, zowel wanneer ik het nodig had als wanneer dat niet het geval was. De motivatiecampagne per post, hoe klein ook, heet me elke keer weer toch een hoop plezier bezorgd. Dank jullie wel, en tot gauw. Repeatedly have I been amazed by the amount of energy and efort invested by Mary and Ritske in supporting me during these four years (and a few months). Whenever I received comments on something I wrote, oten given to you on short notice, the quality of the response has helped me improve the text by a signiicant amount. My blatant ignorance of administrative requirements must also have caused quite some headaches, but even so, the support I have received from both of you has surpassed my expectations on numerous occasions. hank you for your patience, knowledge, and occasional motivating deadlines. hank you to Sébastien, Paul, and others at the CRPG for being friendly, welcoming, and helping me with my French. Lastly, I want to thank my friends. For avoiding asking about work when I didn’t feel like it, and discussing it when I needed it. For being warmly welcomed every time I visit, and being interesting and enjoyable company in general. here. It appears I have taken more than an hour to write this. he premise of the quote stands, for now.

Table of Contents Abstract i Résumé ii Foreword and acknowledgements iii

Résumé détaillé en français 1 Les Pyrénées comme cas d’étude 1 L’évolution nord-pyrénéenne 1 L’évolution sud-pyrénéenne 2 Modèle à l’échelle crustale 3 Modélisation de l’héritage extensif, des processus de surface et du décollement salifère 3 Résultats 3 Distribution du raccourcissement 4 Mécanisme d’empilement antiforme 4 Modélisation des contrôles de décollement sur la structure orogénique 5 Résultats 5 Comparaisons avec les systèmes naturels 5 Conclusions 6

Chapter 1 9 Introduction 1.1 Concepts 10 1.1.1 hrust wedges and critical taper 10 1.1.2 Flexural foreland basins 11 1.1.3 Pro- and retro-wedges and foreland basins 12 1.2 Techniques 13 1.2.1 Cross-section balancing 13 1.2.2 Subsidence analysis 15 1.2.3 Numerical mechanical modelling 17 1.3 Pyrenees 19 1.3.1 Present-day structure 19 1.3.2 Hyperextended rit 21 1.3.3 Convergent evolution 22 1.3.4 Shortening estimates 23 1.4 Limitations and scientiic questions 24 References 25

Chapter 2 33 Insights into the crustal-scale dynamics of a doubly vergent orogen from a quantitative analysis of its forelands: A case study of the Eastern Pyrenees 2.1 Introduction 34 2.2 Geological Setting 35 2.2.1 Plate Kinematics 35 2.2.2 Tectonic Zones and Major Boundaries 35 2.3 Methods 36 2.4 North Pyrenean Foreland 38 2.4.1 Stratigraphy of the North Pyrenean Foreland 38 2.4.2 North Pyrenean Structure 40 2.4.3 Syn-orogenic Subsidence of the North Pyrenean Foreland 44 2.4.4 Evolution of the North Pyrenean Foreland 45 5 South Pyrenean Foreland 49 2.5.1 Stratigraphy of the South Pyrenean Foreland 49 2.5.2 Structure of the South Pyrenean Foreland 49 2.5.3 Subsidence of the South Pyrenean Foreland 50 2.5.4 Evolution of the South Pyrenean Foreland 50 2.6 Discussion 53 2.6.1 Crustal-scale Model of Orogenic Evolution 53 2.6.2 Improved detail in orogen deformation history 55 2.6.3 Linked Pro- and Retro-wedge Dynamics 56 2.7 Conclusions 57 Acknowledgements 58 References 58

Chapter 3 65 Rit inheritance, surface processes, and décollement control on orogenic mountain belt structure explained by dynamic models 3.1 Introduction 66 3.2 Methods 66 3.3 Results 68 3.3.1 Model 1: Efects of extensional inheritance, salt, and surface processes. 68 3.3.2 Shortening distribution 68 3.4 Discussion 69 3.4.1 Pyrenean double vergent orogen 70 3.5 Conclusions 71 Acknowledgements 71 References 71

Chapter 4 73 Strain partitioning in doubly vergent orogens: controls of rit inheritance, surface processes and décollement using numerical models 4.1 Introduction 74 4.2 Methods 74 4.3 Results 78 4.3.1 Model descriptions 78 4.3.2 Strain distribution analysis 84 4.4 Discussion 85 4.4.1 Individual factors 85 4.4.2 Interactions 85 4.4.3 Natural systems 86 4.4.4 Limitations 90 4.5 Conclusions 90 References 90

Chapter 5 95 Conclusions and perspectives How does a doubly vergent orogen evolve? 96 Do the pro- and retro-wedge interact? 96 What factors inluence the shortening distribution, and how? 96 Secondary questions 96 Perspectives 97 Appendix A 99 Supplement to Chapter 2: Pyrenean case study Contents 100

Appendix B 117 Supplement to Chapter 3: Numerical modelling of orogen crustal structure Supplementary methods 118 Mathematical description 118 Model set up 118 Model parameters 119 Shortening distribution measurement method 119 Supplementary model descriptions 121 DR1: rit inheritance + shale + surface processes 121 DR2: no inheritance + shale + surface processes 122 References 123

Appendix C 125 Supplement to Chapter 4: Numerical modelling of strain partitioning Supplementary models 126 Model S1 127 Model S2 127 Model S3 127

Appendix D 129 Raman data Raman spectra of carbonaceous material 130 Unsuccessful analysis 130 References 130 Raman results 131

Stellingen 135

Résumé en français 1 Résumé détaillé en français

Les piliers de notre compréhension des orogènes et des principes physiques qui régissent leur formation reposent depuis longtemps sur les modèles du prisme critique de Coulomb et du bassin lexural d’avant- pays. Ces théories physiques se sont avérées particulièrement cruciales dans la description, la rationalisation, l’explication et la prédiction des géométries observées dans la nature. Cependant, il est très tôt apparu illusoire de généraliser l’application de chacune de ces théories aux prismes et bassins d’avant-pays qui leur sont associés de part et d’autre des orogènes (pro- et rétro-). Il apparait en efet que les rétro-prismes atteignent un état critique dynamique caractérisé par une ouverture généralement plus importante que celle caractéristique des pro-prismes. Par ailleurs, les rétro-bassins d’avant-pays enregistrent une histoire de subsidence bien diférente de celles propres aux pro-bassins d’avant-pays. Les relations physiques (sens large) qui existent entre pro- et rétro-prismes au sein d’un système orogénique naturel restent encore méconnues, tout comme leur évolution spatio-temporelle et les moteurs de cette évolution. Pour répondre à ces questions, il est essentiel de comparer de manière précise la structure et l’évolution de chacun de ces prismes et des bassins d’avant-pays qui leur sont associés.

Les Pyrénées comme cas d’étude Ce travail s’attache à étudier les Pyrénées Orientales, et plus particulièrement à mieux comprendre l’évolution de l’orogène pyrénéen dans sa globalité en comparant et corrélant les avant-pays nord-pyrénéen et sud-pyrénéen au moyen de deux sections longitudinales. Le système sud-pyrénéen est extrêmement bien documenté de par les conditions d’aleurement qui y sont excellentes et la stratigraphie qui y est très bien déinie. Le système nord-pyrénéen, en revanche, ne bénéicie pas de telles conditions d’aleurement et donne lieu à l’heure actuelle à de nombreuses interrogations de par sa complexité structurale et stratigraphique. Ce constat montre alors l’importance d’une étude détaillée de l’évolution tectono-stratigraphique de la Zone Nord-Pyrénéenne et du Bassin d’avant-pays Aquitain. Les Pyrénées font parties de la chaine alpine qui se forma par collision de la microplaque ibérique avec le continent européen. Cette collision succéda à un système extensif oblique qui se développa de l’Aptien au Cénomanien précoce et fut associé à l’exhumation de roches mantéliques entre les deux plaques. La mise en place de la convergence N-S entre ces deux plaques est datée à ~84 Ma. Dans les Pyrénées Orientales, l’orogène peut être divisé en six zones tectono-stratigraphiques majeures. Ainsi, du nord au sud, peuvent être rencontrés : le Bassin d’avant-pays Aquitain, la Zone Sous-Pyrénéenne, le Chevauchement Frontal Nord-Pyrénéen, la Zone Nord-Pyrénéenne, la Faille Nord-Pyrénéenne et la Zone Axiale. La Zone Sous-Pyrénéenne est une zone plissée et faillée que l’on sépare couramment du Bassin Aquitain par un chevauchement masqué en profondeur et de la Zone Nord-Pyréenne par le Chevauchement Frontal Nord-Pyrénéen. La Zone Nord-Pyrénéenne est une ceinture étroite de plissement et de chevauchement à vergence nord au front de la chaine qui comprend, dans la zone étudiée, des bassins extensifs inversés (exemple du Bloc de Fougax), des massifs cristallins inversés (exemple du Massif de Saint-Barthélémy) et la Zone Interne Métamorphique. La Zone Nord-Pyrénéenne est séparée de la Zone Axiale par la Faille Nord-Pyrénéenne, traditionnellement interprétée comme la suture entre l’Europe et l’Ibérie. A environ 50 km à l’ouest de la zone d’étude, sur le proil sismique pyrénéen profond ECORS, la Zone Axiale apparait comme un empilement antiforme de nappes de socle ibérique à vergence sud. Dans ce travail, nous avons utilisé la restauration de coupes transversales et l’analyse de la subsidence pour reconstruire l’évolution de l’avant-pays nord-est pyrénéen. Nous avons ensuite comparé nos résultats à des données de même nature, issues de l’avant-pays sud-est pyrénéen, pour mieux comprendre la dynamique d’orogènes à double vergence à l’échelle crustale. L’évolution nord-pyrénéenne L’extension au Crétacé inférieur fut principalement concentrée au sud, où l’amincissement de la croûte entraina l’exhumation du manteau entre les plaques ibérique et européenne. Cette zone de manteau exhumé aurait été recouverte par la Zone Interne Métamorphique. Les failles de Roquefeuil et de Benaix 2 Résumé en français représentent la limite nord (proximale) du système extensif et ont accomodé ~ 1,5 km d’extension N-S. Les sédiments marins profonds du Groupe Black Flysch ont été déposés dans le bassin extensif de l’Aptien au Cénomanien inférieur. Les sédiments post-rit du Groupe Gray Flysch enregistrent un approfondissement progressif des environnements de dépôts jusqu’au Turonien moyen. Cet approfondissement est aussi marqué par le recouvrement du socle varisque qui ne fut pas afecté par la phase extensive et que l’on ne retrouve pas à l’aleurement plus au nord. Au début de la convergence (84-66 Ma), la Zone Interne Métamorphique fut mise en place vers le nord, le long de la faille 3M. L’inversion, alors de faible ampleur, afecta les failles normales héritées de la marge européenne et entraina le plissement des successions mésozoïques des bassins extensifs (Nalzen, Fougax). Les taux de raccourcissement au début de la convergence sont estimés entre 0,6 et 0,3 mm/an, selon que l’on considère ou non un minimum de raccourcissement lié à la mise en place de la zone interne métamorphique. Le bassin lexural marin naissant au nord fut alimenté par des sédiments provenant de l’Est (Groupe Petites Pyrénées). Nous relions la subsidence tectonique de cette phase précoce (0,09 mm/ an) à la fois à la mise en place de la Zone Interne Métamorphique, à l’inversion précoce et à l’épaississement crustal de la marge européenne. Tout comme le met en évidence la stratigraphie, l’édiice orogénique précoce présentait alors un relief faible et restait largement sous-marin. Au début du Paléocène (66-59 Ma), aucune faille n’est active et la subsidence tectonique ralentit voire s’arrête dans l’avant-pays. Seule une mince couche de sédiments continentaux à grains ins (Groupe Aude Valley) s’y dépose. L’activité tectonique revient inalement durant le hanétien (59 Ma) avec un raccourcissement distribué dans le socle à travers tout l’avant-pays nord-pyrénéen. Les taux de raccourcissement durant cette phase (0,3 mm/an) sont plus faibles qu’au Crétacé supérieur. Le front de déformation migre vers le nord et implique progressivement le socle dans l’avant-pays (chevauchements de Tréziers et d’Orsans). Une incursion marine mineure a lieu au hanétien dans la moitié sud de l’avant-pays (Groupe Rieubach). A l’Yprésien, une élévation eustatique du niveau de la mer donne lieu à une nouvelle incursion marine de plus grande extension (Groupe Coustouge). La sédimentation continentale (Groupe Carcassonne) s’airme régionalement dès la in de l’Yprésien et se poursuit au moins jusqu’au Rupélien. Le soulèvement et l’érosion à cette période de la Zone Interne Métamorphique, du Massif de Saint Barthélémy et/ou de la Zone Axiale sont tous deux enregistrés par la présence de galets de calcaire métamorphique et de socle cristallin dans les conglomérats du Groupe Carcassonne. La subsidence tectonique à cette époque y est encore plus lente qu’au Crétacé supérieur, ne représentant que ~0,03 mm/y près du dépocentre du bassin. La dernière évidence de déformation dans l’avant-pays nord-pyrénéen est un léger plissement des niveaux supérieurs du Groupe Carcassonne, situé au-dessus du chevauchement de Tréziers et daté approximativement du Priabonien (34 Ma). En raison du soulèvement miocène du Massif Central au nord de cette partie du bassin d’avant-pays pyrénéen, l’enregistrement oligocène précoce est incomplet. Nous n’avons enregistré aucune activité de faille durant cette phase (34-28 Ma), et il semble que les poussées de Tréziers et d’Orsans aient été abandonnées à la in de la phase précédente. L’évolution sud-pyrénéenne Le début de la déformation sud-pyrénéenne ne peut être ni identiié dans l’enregistrement préservé de la subsidence ni clairement distingué du reste de nos données. Cependant, un bassin lexural formé au Crétacé supérieur est préservé en tant que nappe de chevauchement allochtone (Pedraforca Inférieure). Nous interprétons ce bassin lexural comme une conséquence de l’inversion précoce de la marge distale ibérique. Un bassin extensif formé au Crétacé inférieur a été inversé au Crétacé supérieur et au Paléocène (Pedraforca Supérieure, ~84-56 Ma). Le raccourcissement accommodé durant cette phase d’inversion précoce est d’environ 10 km, ce qui donne un taux de raccourcissement de ~ 0,4 mm/an. Au début de l’Éocène, le front de poussée dans les Pyrénées méridionales avance rapidement. Le raccourcissement atteint un maximum de 4,0 mm/an et s’accompagne d’une augmentation de la subsidence Résumé en français 3 tectonique (0,53 mm/y au maximum) directement au sud du front de poussée. Ceci conduit à la formation d’un bassin marin profond. Le raccourcissement de l’avant-pays sud-pyrénéen ralentit considérablement après la in de l’Yprésien, chutant à 1,0 mm/an. Cependant, à partir de ~50 Ma, le raccourcissement interne de la Zone Axiale accommode en moyenne 1,1 mm/an de plus (~ 23 km au total), comme en témoigne le début de son exhumation. La sédimentation dans les Pyrénées méridionales devient continentale après ~36 Ma, après que le soulèvement des Pyrénées Occidentales ait isolé de l’océan Atlantique le Bassin d’avant- pays de l’Èbre. Le raccourcissement inal est alors principalement accommodé par un plissement de détachement au-dessus d’un décollement évaporitique. Dans cette partie de l’avant-pays sud-pyrénéen, les plus jeunes sédiments préservés (Rupélien) sont déformés, suggérant une déformation continue jusqu’au moins ~28 Ma. Modèle à l’échelle crustale La comparaison des évolutions de chacun des deux avant-pays, décrites ci-avant, nous permet de diviser l’évolution des Pyrénées Orientales en quatre phases. La première phase (Crétacé supérieur) se caractérise d’une part par la fermeture d’un domaine de manteau exhumé entre les plaques ibérique et européenne, et d’autre part par l’inversion d’un système extensif riche en sel et non équilibré thermiquement. Le raccourcissement global (~1 mm/an) est alors réparti à peu près équitablement entre les deux marges durant une vingtaine de millions d’années. Vient ensuite la deuxième phase, à savoir une phase de quiescence (Paléocène) apparemment restreinte au rétro-système d’avant-pays et marquée par une déformation lente et continue dans le pro-système d’avant-pays (~0,4 mm/an). Dans les Pyrénées Orientales, cette deuxième phase précède le début de la collision principale (Eocène) qui enregistre le taux de raccourcissement moyen le plus élevé (~3,1 mm/an). Durant cette troisième phase, le raccourcissement est principalement accommodé dans le pro-système d’avant-pays. Enin, au cours de la phase inale (Oligocène), le rétro- système d’avant-pays apparait inactif et le raccourcissement du pro-système d’avant-pays ralentit (~2,2 mm/an). En ignorant le raccourcissement que pourrait engendrer la fermeture du domaine de manteau exhumé, le raccourcissement total des Pyrénées Orientales est d’environ 111 km. Le rétro-système d’avant- pays accommode au total environ 20 km de raccourcissement, principalement le long du Chevauchement Frontal Nord-Pyrénéen. Nous voyons ainsi que le raccourcissement, à l’origine réparti de manière équitable de part et d’autre de l’orogène, devient bien plus prononcé dans le pro-système d’avant-pays. Cette modiication de la distribution du raccourcissement coïncide avec le début de la subduction de la croûte inférieure ibérique et du manteau lithosphérique (collision). Par conséquent, le changement dans la distribution de raccourcissement peut être dû à l’atteinte de seuils internes intrinsèques aux systèmes extensifs inversés, sans pour autant dépendre d’un quelconque forçage externe tel que la cinématique des plaques.

Modélisation de l’héritage extensif, des processus de surface et du décollement salifère Nous avons utilisé des modèles numériques pour tester l’hypothèse que les systèmes extensifs inversés évoluent en orogènes à double vergence avec un changement dans la distribution de raccourcissement. En outre, nous avons cherché à expliquer le mécanisme responsable de la création d’une pile crustale antiforme, telle que celle formant la Zone Axiale dans les Pyrénées. Nous avons utilisé des modèles numériques thermo-mécaniques de haute-résolution, ce qui nous a permis de modéliser la déformation à l’échelle lithosphérique tout en conservant une bonne résolution des déformations dans les ceintures de plissement et de chevauchement et dans les bassins sédimentaires associés. Résultats Les efets de l’héritage extensif, du décollement salifère et des processus de surface aboutissent à une évolution en trois phases. La phase 0 est la phase d’extension pré-orogénique, où 50 km d’extension résultent en un système extensif étroit, à peu près symétrique et délimité par deux zones de cisaillement friction- plastique d’échelle crustale qui s’enracinent dans la croûte moyenne, plus faible rhéologiquement. Deux zones de cisaillement friction-plastique conjuguées dans le manteau lithosphérique permettent l’extension dans ce dernier. 4 Résumé en français

Pendant la phase 1, le raccourcissement lithosphérique initial conduit à une inversion symétrique de la zone extensive. Les zones de cisaillement extensives héritées sont préférentiellement réactivées, ce qui entraine la soulèvement d’un bloc central en clé de voûte. La distribution de la vitesse de déformation montre que les deux zones de cisaillement conjuguées dans la croûte supérieure et le manteau supérieur lithosphérique sont simultanément activées pendant cette première phase d’inversion. La phase 2 comprend le développement d’un édiice orogénique asymétrique à l’échelle crustale. Après une inversion symétrique, la localisation de la déformation sur une seule zone de cisaillement à grande échelle déclenche une subduction asymétrique de la croûte inférieure et du manteau lithosphérique, ce qui donne lieu à l’abandon des autres zones de cisaillement précédemment actives. Les premières ceintures étroites de plissement et de chevauchement dites de couverture se développent des deux côtés, s’enracinant à la fois dans les zones de cisaillement profond d’arrière-pays et dans le glissement de couverture du bloc central en clé de voûte. Dans le pro-système d’avant-pays, la sédimentation syn-tectonique favorise une large ceinture de plissement et de chevauchement. La déformation de socle se propage vers l’extérieur de l’orogène au-dessous de la ceinture de couverture en accrétant de nouvelles nappes de croûte supérieure. Les plus anciennes nappes du pro-prisme migrent lentement sur la plaque supérieure, réactivant la zone de cisaillement du rétro-prisme. Après 230 km de convergence, un découplage eicace le long du décollement salifère et une sédimentation sur le prisme, en partie distale de ce dernier, facilitent la propagation de la ceinture de plissement et de chevauchement de couverture du pro-prisme vers l’extérieur de l’orogène. Les nappes de socle du pro-prisme forment un empilement antiforme dont la croissance se fait par accrétion basale. Distribution du raccourcissement La répartition du raccourcissement entre le pro-prisme et le rétro-prisme est décrite ci-dessous. L’inversion précoce de la phase 1 est symétrique, les deux côtés accommodant environ 50% de la convergence totale. Après le début de la subduction de la croûte inférieure et du manteau lithosphérique durant la phase 2, le pro-prisme accommode ~80% de la convergence totale tandis que le rétro-prisme en accommode seulement 20%. Les modèles sans héritage extensif ne montrent pas ce changement dans la distribution du raccourcissement. Mécanisme d’empilement antiforme La formation d’un empilement antiforme est le résultat des efets combinés d’un décollement salifère, de la sédimentation syn-tectonique et de l’érosion. Le décollement salifère crée un système à deux prismes : 1) Un prisme de socle caractérisé par un biseau critique relativement élevé et contrôlé par le décollement en croûte moyenne (« prisme crustal »), et 2) un prisme de couverture caractérisé par un biseau critique faible (« prisme de couverture ») contrôlé par le décollement salifère. Puisque les volumes élémentaire de ces prismes ignorent leurs voisins et la géométrie globale du système, ils ne se déforment ou ne se déplacent que sous l’inluence des forces suivantes : (a) le frottement le long de leur décollement basal respectif, (b) les contraintes liées à la convergence et (c) la charge de la pile sédimentaire sus-jacente. Le prisme de couverture recouvre le socle, contribuant ainsi à cette charge de la pile sédimentaire sus-jacente, et devrait donc être considéré comme faisant partie du prisme crustal. Efectivement, le système s’agence comme un prisme au sein d’un prisme, plutôt que comme un prisme au sommet d’un autre. Cette disposition signiie que les deux prismes partagent le même angle alpha (inclinaison de la surface sommitale), à savoir celui dicté par le décollement le plus faible. Le prisme de couverture dicte ainsi l’ouverture du biseau en dessous de celle du biseau critique du prisme crustal. Le prisme crustal se déforme de manière interne (raccourcissement interne) pour augmenter l’ouverture du biseau. Cependant, le prisme salifère ne peut pas supporter cette ouverture élevée, et se déforme pour maintenir une ouverture faible, obligeant le prisme crustal à se déformer avec un raccourcissement plus interne. La sédimentation syn-tectonique réduit également l’ouverture dans l’avant-pays, tandis que l’érosion la réduit dans l’arrière-pays. Combinés, ces efets aboutissent inalement à la formation d’une pile antiforme de nappes de socle. Résumé en français 5

Modélisation des contrôles de décollement sur la structure orogénique D’après notre modèle, que nous avons décrit précédemment, l’héritage extensif semble permettre la formation d’un orogène à double vergence avec une évolution et une distribution des déformations similaires à celles observées dans les Pyrénées. Cependant, la répartition des contraintes entre le pro- prisme et rétro-prisme varie considérablement d’un système naturel à l’autre. En outre, notre modèle montre que la déformation de couverture peut contrôler la structure crustale. Par conséquent, nous avons étudié l’inluence de plusieurs facteurs sur la distribution des déformations et l’évolution des orogènes à double vergence. Nous nous sommes concentrés sur l’inluence de la rhéologie et de la distribution d’un niveau de décollement sédimentaire. Résultats Les modèles sans héritage extensif aboutissent à un orogène hautement asymétrique où ~95% du raccourcissement est accommodé dans un pro-prisme très large. Tous les modèles avec héritage extensif aboutissent à une évolution en trois phases que nous avons explicité ci-dessus. Un décollement schisteux engendre des ceintures de plissement et de chevauchement de couverture aussi bien dans le pro-système d’avant-pays que dans le rétro-système d’avant-pays. Ces ceintures de déformation ne sont actives qu’à proximité de la zone de cisaillement frontale qui implique le socle. Elles sont abandonnées dès qu’une nouvelle zone de cisaillement plus distale est formée dans le pro-prisme. Un décollement salifère, plus eicace dans le découplage de la couverture sédimentaire par rapport à son socle, permet le glissement gravitationnel, ce qui crée des ceintures de couverture plus larges. Ces dernières transportent la couverture sédimentaire du coeur de l’orogène vers l’avant-pays externe, ce qui conduit à un pro-prisme crustal plus étroit. La sédimentation syn-tectonique stabilise ces ceintures de plissement et de chevauchement de couverture, ajoutant encore à la charge d’avant-pays. L’érosion est plus eicace dans la zone interne de l’orogène et engendre un orogène plus étroit. Lorsque les processus de surface ne sont pas pris en compte, le décollement salifère favorise un raccourcissement dans le rétro-prisme plus important que ne le fait le décollement schisteux. Cependant, lorsque les processus de surface sont considérés dans la modélisation, quelle que soit la rhéologie du décollement, le pro-prisme accommode une fois encore ~80% de la convergence totale tandis que le rétro-prisme en accommode seulement 20%. Une distribution asymétrique du décollement peut inluencer la polarité de la plaque, plaçant préférentiellement le décollement dans le pro-système. Cependant, d’autres facteurs dont nous n’avons pas tenu compte peuvent avoir une plus grande inluence sur la polarité des plaques. Comparaisons avec les systèmes naturels Nos modèles reproduisent systématiquement une évolution en trois phases lorsque l’héritage extensif est considéré. Ceci nous permet de proposer un modèle évolutif générique pour les systèmes extensifs inversés comme suit : (1) la phase 0 est la phase d’extension, (2) la phase 1 voit une inversion à peu près symétrique des structures extensives héritées, et (3) la phase 2 est la phase de collision asymétrique, où la subduction continentale entraine une accommodation d’environ 80% de la convergence totale dans le pro-prisme. Nous avons comparé plusieurs systèmes naturels au modèle proposé. Le Haut Atlas (Maroc) est un système extensif inversé d’âge alpin. Les structures extensives héritées ne sont pas complètement inversées et le raccourcissement global est très faible (~26 km). Les failles crustales du côté nord plongent vers le sud et vice versa. La répartition du raccourcissement est symétrique, avec ~13 km de raccourcissement accommodés de chaque côté d’un bloc central. Nous proposons que le Haut Atlas se conforme à la phase 1 de notre modèle évolutif. L’orogène pyrénéen, comme nous l’avons décrit précédemment, est également un système extensif inversé. Les estimations de raccourcissement total vont de ~100 km à ~165 km dans les Pyrénées Centrales. Dans le cas d’un raccourcissement total maximal (~165 km), le pro-prisme accommode ~128 km de raccourcissement tandis que le rétro-prisme en accommode ~37 km. Notre reconstruction dans les Pyrénées Orientales se traduit par ~ 111 km de raccourcissement total sans tenir compte de la fermeture du domaine de manteau exhumé. Le pro-prisme accommode alors ~92 km de raccourcissement alors que le rétro-wedge 6 Résumé en français

en accommode ~19 km. L’estimation de la distribution de raccourcissement dans les Pyrénées Centrales et notre estimation pour les Pyrénées Orientales sont très proches de la distribution pro/rétro avoisinant les 80%/20% prédite par la phase 2 de notre modèle évolutif. Nous observons également le passage d’un raccourcissement à peu près symétrique à un raccourcissement asymétrique dans les Pyrénées Orientales. Par conséquent, nous proposons que les Pyrénées se conforment à la phase 2. Les Alpes sont plus complexes, de par la convergence bien plus importante qu’elles ont connue et de la mise en place d’une asymétrie de la subduction océanique avant la collision. Notre modèle prédirait une distribution fortement asymétrique du raccourcissement, ce qui n’est pas le cas. Lorsque l’on ne tient compte que du raccourcissement post-suture, on estime que les Alpes centrales peuvent accommoder environ 164 km de raccourcissement : ~108 km dans le pro-prisme (plaque européenne) et ~56 km dans le rétro-prisme. Les estimations de raccourcissement varient entre les auteurs et la distribution de raccourcissement peut même être symétrique. Dans le cas des Alpes, nous suggérons donc l’éventualité que notre modèle évolutif pour les systèmes extensifs inversés ne soit pas applicable.

Conclusions Les travaux résumés ci-dessus ont conduit aux nouvelles idées décrites ci-après. D’une manière générale, les Pyrénées Orientales ont évolué en deux phases. La phase d’inversion précoce fut caractérisée par une distribution du raccourcissement à peu près similaire de part et d’autre de l’orogène, entre les marges européenne et ibérique. La phase de collision fut asymétrique avec environ 80% du raccourcissement accommodé dans le pro-système (Ibérie) pour 20% dans le rétro-système. Cette asymétrie résulte du retournement du pro-prisme, alors trop épaissi, sur la plaque supérieure (Europe). La transition entre ces deux phases coïncide avec le début de la subduction continentale de la croûte inférieure ibérique et du manteau lithosphérique. Les modèles numériques qui testent un héritage extensif reproduisent une évolution biphasée similaire à celle observée dans les Pyrénées Orientales, avec des distributions de raccourcissement comparables. Ceci supporte l’hypothèse que le passage d’un raccourcissement équitablement réparti de part et d’autre de l’orogène à un raccourcissement asymétrique peut résulter de la seule atteinte d’un seuil interne intrinsèque aux systèmes extensifs inversés, indépendamment d’un quelconque forçage externe tel que la cinématique des plaques. Nos résultats montrent également que la déformation de couverture peut inluencer la structure impliquant le socle. Associé à la sédimentation syn-tectonique et à l’érosion, le découplage eicace de la couverture sédimentaire favorise la formation d’un empilement crustal antiforme. Enin, nos modèles montrent que l’évolution du pro-prisme et celle du rétro-prisme sont inextricablement liées : un événement ou une condition modiiée d’un côté d’un orogène a double vergence a un efet immédiat de l’autre côté. L’analyse comparative du comportement des deux prismes révèle l’auto-organisation d’un système complexe. L’application à d’autres orogènes d’une approche de détail telle que celle présentée dans ce travail peut potentiellement permettre d’identiier une auto-organisation comparable et des seuils intrinsèques à l’orogenèse comparables. Résumé en français 7

Chapter 1

Introduction 10 Chapter 1. Introduction

his thesis investigates the relationship between, and evolution of, pro- and retro-wedges in doubly vergent orogens and how this is inluenced by a number of factors, in particular extensional inheritance and the rheology of a décollement in the sedimentary cover. he approach uses numerical modelling coupled with a detailed study of both sides of the eastern Pyrenean orogen. he work is based around several concepts and techniques that have been developed to describe the processes involved in orogenesis. hese concepts and techniques are introduced in this chapter. A summary of the state of knowledge on the Pyrenean orogeny provides the regional context for the East Pyrenean case study (Chapter 2). Finally, several limitations of the concepts and techniques explained in this chapter are outlined, and the scientiic questions addressed in this thesis are presented.

1.1 Concepts 1.1.1 hrust wedges and critical taper

Figure 1.1. Example of a natural fold-and-thrust belt taking on a wedge shape. Ater Suppe [1980].

To create a mountain belt, the crust has to be shortened and thickened. In the brittle part of the crust, this shortening is accommodated along a series of thrust faults that stack slices of crust on top of each other. hese thrust systems oten result in an approximate wedge shape that tapers toward the foreland (Figure 1.1) [e.g., Suppe, 1980]. Critical taper theory is a mathematical model, developed in the 1980’s, that describes in mechanical terms how these wedges form, and what factors inluence the shape of the wedge [e.g., Davis et al., 1983; Dahlen, 1984, 1990]. he wedge shape is deined by the slope of the top surface (angle ) and the slope of the basal décollement surface (angle ). he taper is the sum of these angles ( ; Figure 1.2a). A critically tapered wedge will grow self-similarly to maintain the same taper. Note that under most conditions, there are two values for critical taper, and a wedge with a taper between those two values is stable and will not deform internally (Figure 1.2b) [Dahlen, 1984]. For frontal accretion, it suices to only regard the lower of the two critical tapers, as this is the irst taper a newly formed wedge will encounter. If a wedge has a subcritical taper, it will deform internally and not slide along the basal décollement until the taper has increased to become critical. If a wedge is supercritical, it will either extend internally (taper above stable ield) or slide along the décollement without deforming (taper in stable ield) until the taper is lowered through frontal accretion and/or surface erosion. Assuming a Mohr-Coulomb rheology, the critical taper depends on the relative magnitudes of the coeicient of friction along the basal décollement and the angle of internal friction of the wedge material [Dahlen, 1990]. Lowering the coeicient of friction along the basal décollement will reduce critical taper, and increasing it will increase critical taper. Lowering the angle of internal friction of the wedge material (making it weaker) will increase critical taper, because that makes it easier to deform the wedge than to slide along the basal décollement. he wedge will then deform internally until the new, steeper critical taper has been reached. However, if the wedge material is weakened to the point it can no longer support the topographic slope ( ), the wedge will collapse. Increasing the angle of internal friction of the wedge material lowers the critical taper. he Chapter 1. Introduction 11

Figure 1.2. a) Cross-section through a critically tapered wedge. b) stability ield of the wedge. Ater Buiter [2012]. wedge material is now strong enough to withstand the horizontal stress that drives convergence and it slides along the basal décollement instead. hese relationships are clear in the mathematical expression of this model [Dahlen, 1990]:

(1.1) where is the angle of internal friction of the wedge material and is the friction coeicient along the basal décollement. his equation assumes noncohesive Coulomb failure of dry sand, and uses several simpliications for the case that [Dahlen, 1990]. Improvements on this basic mathematical model account for cohesion and pore luid pressure [Davis et al., 1983; Dahlen, 1990]. In thrust wedges above a salt décollement, basal friction is not dependent on normal stress. he yield strength of salt instead depends on temperature and strain rate. If one assumes a constant yield strength , the critical taper of the wedge becomes [Davis and Engelder, 1985]:

(1.2) where is rock density, the depth to the basal detachment, and the pore luid pressure relative to lithostatic pressure. If =1 MPa, appropriate for salt, the resulting taper is very low, ~1° [Davis and Engelder, 1985]. Following the development of critical taper theory, further investigations aimed to explain the inluence of other factors, such as syn-tectonic sedimentation [e.g., Storti and McClay, 1995; Ford, 2004], erosion [e.g., Persson and Sokoutis, 2002; Konstantinovskaya and Malavieille, 2005], and a ductile crustal root [Willett et al., 1993]. Sedimentation efectively strengthens the distal wedge and undeformed foreland, focusing deformation in the interior of the wedge [Storti and McClay, 1995; Fillon et al., 2013b]. Erosion reduces the topographic slope ( ), making the wedge subcritical. Internal deformation to restore taper then results in a smaller wedge that exhumes deep materials [Persson and Sokoutis, 2002]. For an overview of wedge models, see Buiter [2012]. 1.1.2 Flexural foreland basins he association of foreland basins with orogens has long been recognised [e.g., Abouin, 1965]. A lexural model of foreland basins was developed in which the topographic load of an adjacent orogen creates a downward lexure of the crust, thus forming a foreland basin (Figure 1.3) [Walcott, 1970; Price, 1973; Beaumont, 1981]. his link between orogen growth and foreland lexure means that investigating the foreland basin can reveal information about the growth of the orogen. In the lexural foreland model, the width and depth of a foreland basin depend on two things: the magnitude of the load, and the lexural 12 Chapter 1. Introduction

Figure 1.3. he delection of an elastic plate under an applied load, forming a lexural foreland basin. he diferent lines show how the delection depends on the lexural rigidity. Ater Beaumont [1981]. rigidity of the plate (oten expressed as ‘efective elastic thickness’ [Watts, 1992]). he larger the load, the greater the downward delection of the plate, thus the deeper the basin. he more rigid the plate, the shallower the basin, but also the wider the lexed zone (increased diameter of the curve in the lexed plate). his model has been reined and adapted for a broken plate that is unconnected at the edge below the load [Walcott, 1970; Turcotte and Schubert, 1982], a load that is distributed along the lexed plate instead of placed at the edge [Beaumont, 1981], lateral variations in lexural rigidity, and buried loads such as subducted oceanic slabs [Royden, 1993]. Classical foreland basins form on a plate that is overthrust by an orogenic wedge. Any record of the downward lexure through time (subsidence) is thus transported toward the orogen during convergence. In the distal basin the plate is almost horizontal, thus a given amount of convergence will result in relatively little subsidence. In the proximal basin, the dip of the plate is much steeper (max. 9°), thus the same point on the mobile lower plate will experience gradually increasing subsidence. It is this hinterlandward transport of a point along a curved plate that gives foreland basin subsidence its classical accelerating pattern [e.g., Allen et al., 1986; Vergés et al., 1998; Haddad and Watts, 1999]. 1.1.3 Pro- and retro-wedges and foreland basins Following the development of orogenic wedge models with a rigid backstop came the realisation that the upper plate, represented by the backstop in the experiments, should be deformable, just as the lower plate is. A new category of models was developed, where the substrate of a continuous layer of material is split into two halves (Figure 1.4) [Willett et al., 1993]. One half, representing the lower plate, is dragged towards the central singularity or S-point where the plates meet and the substrate of the lower plate is subducted. he other half, representing the upper plate, remains stationary. his asymmetrical boundary condition results in two wedges joined at the back: he wedge on the lower plate (pro-wedge) grows by frontal accretion and has a low critical taper (the lower of two critical tapers, see above). he wedge on the upper plate (retro- wedge) grows outward by receiving material from the pro-wedge, resulting in a higher critical taper (the highest of two critical tapers, see above). he original description of a doubly vergent orogen with a pro- and retro-wedge was based on numerical models [Willett et al., 1993; Beaumont et al., 1994, 1996], but the same geometry has been observed in analogue models with a similar setup [e.g., Storti et al., 2000; McClay et al., 2004; Hoth et al., 2007, 2008]. Modelling of doubly vergent orogens quickly evolved to account for plate lexure [e.g., Simpson, 2010], sedimentation [Duerto and McClay, 2009], asymmetric erosion [Willett et al., 1993, 2001; Willett, 1999], friction along the base [Beaumont et al., 1994; Naylor et al., 2005], partial subduction of the lower plate [Beaumont et al., 1994, 1996], and viscous rheologies and convergence rates [Rossetti et al., 2002]. Eventually, numerical experiments simply replaced the S-point boundary condition at the base with full modelling of lithospheric plates resting on top of the asthenosphere and imposing convergence at the side boundaries [e.g., Jammes and Huismans, 2012; Erdős et al., 2014, 2015]. Chapter 1. Introduction 13

Figure 1.4. S-point model resulting in a doubly vergent geometry with a pro- and retro-wedge. Ater Willett et al. [1993]. he asymmetry in doubly vergent orogens led to the realisation that the lexural foreland basins also record diferent subsidence histories (Figure 1.5) [Naylor and Sinclair, 2008; Sinclair and Naylor, 2012]. he pro-foreland basin, on the lower plate, records the classical accelerating subsidence pattern as a result of being transported towards the orogen. However, the retro-foreland basin is not transported toward the orogen, because the upper plate remains stationary relative to the plate boundary (S-point). he retro- foreland basin thus only records subsidence as the orogen grows and lexure increases, and retro-foreland subsidence tends to be slow and long-lasting as a result [Naylor and Sinclair, 2008; Sinclair and Naylor, 2012]. If the orogen is in a lux steady state (accretional inlux = erosional outlux), both foreland basins remain the same size. Points in the retro-foreland basin thus do not record subsidence during steady state, and points in the pro-foreland record only subsidence due to convergence toward the orogen. his diference between pro- and retro-foreland basins also implies they preserve diferent parts of the orogenic history. Both basins record the early growth history of the orogen, but the early pro-foreland record is eventually destroyed. During steady state, only the pro-foreland records subsidence. In a suiciently mature orogen, combining the subsidence histories of the pro- and retro-foreland should thus result in a more complete picture of orogenic history [Naylor and Sinclair, 2008].

1.2 Techniques 1.2.1 Cross-section balancing In the analysis of natural foreland fold-and-thrust belts, an important technique is the construction of cross sections. he available data in natural systems is usually incomplete, and thus permits several diferent interpretations that imply diferent amounts of shortening accommodated along the section. For this reason, it is important to ensure that the cross section is geologically valid, and that it is physically possible to achieve the proposed geometry. Balancing of cross sections is a technique where a deformed cross section is restored to its undeformed state using set rules (Figure 1.6), and then corrected for impossible geometries [Dahlstrom, 1969; Hossack, 1979; Woodward et al., 1989]. In constructing a valid balanced cross section, one has to take into account the following considerations: 14 Chapter 1. Introduction

Figure 1.5. Pro- and retro-foreland basins record diferent subsidence histories. Ater Naylor and Sinclair [2008]. 1) Choosing the right orientation of the section. Restoring a 2D cross section assumes plane strain, thus material is transported parallel to the section line. A section that is oblique to the transport direction results in an overestimation of shortening. he transport direction can be determined based on geological evidence. his can be simply perpendicular to the strike of large-scale folds and faults [Woodward et al., 1989]. For numerous small-scale folds, the strike is not a reliable guide for transport direction, but the transport direction can still be estimated using stereonet analysis [Hansen, 1971]. Mineral stretching lineations in metamorphic terranes can also be used as a guide to the (syn-metamorphic) tectonic transport direction [Woodward et al., 1989]. 2) Conserving area between the deformed and restored section [Dahlstrom, 1969]. Since the section is parallel to the direction of transport and the restoration assumes plane strain, the volume of material in the deformed and restored sections should be the same. Only in exceptional cases where out-of-plane transport, compaction, or dissolution play a signiicant role can the volume change between sections [Rowan and Ratlif, 2012]. In the case of compaction and dissolution, the volume of the restored section should always be greater than that of the deformed section. Chapter 1. Introduction 15

3) Conserving bed length [Dahlstrom, 1969]. Competent layers can be folded and faulted, but tend to keep the same length. his serves as an important constraint on the restored cross section as it will determine how long the restored section is going to be. 4) Ensuring that the restored section displays viable dips and geometries for faults and stratigraphy, in line with well-documented analogous structures from a diferent location [Woodward et al., 1989]. his includes eliminating voids in the restored section, which may result from restoring an incorrectly interpreted deformed section. Adhering to geologically sound geometries requires a knowledge of processes and geometries involved in the deformation of strata, such as fault-bend folds [e.g., Rich, 1934; Medwedef and Suppe, 1997], fold-propagation folds and trishear [e.g., Suppe and Medwedef, 1990; Erslev, 1991], and forced folds [e.g., Mitra and Mount, 1998; Miller and Mitra, 2011]. 5) It is good practice to keep shortening within the section to a minimum, and keep the geometry as simple as the data will allow. When constructing multiple sections, consistency between adjacent sections is preferable [Dahlstrom, 1969; Woodward et al., 1989]. Taking into account these guidelines, the process of balancing a cross section is simply iteratively restoring and correcting the deformed section until it is valid [Woodward et al., 1989]. he process of balancing sections was originally developed for use in thin-skinned foreland fold-and-thrust belts, but these principles can also be applied to crustal scale sections that involve basement in the deformation. he balancing technique remains useful and is still used today [e.g., Suppe, 1980; Roeder, 1992; Vergés, 1993; McQuarrie, 2004; Labaume et al., 2016].

Figure 1.6. A balanced and restored cross section through the Zagros mountains. Ater McQuarrie [2004].

1.2.2 Subsidence analysis As described above, foreland basins record subsidence that relates to the evolution of the adjacent orogen. Subsidence analysis is the process that extracts this information from the sedimentary basin. he aim is to relate the accumulation of sediments and vertical movements of the top of the sedimentary column relative to the eustatic sea level, in order to obtain the vertical movements of the basement, and eventually isolate the subsidence signal that is purely due to tectonic forces. Because subsidence analysis involves the thickness of a column of sediments through time, it is important to understand compaction. he thickness of a sedimentary layer decreases as it is buried, and the weight of the overburden literally squeezes it into a thinner layer. his reduction of volume is accommodated by a reduction in pore space, lowering porosity (Figure 1.7a). For normally pressured sediments, the relationship between depth and porosity is: (1.3) where is the porosity at depth, the initial porosity (a material constant), the compaction coeicient (another material constant), and is depth [Allen and Allen, 2005]. In overpressured sediments, the equation becomes: 16 Chapter 1. Introduction

(1.4)

where is the bulk density of the layer (grains + pore ill), the density of the pore luid (water), and the ratio of pore luid pressure relative to lithostatic pressure [Allen and Allen, 2005]. Material constants and are highly lithology-speciic [Sclater and Christie, 1980; Halley and Schmoker, 1983; Sonnenfeld, 1984]. Each lithology (or mix of lithologies) thus has its own porosity-depth curve depending on initial porosity, lithology, pore luid pressure, diagenesis and other factors. Using the porosity-depth relationship, it is possible to calculate the thickness of a given layer at any depth. With the assumption that the thickness change is entirely the result of reducing pore volume during compaction, the thickness of a decompacted layer takes the form: (1.5) with representing thickness of the layer and pore volume. Together with equation 1.3, this gives the general decompaction equation [Allen and Allen, 2005]:

(1.6)

where is the thickness of the decompacted layer (depth of base minus depth of top), is the thickness of the compacted layer, the second-to-last term represents the pore volume of the compacted layer, and the inal term represents the pore volume of the decompacted layer. here is no exact solution for this equation. Instead, it is solved by iteratively adjusting the values and until both sides of the equation are equal.

Figure 1.7. a) he generic porosity-depth relationship. b) Schematic of decompaction. Ater Allen and Allen [2005].

Using the general decompaction equation, a decompaction curve can be constructed. his is done by removing the top layer and recalculating the decompacted thickness of all layers underneath, then repeating the process until there is no sediment let on top of basement (Figure 1.7b). When correcting for paleobathymetry and eustatic sea level variations, the thickness of the sedimentary column (i.e. depth to basement) can be plotted against time to give total subsidence [Allen and Allen, 2005]. Total subsidence represents the ‘true’ vertical movements of basement through time, under the inluence of tectonic forces and isostatic adjustments to the sediment mass, and changes in the mass of the water Chapter 1. Introduction 17 column due to sea level variations. To isolate only the tectonic signal, it is necessary to correct for the isostatic response to sediments, paleobathymetry and eustatic sea level, a process called backstripping [Steckler and Watts, 1978; Allen and Allen, 2005]:

(1.7) where is depth to basement relative to present-day sea level, is the decompacted sediment column, is the density of the mantle, is the paleobathymetry, and the paleo-sea level relative to present-day. he backstripping equation assumes Airy isostasy to compensate for the load of the sediment column and part of the water column. he result is tectonic subsidence, which represents the depth of basement if there had been no sediment, and sea level would have been the same as today (Figure 1.8). It is a measure of the tectonic forcing in a basin’s evolution [Steckler and Watts, 1978]. Tectonic subsidence is a valuable tool in the analysis of foreland basins [e.g., Desegaulx and Brunet, 1990; Vergés et al., 1998]. It is equally useful in other tectonic settings, as long as there is accumulation of sediment relative to a known datum (i.e. sea level) [e.g., Xie and Heller, 2009].

Figure 1.8. Total versus tectonic subsidence. Ater Steckler and Watts [1978].

1.2.3 Numerical mechanical modelling Numerical modelling of deformation has proven to be a very efective tool in geoscience. With the availability of ever increasing computational power, numerical models show a clear trend of increasing complexity and resolution. he earliest numerical investigations applied purely thermal models to mantle convection in subduction zones [e.g., Richter and McKenzie, 1978; Schubert, 1992]. Mechanical models of continental collision followed [e.g., Bird, 1978; Daignières et al., 1978]. Since then, model codes have become steadily more capable, including thermo-mechanically coupled deformation, higher resolutions, 3D capable codes, coupled surface processes, etc. [e.g., Fullsack, 1995; Braun et al., 2008; hieulot, 2011]. 18 Chapter 1. Introduction hree main categories of numerical implementation have been developed: inite element (Figure 1.9a) [e.g., Fullsack, 1995; Braun et al., 2008], inite diference [e.g., Waltham, 1989; Waltham and Hardy, 1995; Gerya, 2009], and discrete element (Figure 1.9b) [e.g., Finch et al., 2004; Naylor et al., 2005]. Finite element and inite diference models are both implementations of deformation in a continuum, whereas discrete element models consist of discrete, rigid particles that are loosely packed together, analogous to sand.

Figure 1.9. a) Example of a inite element model. Ater Jammes and Huismans [2012]. b) Example of a discrete element model. Ater Naylor et al. [2005].

Mechanical models that are based on continuum mechanics model deformation by solving the Stokes equation. When assuming incompressibility and negligible inertial forces, it can be written as [e.g., Currie et al., 2008]:

(1.8) where is the stress tensor, the spatial dimensions, is density and the gravitational acceleration. Materials in the model deform by lowing as a viscous luid, following a power law rheology that relates the strain rate to the creep stress to the power :

(1.9) where is the strain rate tensor, is a pre-exponential material constant, is the creep stress, the exponent, is the activation energy, is pressure, is the activation volume, is the gas constant and is temperature. Material constants , , , and are determined from laboratory data [Karato and Wu, 1993; Gleason and Tullis, 1995]. Extrapolating the mechanical properties of rocks from laboratory experiments to geological timescales and rates may introduce errors in the mechanical model. Another limitation of numerical mechanical models using continuum mechanics is that they cannot model ‘true’ frictional (brittle) deformation. A method to simulate frictional deformation is to calculate the theoretical yield stress and adjust the viscosity to ensure the creep stress does not exceed the yield stress [Fullsack, 1995; hieulot, 2011]. For a more in-depth explanation of the governing equations of the numerical models used in this work, see appendix B. Mechanical models (numerical and analogue) can be used to test the inluence of various factors on a certain deformational process. An advantage of numerical models over analogue models is that all experiments Chapter 1. Introduction 19 can be performed at full scale with geological deformation rates. Numerical models do not have to rely on suitably scaled analogue materials. In addition, the control over factors such as sedimentation and erosion is more reined in numerical models. However, because numerical models can simulate processes at the spatial and temporal scales of geological processes, material properties obtained from laboratory experiments must be extrapolated, introducing some uncertainty. his demonstrates the importance of benchmarking a numerical model against known analogue and numerical experiments. Numerical deformation models have been applied to a wide variety of problems: subduction zones [e.g., Schellart et al., 2007; Currie et al., 2008], continental collision [e.g., Willett et al., 1993; Pifner et al., 2000; Sinclair et al., 2005; Jammes and Huismans, 2012; Erdős et al., 2014; Ueda et al., 2015], riting [e.g., Lesne et al., 2000; Huismans and Beaumont, 2007], salt décollements [e.g., Ruh et al., 2012; Ghazian and Buiter, 2014], and many more. In this work, we use a modiied version of the numerical code FANTOM, a high-resolution arbitrary Eulerian-Lagrangian inite-element code for modelling thermo-mechanically coupled deformation [hieulot, 2011]. his code is described in more detail in Chapters 3, 4 and Appendix B.

1.3 Pyrenees 1.3.1 Present-day structure

Figure 1.10. Crustal structure of the Central Pyrenees based on the ECORS Pyrenees deep seismic section. Ater Muñoz [1992]. he Pyrenees form a WNW-ESE striking orogen at the boundary between Spain and France that was created by Late Cretaceous to Oligocene collision of the Iberian and European plates. he orogen can be divided into the following tectonostratigraphic zones from north to south (Figure 1.10): 1) he Aquitaine retro-foreland basin, a mildly deformed sedimentary basin. In the east, this basin comprises successive sub-basins of Late Cretaceous to Eocene age that overlie Paleozoic basement [Plaziat, 1981; Tambareau et al., 1995; Martin-Martin et al., 2001; Christophoul et al., 2003]. In general, younger sub- basins are situated to the north of older sub-basins. In the west, syn-orogenic sediments overlie Triassic salt to Lower Cretaceous rit sediments [Puigdefàbregas and Souquet, 1986; Desegaulx and Brunet, 1990]. 2) he Sub-Pyrenean Zone, a more deformed (folded and faulted) part of the foreland basin immediately north of the North Pyrenean Frontal hrust, usually limited to the north by a blind thrust [Bilotte et al., 1988; Deramond et al., 1993]. 3) he North Pyrenean Zone, a narrow belt between the North Pyrenean Frontal hrust to the north and the North Pyrenean Fault to the south. he intensity of deformation in this zone is signiicantly higher than in the zones further north, and mainly north-vergent [e.g., Choukroune, 1974]. his zone comprises inverted Early Cretaceous rit basins [Souquet et al., 1985; Debroas, 1990], allochthonous 20 Chapter 1. Introduction basement massifs [Souquet et al., 1977; Souquet and Peybernès, 1987; Baby, 1988; Baby et al., 1988; de Saint Blanquat et al., 1990], and the Metamorphic Internal Zone immediately north of the North Pyrenean Fault. he Metamorphic Internal Zone mostly consists of Mesozoic sediments that have undergone high temperature, low pressure metamorphism and are highly deformed and brecciated [Golberg and Leyreloup, 1990; Debroas et al., 2010a, 2010b; de Saint Blanquat et al., 2016]. Blocks of mantle peridotites are embedded within the metamorphosed sediments and breccias (Figure 1.11) [Lagabrielle et al., 2010]. hese ultrabasic mantle rocks are named lherzolites, ater Lherz, a locality south of the Trois Seigneurs Massif, where they were irst described [e.g., Bonney, 1877]. Due in part to the presence of lherzolites, the North Pyrenean Fault that forms the southern border of this zone is generally regarded as the suture between the Iberian and European plates. he North Pyrenean Fault is a major vertical fault, whether it accommodated a large strike-slip displacement or not was diicult to constrain, and thus debated [e.g., Casteras, 1933; Choukroune, 1974; Souquet et al., 1977; Choukroune and Mattauer, 1978]. Due to the lack of plate kinematic consensus, part of this debate is ongoing [e.g., Olivet, 1996; Jammes et al., 2009; Vissers and Meijer, 2012b].

Figure 1.11. Location of lherzolite outcrops in the Pyrenees. Ater Lagabrielle et al. [2010].

4) he Axial Zone comprises Paleozoic rocks deformed mainly by the Variscan orogeny. he ECORS Pyrenees deep seismic relection proile across the Central Pyrenees constrains the deep structure [Choukroune and ECORS-team, 1989; Roure et al., 1989; Muñoz, 1992; Roure and Choukroune, 1998; Beaumont et al., 2000]. he Iberian lower crust subducts below the European plate, and Iberian upper crustal thrust sheets form a south-vergent antiformal stack [Muñoz, 1992]. To the west and to the east, the stack of Iberian thrust sheets is interpreted to form a less well developed antiformal stack [Vergés et al., 1995; Teixell, 1998]. 5) he South Pyrenean Zone, a south verging thin-skinned fold-and-thrust belt of variable width that is delimited by the Vallfogona thrust in the east, the Sierras Marginales in the Central Pyrenees, and the external Sierras in the west. he thrust sheets in the South Pyrenean Zone were thrust southward along an evaporite décollement. For the upper, older thrust sheets this décollement consists of Triassic Keuper evaporites, while for the lower, younger structures this décollement consists of two Eocene evaporite layers (Beuda Fm. and Cardona Fm.) [Puigdefàbregas et al., 1992; Vergés et al., 1992; Vergés, 1993]. Chapter 1. Introduction 21

6) he Ebro pro-foreland basin. his basin consists of Eocene and Oligocene stratigraphic units that have been mildly deformed by detachment folding above the upper Eocene Cardona evaporite formation [Vergés et al., 1992; Vergés, 1993]. he Ebro foreland basin is bordered to the south by the north-vergent Catalan Coastal Ranges. 1.3.2 Hyperextended rit

0 20 W 0 A A33(80 Ma) Olivet (1996) SO N

1 ) / 1

40N 1 40 N

M0(118 Ma) Ol.ivet (1 996) 0 20 W 0

SON

1 ) / 1

40N 1 40 N

0

SON -SON

MO (1 18 Ma) Jammes et al, (2009) Jammes et al. (2009) 40 N 40 N 80Ma

Figure 1.12. hree competing plate kinematic reconstructions of the Pyrenean domain. Ater Mouthereau et al. [2014].

Extension and subsidence between the Iberian and European plates is recorded since Triassic times as the Iberian plate lay between the opening North Atlantic and Tethys oceans [Biteau et al., 2006]. he principal rit phase occurred in the Early Cretaceous. he plate kinematics of this phase remain diicult to constrain, resulting in several competing models (Figure 1.12) [Olivet, 1996; Sibuet et al., 2004; Jammes et al., 2009]. hese models predict transtension, strike-slip or orthogonal extension, broadly categorised into 22 Chapter 1. Introduction

three competing models: 1) A strike slip model that implies several hundred kilometres of sinistral strike- slip accommodated along the North Pyrenean Fault [Olivet, 1996]. 2) A scissor type model that implies opening of the Bay of Biscay in the west contemporaneous with oceanic subduction in the east [Srivastava et al., 1990; Roest and Srivastava, 1991; Sibuet et al., 2004; Vissers and Meijer, 2012a, 2012b], and a hybrid model that combines an anticlockwise rotation and eastward movement of Iberia with transtension between Iberia and Europe, followed by orthogonal extension during the Aptian and orthogonal convergence from the Late Cretaceous onwards [Jammes et al., 2009].

Figure 1.13. Schematic cross section of the Pyrenean rit that exhumes mantle. Ater Lagabrielle et al. [2010].

Regardless of the plate kinematics, the presence of lherzolites in the Metamorphic Internal Zone shows that mantle was exhumed during the pre-orogenic rit phase. he metamorphosed Mesozoic sediments are assumed to have covered the exhumed mantle during this rit phase (Figure 1.13). his is supported by roughly coeval ages for riting and metamorphism [Golberg and Leyreloup, 1990; Clerc, 2012] and syn-rit breccias with Lherzolite clasts [Jammes et al., 2009; Lagabrielle et al., 2010; de Saint Blanquat et al., 2016]. Triassic Keuper salt may also have provided a décollement for the Mesozoic cover to slide into the rit [Jammes et al., 2010; Teixell et al., 2016; Saura et al., 2016]. 1.3.3 Convergent evolution he onset of convergence is recorded by lexural foreland basins and syn-depositional thrusts in both forelands starting around 85-80 Ma [Desegaulx et al., 1990; Deramond et al., 1993]. he onset of convergence is usually placed at late Santonian to early Campanian (~83 Ma) [Puigdefàbregas and Souquet, 1986; Mouthereau et al., 2014]. he Metamorphic Internal Zone was emplaced northward (and likely also southward) during closure of the exhumed mantle domain [Jammes et al., 2009; Lagabrielle et al., 2010]. Marine lexural basins formed on both the Iberian and European margins. In the northern Pyrenees, paleocurrent data and the westward prograding geometry of Upper Cretaceous lysch indicates that sediments were supplied from the east (Figure 1.14) [Bilotte, 1985]. In the southern Pyrenees, sediments were supplied from the south and east [Rosell et al., 2001]. Paleocene sedimentation was continental throughout the Pyrenees, and tectonic activity appears to have been slow or nonexistent [Plaziat, 1981; Desegaulx and Brunet, 1990; Vergés et al., 2002]. his phase of slow shortening may be due to a temporary halt in plate convergence, an idea based on plate kinematic reconstructions [Rosenbaum et al., 2002], but this has been contested by other authors [Vissers and Meijer, 2012a]. Following a marine transgression Chapter 1. Introduction 23

Figure 1.14. Lateral stratigraphic section through the northeastern Pyrenees showing the westward prograding geometry of Upper Cretaceous Flysch. Ater Ricateau and Villemin [1973] and Bilotte [1985]. from the west in the Ypresian, the south Pyrenean foreland developed as underilled and marine from 55 to 37 Ma [Puigdefàbregas and Souquet, 1986; Puigdefàbregas et al., 1992; Vergés et al., 1995, 1998]. he transition to continental conditions in the southern Pyrenees occurred because the Ebro foreland basin was isolated from the Atlantic Ocean ater uplit of the Western Pyrenees [Costa et al., 2010]. In the northeastern Pyrenees, the Ypresian marine transgression was followed by a westward regression and deposition of continental sandstones and conglomerates [Martin-Martin et al., 2001; Christophoul et al., 2003]. Syn-depositional deformation was recorded in both the north and south Pyrenean forelands [Ford et al., 1997; Martin-Martin et al., 2001; Ramos et al., 2002; Christophoul et al., 2003; Labaume et al., 2016]. Uplit and exhumation of the Axial Zone starting around 50 Ma is recorded by both in-situ and detrital thermochronology data [Fitzgerald et al., 1999; Sinclair et al., 2005; Metcalf et al., 2009; Rahl et al., 2011; Whitchurch et al., 2011]. he peak of exhumation of the Axial Zone appears to have occurred in the Early Oligocene (~36 Ma), simultaneous with a large inlux of sediment into the south Pyrenean foreland [Fitzgerald et al., 1999; Sinclair et al., 2005; Fillon et al., 2013a]. Exhumation migrated both southward and westward in the Pyrenees [Sinclair et al., 2005; Jolivet et al., 2007; Metcalf et al., 2009]. he end of deformation in the Pyrenees is placed in the Late Oligocene (~24.7 Ma) by magnetostratigraphic dating of growth strata along the south Pyrenean thrust front [Meigs, 1997]. Following the end of convergence, post-orogenic exhumation of the Ebro foreland basin remained relatively minor until 9 Ma, most likely because of a drop in base level ater a connection to the Atlantic Ocean was re-established [Fillon and van der Beek, 2012]. 1.3.4 Shortening estimates Estimates of the amount of shortening in the Pyrenees vary. Before the acquisition of deep seismic data, minimum shortening was estimated between 55 km in the west to 80 km in the east [Seguret and Daignieres, 1986]. Following the irst results of the ECORS deep seismic proile, shortening was estimated at a minimum, but poorly constrained, 120 km [Roure et al., 1989]. Later, better constrained balanced sections along the ECORS line placed shortening at 147 km and 165 km, using upper crustal imbricates in the Axial Zone and subducting Iberian lower crust [Muñoz, 1992; Beaumont et al., 2000]. he latter estimate is partitioned between 128 km for the southern wedge and 37 km for the northern one [Beaumont et al., 2000]. More recently, a much lower estimate of 92 km of shortening (excluding closure of the exhumed 24 Chapter 1. Introduction

mantle domain) was obtained, resulting from full crustal imbrication and a revised shortening estimate of the southern Pyrenees [Mouthereau et al., 2014]. Shortening in the Eastern Pyrenees is estimated at 125 km [Vergés et al., 1995]. his is partitioned as 32 km in the northern Pyrenees [Baby et al., 1988], ~23 km of internal shortening of the Axial Zone, and 70 km in the south Pyrenean foreland [Vergés, 1993]. In the Western Pyrenees, shortening is estimated to be lower, at 75 to 80 km [Teixell, 1998], and more recently 115 km including closure of a 15 km wide exhumed mantle zone [Teixell et al., 2016].

7 a Western Pyrenees b Central Pyrenees (ECORS)

6

5 Teixell et al. (2016) Beaumont et al. (2000)

4 mantle protocollision closure 3

shortening rate (mm/y) full collision 2

1 Mouthereau et al. (2014) mantle closure

0 Phase 1 Ph2 Phase 3 Ph4 Phase 1 Ph2 Phase 3 Ph4

90 80 70 60 50 40 30 20 90 80 70 60 50 40 30 20 Time (Ma) Time (Ma)

Figure 1.15. Estimates of shortening rate through time across the Pyrenees.

he evolution of shortening rates in the Pyrenees remains debated. Several diferent shortening histories have been proposed, some of them contradicting others (Figure 1.15). In the Central Pyrenees, Beaumont et al., [2000] show an overall acceleration of the shortening rate, in particular at 36 Ma. his is based around a peak in exhumation rates of the Axial Zone around that time [e.g., Fitzgerald et al., 1999; Metcalf et al., 2009]. In contrast, Mouthereau et al., [2014] propose a deceleration of shortening, starting at ~3 mm/y in the Late Cretaceous and Paleocene and gradually decelerating until the Miocene. In the Western Pyrenees, shortening rates are estimated to have peaked in the Eocene at 3-2.5 mm/y [Teixell et al., 2016]. In the Eastern Pyrenees, a shortening rate history exists only for the southern foreland. his also places the peak of shortening during the Eocene, at ~4.5 mm/y [Vergés et al., 1995].

1.4 Limitations and scientiic questions Critical taper theory is useful to describe the average shape and deformation of an orogenic wedge. However, it does not describe the episodic nature of deformation along discrete fault planes as observed in modelled thrust wedges [Hoth et al., 2007; Simpson, 2011; Buiter, 2012]. In doubly vergent models, it has been shown that pro- and retro-wedges and pro- and retro-foreland basins behave diferently [Willett et al., 1993; Naylor and Sinclair, 2008]. Analogue models have hinted at a possible mechanical coupling between the pro- and retro-wedge, where shortening alternates between them [Hoth et al., 2007]. his behaviour cannot be explained with critical taper theory alone, and is currently poorly understood. he relationship between the pro- and retro-wedge, whether that relationship changes through time, and what would cause such a change remains poorly constrained in natural systems. Detailed comparisons of the two wedges and their foreland basins in terms of structure and evolution are required to investigate these issues. Chapter 1. Introduction 25

From these limitations of our understanding of doubly vergent orogens come the following main scientiic questions that are investigated in this thesis.

• How does a doubly vergent orogen evolve? • Do the pro- and retro-wedge interact? • What factors inluence shortening distribution, and how? Associated with these questions are secondary questions: What causes the Paleocene quiescence in the Pyrenees? Can thin-skinned deformation inluence the crustal structure of an orogen? In this work, a dual approach is used that combines detailed study of the Eastern Pyrenees and numerical modelling of controlling factors to constrain the evolution of doubly vergent orogens. Both methods were used in parallel, so that insights from one may beneit the other. he results of the East Pyrenean case study are described in Chapter 2 (in revision ater review at Tectonics). he results of numerical modelling are described in Chapters 3 and 4 (destined for submission to Geology and Journal of Geophysical Research: Solid Earth, respectively). Based on the results of these studies, a generic evolutionary model for inverted rit systems is developed and compared against several natural doubly vergent orogens, also in Chapter 4.

References Abouin, J. (1965), Geosynclines, Elsevier B.V., Amsterdam. Allen, P. A., and J. R. Allen (2005), Basin Analysis, Second Edi., Blackwell Publishing Ltd., Oxford, UK. Allen, P. A., P. Homewood, and G. D. Williams (1986), Foreland Basins: An Introduction, International Association of Sedimentologists, Oxford. Baby, P. (1988), Chevauchements dans une zone à structure complexe: La zone nord-pyrénéenne ariégeoise, Université Paul Sabatier Toulouse III, Toulouse, France. Baby, P., G. Crouzet, M. Specht, J. Deramond, M. Bilotte, and E.-J. Debroas (1988), Rôle des paléostructures albo- cénomaniennes dans la géométrie des chevauchements fronaux nord-pyrénéens, C. R. Acad. Sci. Paris, 306(2), 307–313. Beaumont, C. (1981), Foreland Basins, Geophys. J. R. Astron. Soc., 65, 291–329. Beaumont, C., P. Fullsack, and J. Hamilton (1994), Styles of crustal deformation in compressional orogens caused by subduction of the underlying lithosphere, Tectonophysics, 232(1–4), 119–132, doi:10.1016/0040- 1951(94)90079-5. Beaumont, C., S. Ellis, J. Hamilton, and P. Fullsack (1996), Mechanical model for subduction-collision tectonics of Alpine-type compressional orogens, Geology, 24(8), 675–678, doi:10.1130/0091- 7613(1996)024<0675:MMFSCT>2.3.CO. Beaumont, C., J. A. Muñoz, J. Hamilton, and P. Fullsack (2000), Factors controlling the Alpine evolution of the central Pyrenees inferred from a comparison of observations and geodynamical models, J. Geophys. Res. Solid Earth, 105(B4), 8121–8145, doi:10.1029/1999JB900390. Bilotte, M. (1985), Le Crétacé supérieur des plates-formes est-pyrénéennes, Université Paul-Sabatier, Toulouse, France. Bilotte, M., J. Cosson, B. Crochet, B. Peybernès, J. Roche, F. Taillefer, Y. Tambareau, Y. Ternet, and J. Villatte (1988), Notice explicative de la feuille Lavelanet à 1/50 000 (Feuille №1076), BRGM, Orléans, France. Bird, P. (1978), Finite element modeling of lithosphere deformation: the Zagros collision orogeny, Tectonophysics, 50, 307–336. Biteau, J.-J., a. Le Marrec, M. Le Vot, and J.-M. Masset (2006), he Aquitaine Basin, Pet. Geosci., 12(3), 247–273, doi:10.1144/1354-079305-674. Bonney, T. G. (1877), he Lherzolite of Ariège, Geol. Mag., IV(2). Braun, J., C. hieulot, P. Fullsack, M. DeKool, C. Beaumont, and R. S. Huismans (2008), DOUAR: A new three- dimensional creeping low numerical model for the solution of geological problems, Phys. Earth Planet. Inter., 171(1–4), 76–91, doi:10.1016/j.pepi.2008.05.003. Buiter, S. J. H. (2012), A review of brittle compressional wedge models, Tectonophysics, 530–531(40), 1–17, doi:10.1016/j.tecto.2011.12.018. Casteras, M. (1933), Recherches sur la structure du versant nord des Pyrénées centrales et orientales, Bull. Serv. Cart. géol. Fr., XXXVII, 515. 26 Chapter 1. Introduction

Choukroune, P. (1974), Structure et evolution tectonique de la zone nord-Pyrénéenne. Analyse de la déformation dans une portion de chaine a schistosité sub-verticale., Université des Sciences et Techniques du Languedoc, Montpellier, France. Choukroune, P., and ECORS-team (1989), he Ecors Pyrenean deep seismic proile relection data and the overall structure of an orogenic belt, Tectonics, 8, 23, doi:10.1029/TC008i001p00023. Choukroune, P., and M. Mattauer (1978), Tectonique des plaques et Pyrénées: sur le fonctionnement de la faille transformante nord-Pyrénéenne; comparaisons avec des modèles actuels, Bull. la Soc. Geol. Fr., 7 t. XX(5), 689–700. Christophoul, F., J.-C. Soula, S. Brusset, B. Elibana, M. Roddaz, G. Bessiere, and J. Deramond (2003), Time, place and mode of propagation of foreland basin systems as recorded by the sedimentary ill: examples of the Late Cretaceous and Eocene retro-foreland basins of the north-eastern Pyrenees, in Tracing Tectonic Deformation Using the Sedimentary Record. Geological Society, London, Special Publications, vol. 208, edited by T. McCann and A. Saintot, pp. 229–252, Geological Society of London, London, United Kingdom. Clerc, C. (2012), Evolution du domaine nord-Pyrénéen au Cretace. Amincissement crustal extreme et thermicité élévee: un analogue pour les marges passives, Université Pierre et Marie Curie. Costa, E., M. Garcés, M. López-Blanco, E. Beamud, M. Gómez-Paccard, and J. C. Larrasoaña (2010), Closing and continentalization of the South Pyrenean foreland basin (NE Spain): magnetochronological constraints, Basin Res., 22(6), 904–917, doi:10.1111/j.1365-2117.2009.00452.x. Currie, C. a., R. S. Huismans, and C. Beaumont (2008), hinning of continental backarc lithosphere by low- induced gravitational instability, Earth Planet. Sci. Lett., 269, 435–446, doi:10.1016/j.epsl.2008.02.037. Currie, C. A., C. Beaumont, and R. S. Huismans (2007), he fate of subducted sediments: A case for backarc intrusion and underplating, Geology, 35(12), 1111–1114, doi:10.1130/G24098A.1. Dahlen, F. a. (1984), Noncohesive critical Coulomb wedges: An exact solution, J. Geophys. Res., 89(B12), 10125, doi:10.1029/JB089iB12p10125. Dahlen, F. a. (1990), Critical Taper Model of Fold-And-hrust Belts and Accretionary Wedges, Annu. Rev. Earth Planet. Sci., 18(1), 55–99, doi:10.1146/annurev.ea.18.050190.000415. Dahlstrom, C. D. A. (1969), Balanced cross sections, Can. J. Earth Sci., 6(4), 743–757. Daignières, M., M. Fremon, and A. Friaa (1978), Modèle de type Norton-Hof généralisé pour l’étude des déformation lithosphériques, Comptes Rendus l’Académie des Sci., Série B(18), 371–74. Davis, D., J. Suppe, and F. a. Dahlen (1983), Mechanics of fold-and- thrust belts and accretionary wedges., J. Geophys. Res., doi:10.1029/JB088iB02p01153. Davis, D. M., and T. Engelder (1985), he role of salt in fold-and-thrust belts, Tectonophysics, 119(1–4), 67–88, doi:10.1016/0040-1951(85)90033-2. Debroas, E.-J. (1990), Le lysch noir albo-cénomanien témoin de la structuration albienne à sénonienne de la Zone nord-pyrénéenne en Bigorre (Hautes-Pyrénées, France), Bull. la Soc. Geol. Fr., VI(2), 273–285, doi:10.2113/gssgbull.VI.2.273. Debroas, E.-J., J. Canérot, and M. Bilotte (2010a), Les brèches d’urdach, témoins de l’exhumation du manteau pyrénéen dans un escarpement de faille vraconnien-cénomanien inférieur (zone nord-pyrénéenne, pyrénées- atlantiques, France), Geol. la Fr., (2), 53–65. Debroas, E.-J., M. Bilotte, J. Canérot, and J. G. Astruc (2010b), Réinterprétation des brèches de la Faille nord- pyrénéenne ariégeoise ( France ), Bull. la Sociétée d’Histoire Nat., 146(1), 77–88. Deramond, J., P. Souquet, M.-J. Fondecave-Wallez, and M. Specht (1993), Relationships between thrust tectonics and sequence stratigraphy surfaces in foredeeps: model and examples from the Pyrenees (Cretaceous-Eocene, France, Spain), Geol. Soc. London, Spec. Publ., 71(1), 193–219, doi:10.1144/GSL. SP.1993.071.01.09. Desegaulx, P., and M.-F. Brunet (1990), Tectonic subsidence of the Aquitaine Basin since Cretaceous times, Bull. la Soc. Geol. Fr., VI(2), 295–306, doi:10.2113/gssgbull.VI.2.295. Desegaulx, P., F. Roure, and A. Villien (1990), Structural evolution of the Pyrenees: tectonic inheritance and lexural behaviour in the continental crust, Tectonophysics, 182(3–4), 211–225, doi:10.1016/0040- 1951(90)90164-4. Duerto, L., and K. R. McClay (2009), he role of syntectonic sedimentation in the evolution of doubly vergent thrust wedges and foreland folds, Mar. Pet. Geol., 26(7), 1051–1069, doi:10.1016/j. marpetgeo.2008.07.004. Chapter 1. Introduction 27

Erdős, Z., R. S. Huismans, P. van der Beek, and C. hieulot (2014), Extensional inheritance and surface processes as controlling factors of mountain belt structure, J. Geophys. Res. Solid Earth, 119, 9042–9061, doi:10.1002/2014JB011408. Erdős, Z., R. S. Huismans, and P. van der Beek (2015), First-order control of syntectonic sedimentation on crustal- scale structure of mountain belts, J. Geophys. Res. Solid Earth, 120, 1–16, doi:10.1002/2014JB011785. Erslev, E. A. (1991), Trishear fault-propagation folding, Geology, 19(6), 617–620, doi:10.1130/0091- 7613(1991)019<0617:TFPF>2.3.CO. Fillon, C., and P. van der Beek (2012), Post-orogenic evolution of the southern Pyrenees: Constraints from inverse thermo-kinematic modelling of low-temperature thermochronology data, Basin Res., 24(4), 418–436, doi:10.1111/j.1365-2117.2011.00533.x. Fillon, C., C. Gautheron, and P. van der Beek (2013a), Oligocene-Miocene burial and exhumation of the Southern Pyrenean foreland quantiied by low-temperature thermochronology, J. Geol. Soc. London., 170(Gallagher 2012), 67–77, doi:10.1144/jgs2012-051. Fillon, C., R. S. Huismans, P. van der Beek, and J. A. Muñoz (2013b), Syntectonic sedimentation controls on the evolution of the southern Pyrenean fold-and-thrust belt: Inferences from coupled tectonic-surface processes models, J. Geophys. Res. Solid Earth, 118(10), 5665–5680, doi:10.1002/jgrb.50368. Finch, E., S. Hardy, and R. Gawthorpe (2004), Discrete-element modelling of extensional fault-propagation folding above rigid basement fault blocks, Basin Res., 16(4), 489–506, doi:10.1111/j.1365-2117.2004.00241.x. Fitzgerald, P. G., J. A. Muñoz, P. J. Coney, and S. L. Baldwin (1999), Asymmetric exhumation across the Pyrenean orogen: Implications for the tectonic evolution of a collisional orogen, Earth Planet. Sci. Lett., 173, 157–170, doi:10.1016/S0012-821X(99)00225-3. Ford, M. (2004), Depositional wedge tops: Interaction between low basal friction external orogenic wedges and lexural foreland basins, Basin Res., 16(3), 361–375, doi:10.1111/j.1365-2117.2004.00236.x. Ford, M., E. A. Williams, A. Artoni, J. Vergés, and S. Hardy (1997), Progressive evolution of a fault-related fold pair from growth strata geometries, Sant Llorenç de Morunys, SE Pyrenees, J. Struct. Geol., 19(3–4), 413–441, doi:10.1016/S0191-8141(96)00116-2. Fullsack, P. (1995), An arbitrary Lagrangian-Eulerian formulation for creeping lows and its application in tectonic models, Geophys. J. Int., 120, 1–23, doi:10.1111/j.1365-246X.1995.tb05908.x. Gerya, T. (2009), Introduction to Numerical Geodynamic Modelling, Cambridge University Press, Cambridge, United Kingdom. Ghazian, R. K., and S. J. H. Buiter (2014), Numerical modelling of the role of salt in continental collision: An application to the southeast Zagros fold-and-thrust belt, Tectonophysics, 632, 96–110, doi:10.1016/j. tecto.2014.06.006. Gleason, G. C., and J. Tullis (1995), A low law for dislocation creep of quartz aggregates determined with the molten salt cell, Tectonophysics, doi:10.1016/0040-1951(95)00011-B. Golberg, J. M., and A. F. Leyreloup (1990), High temperature-low pressure Cretaceous metamorphism related to crustal thinning (Eastern North Pyrenean Zone, France), Contrib. to Mineral. Petrol., 104(2), 194–207, doi:10.1007/BF00306443. Haddad, D., and A. B. Watts (1999), Subsidence history, gravity anomalies, and lexure of the northeast Australian margin in Papua New Guinea, Tectonics, 18(5), 827–842, doi:10.1029/1999TC900009. Halley, R. B., and J. W. Schmoker (1983), High-porosity Cenozoic carbonate rocks of south Florida: progressive loss of porosity with depth., Am. Assoc. Pet. Geol. Bull., 67(2), 191–200, doi:10.1306/03B5ACE6-16D1- 11D7-8645000102C1865D. Hansen, E. (1971), Methods of Deducing Slip-Line Orientations from the Geometry of Folds, in Strain Facies, pp. 27–51, Springer Berlin Heidelberg, Berlin, Heidelberg. Hossack, J. R. (1979), he use of balanced cross-sections in the calculation of orogenic contraction: A review, J. Geol. Soc. London., 136(6), 705–711, doi:10.1144/gsjgs.136.6.0705. Hoth, S., A. Hofmann-Rothe, and N. Kukowski (2007), Frontal accretion: An internal clock for bivergent wedge deformation and surface uplit, J. Geophys. Res. Solid Earth, 112(January), B06408, doi:10.1029/2006JB004357. Hoth, S., N. Kukowski, and O. Oncken (2008), Distant efects in bivergent orogenic belts — How retro-wedge erosion triggers resource formation in pro-foreland basins, Earth Planet. Sci. Lett., 273(1–2), 28–37, doi:10.1016/j.epsl.2008.05.033. Huismans, R. S., and C. Beaumont (2007), Roles of lithospheric strain sotening and heterogeneity in determining the geometry of rits and continental margins, Geol. Soc. London, Spec. Publ., doi:10.1144/SP282.6. 28 Chapter 1. Introduction

Jammes, S., and R. S. Huismans (2012), Structural styles of mountain building: Controls of lithospheric rheologic stratiication and extensional inheritance, J. Geophys. Res. Solid Earth, 117(B10), B10403, doi:10.1029/2012JB009376. Jammes, S., G. Manatschal, L. L. Lavier, and E. Masini (2009), Tectonosedimentary evolution related to extreme crustal thinning ahead of a propagating ocean: Example of the western Pyrenees, Tectonics, 28(4), TC4012, doi:10.1029/2008TC002406. Jammes, S., G. Manatschal, and L. L. Lavier (2010), Interaction between prerit salt and detachment faulting in hyperextended rit systems: he example of the Parentis and Mauléon basins (Bay of Biscay and western Pyrenees), Am. Assoc. Pet. Geol. Bull., 94(7), 957–975, doi:10.1306/12090909116. Jolivet, M., P. Labaume, P. Monié, M. Brunel, N. Arnaud, and M. Campani (2007), hermochronology constraints for the propagation sequence of the south Pyrenean basement thrust system (France-Spain), Tectonics, 26, TC5007, doi:10.1029/2006TC002080. Karato, S. -i., and P. Wu (1993), Rheology of the Upper Mantle: A Synthesis, Science (80-. )., 260(5109), 771–778, doi:10.1126/science.260.5109.771. Konstantinovskaya, E. A., and J. Malavieille (2005), Accretionary orogens: Erosion and exhumation, Geotectonics, 39(1), 69–86. Labaume, P., F. Meresse, M. Jolivet, A. Teixell, and A. Lahid (2016), Tectono-thermal history of an exhumed thrust-sheet-top basin: an example from the south Pyrenean thrust belt, Tectonics, 1–34, doi:10.1002/2016TC004192. Lagabrielle, Y., P. Labaume, and M. de Saint Blanquat (2010), Mantle exhumation, crustal denudation, and gravity tectonics during Cretaceous riting in the Pyrenean realm (SW Europe): Insights from the geological setting of the lherzolite bodies, Tectonics, 29(4), TC4012, doi:10.1029/2009TC002588. Lesne, O., E. Calais, J. Deverchère, J. Chéry, and R. Hassani (2000), Dynamics of intracontinental extension in the north Baikal rit from two-dimensional numerical deformation modeling, J. Geophys. Res., 105(B9), 21727–21744, doi:10.1029/2000JB900139. Martin-Martin, M., J. Rey, F. J. Alcala-Garcia, J. Tosquella, J. Deramond, E. Lara-Corona, F. Duranthon, and P. O. Antoine (2001), Tectonic controls on the deposits of a foreland basin: an example from the Eocene Corbières-Minervois basin, France, Basin Res., 13, 419–433, doi:doi:10.1046/j.0950-091x.2001.00158.x. McClay, K. R., P. S. Whitehouse, T. Dooley, and M. Richards (2004), 3D evolution of fold and thrust belts formed by oblique convergence, Mar. Pet. Geol., 21(7), 857–877, doi:10.1016/j.marpetgeo.2004.03.009. McQuarrie, N. (2004), Crustal scale geometry of the Zagros fold-thrust belt, Iran, J. Struct. Geol., 26(3), 519–535, doi:10.1016/j.jsg.2003.08.009. Medwedef, D. A., and J. Suppe (1997), Multibend fault-bend folding, J. Struct. Geol., 19(3–4), 279–292, doi:10.1016/S0191-8141(97)83026-X. Meigs, A. J. (1997), Sequential development of selected Pyrenean thrust faults, J. ofSfructural Geol., 19(96), 3–4, doi:10.1016/S0191-8141(96)00096-X. Metcalf, J. R., P. G. Fitzgerald, S. L. Baldwin, and J. A. Muñoz (2009), hermochronology of a convergent orogen: Constraints on the timing of thrust faulting and subsequent exhumation of the Maladeta Pluton in the Central Pyrenean Axial Zone, Earth Planet. Sci. Lett., 287(3–4), 488–503, doi:10.1016/j.epsl.2009.08.036. Miller, J. F., and S. Mitra (2011), Deformation and secondary faulting associated with basement- involved compressional and extensional structures, Am. Assoc. Pet. Geol. Bull., 95(4), 675–689, doi:10.1306/09131010007. Mitra, S., and V. S. Mount (1998), Foreland basement-involved structures, Am. Assoc. Pet. Geol. Bull., 82, 70–109, doi:10.1306/1D9BC39F-172D-11D7-8645000102C1865D. Mouthereau, F., P.-Y. Filleaudeau, A. Vacherat, R. Pik, O. Lacombe, M. G. Fellin, S. Castelltort, F. Christophoul, and E. Masini (2014), Placing limits to shortening evolution in the Pyrenees: Role of margin architecture and implications for the Iberia/Europe convergence, Tectonics, 33(DECEMBER 2014), 2283–2314, doi:10.1002/2014TC003663. Muñoz, J. A. (1992), Evolution of a continental collision belt: ECORS-Pyrenees crustal balanced cross-section, in hrust Tectonics, edited by K. R. McClay, pp. 235–246, Springer Netherlands, Dordrecht, the Netherlands. Naylor, M., and H. D. Sinclair (2008), Pro- vs. retro-foreland basins, Basin Res., 20(3), 285–303, doi:10.1111/ j.1365-2117.2008.00366.x. Naylor, M., H. D. Sinclair, S. D. Willett, and P. A. Cowie (2005), A discrete element model for orogenesis and accretionary wedge growth, J. Geophys. Res. Solid Earth, 110(12), 1–16, doi:10.1029/2003JB002940. Olivet, J. L. (1996), La cinématique de la plaque Ibérique, Bull. des Centres Rech. Elf Explor. Prod., 20, 191–195. Chapter 1. Introduction 29

Persson, K. S., and D. Sokoutis (2002), Analogue models of orogenic wedges controlled by erosion, Tectonophysics, 356(4), 323–336, doi:10.1016/S0040-1951(02)00443-2. Pifner, O. A., S. Ellis, and C. Beaumont (2000), Collision tectonics in the Swiss Alps: Insight from geodynamic modeling, Tectonics, 20(2), 288. Plaziat, J. (1981), Late Cretaceous to late Eocene paleogeographic evolution of southwest Europe, Palaeogeogr. Palaeoclimatol. Palaeoecol., 36, 263–320. Price, R. A. (1973), Large-scale gravitational low of supracrustal rocks, southern Canadian Rockies, Gravity and tectonics, 491–502. Puigdefàbregas, C., and P. Souquet (1986), Tecto-sedimentary cycles and depositional sequences of the Mesozoic and Tertiary from the Pyrenees, Tectonophysics, 129(1–4), 173–203, doi:10.1016/0040-1951(86)90251-9. Puigdefàbregas, C., J. A. Muñoz, and J. Vergés (1992), hrusting and foreland basin evolution in the Southern Pyrenees, in hrust Tectonics, pp. 247–254, Springer Netherlands, Dordrecht. Rahl, J. M., S. H. Haines, and B. A. van der Pluijm (2011), Links between orogenic wedge deformation and erosional exhumation: Evidence from illite age analysis of fault rock and detrital thermochronology of syn-tectonic conglomerates in the Spanish Pyrenees, Earth Planet. Sci. Lett., 307(1–2), 180–190, doi:10.1016/j.epsl.2011.04.036. Ramos, E., P. Busquets, and J. Vergés (2002), Interplay between longitudinal luvial and transverse alluvial fan systems and growing thrusts in piggyback basin (SE Pyrenees), Sediment. Geol., 146(1–2), 105–131, doi:10.1016/S0037-0738(01)00169-5. Rich, J. L. (1934), Mechanics of low-angle overthrust faulting as illustrated by Cumberland thrust block, Virginia, Kentucky, and Tennessee, Am. Assoc. Pet. Geol. Bull., 18(12), 1584–1596. Richter, F., and D. McKenzie (1978), Simple plate models of mantle convection, J. Geophys., 44, 441–471. Roeder, D. (1992), hrusting and wedge growth, Southern Alps of Lombardia (Italy), Tectonophysics, 207(1–2), 199–243, doi:10.1016/0040-1951(92)90478-O. Roest, W. R., and S. P. Srivastava (1991), Kinematics of the plate boundaries between Eurasia, Iberia, and Africa in the North Atlantic from the Late Cretaceous to the present, Geology, 19(6), 613–616. Rosell, J., R. Linares, and C. Llompart (2001), El “Garumniense” Prepirenaico, Rev. Soc. Geol. España, 14((1-2)), 47–56. Rosenbaum, G., G. S. Lister, and C. Duboz (2002), Relative motions of Africa, Iberia and Europe during Alpine orogeny, Tectonophysics, 359(1–2), 117–129, doi:10.1016/S0040-1951(02)00442-0. Rossetti, F., C. Faccenna, and G. Ranalli (2002), he inluence of backstop dip and convergence velocity in the growth of viscous doubly-vergent orogenic wedges: Insights from thermomechanical laboratory experiments, J. Struct. Geol., 24(5), 953–962, doi:10.1016/S0191-8141(01)00127-4. Roure, F., and P. Choukroune (1998), Contribution of the ECORS seismic data to the Pyrenean geology: crustal architecture and geodynamic evolution of the Pyrenees, Mémoires la Soc. Geol. Fr., 173, 37–52. Roure, F., P. Choukroune, X. Berastegui, J. A. Muñoz, A. Villien, P. Matheron, M. Bareyt, M. Seguret, P. Camara, and J. Deramond (1989), Ecors deep seismic data and balanced cross sections: Geometric constraints on the evolution of the Pyrenees, Tectonics, 8(1), 41–50, doi:10.1029/TC008i001p00041. Rowan, M. G., and R. A. Ratlif (2012), Cross-section restoration of salt-related deformation: Best practices and potential pitfalls, J. Struct. Geol., 41, 24–37, doi:10.1016/j.jsg.2011.12.012. Royden, L. H. (1993), he tectonic expression slab pull at continental convergent boundaries, Tectonics, 12(2), 303–325. Ruh, J. B., B. J. P. Kaus, and J.-P. Burg (2012), Numerical investigation of deformation mechanics in fold-and- thrust belts: Inluence of rheology of single and multiple décollements, Tectonics, 31(3), n/a-n/a, doi:10.1029/2011TC003047. de Saint Blanquat, M., J. M. Lardeaux, and M. Brunel (1990), Petrological arguments for high-temperature extensional deformation in the Pyrenean Variscan crust (Saint Barthélémy Massif, Ariège, France), Tectonophysics, 177(1–3), 245–262, doi:10.1016/0040-1951(90)90284-F. de Saint Blanquat, M., F. Bajolet, A. Grand’Homme, A. Proietti, M. Zanti, A. Boutin, C. Clerc, Y. Lagabrielle, and P. Labaume (2016), Cretaceous mantle exhumation in the central Pyrenees: New constraints from the peridotites in eastern Ariège (North Pyrenean zone, France), Comptes Rendus - Geosci., 348(3–4), 268–278, doi:10.1016/j.crte.2015.12.003. Saura, E., L. Ardèvol I Oró, A. Teixell, and J. Vergés (2016), Rising and falling diapirs, shiting depocenters, and lap overturning in the Cretaceous Sopeira and Sant Gervàs subbasins (Ribagorça Basin, southern Pyrenees), Tectonics, 35(3), 638–662, doi:10.1002/2015TC004001. 30 Chapter 1. Introduction

Schellart, W. P., J. Freeman, D. R. Stegman, L. Moresi, and D. May (2007), Evolution and diversity of subduction zones controlled by slab width., Nature, 446(March), 308–311, doi:10.1038/nature05615. Schubert, G. (1992), Numerical Models of Mantle Convection, Annu. Rev. Fluid Mech., 24, 359–394. Sclater, J. G., and P. A. F. Christie (1980), Continental stretching: an explanation of the post-mid-cretaceous subsidence of the central north sea basin, J. Geophys. Res., 85(B7), 3711–3739. Seguret, M., and M. Daignieres (1986), Crustal scale balanced cross-sections of the Pyrenees; discussion, Tectonophysics, 129(1–4), 303–318, doi:10.1016/0040-1951(86)90258-1. Sibuet, J.-C., S. P. Srivastava, and W. Spakman (2004), Pyrenean orogeny and plate kinematics, J. Geophys. Res., 109(B8), B08104, doi:10.1029/2003JB002514. Simpson, G. (2011), Mechanics of non-critical fold-thrust belts based on inite element models, Tectonophysics, 499(1–4), 142–155, doi:10.1016/j.tecto.2011.01.004. Simpson, G. D. H. (2010), Inluence of the mechanical behaviour of brittle-ductile fold-thrust belts on the development of foreland basins, Basin Res., 22(2), 139–156, doi:10.1111/j.1365-2117.2009.00406.x. Sinclair, H. D., and M. Naylor (2012), Foreland basin subsidence driven by topographic growth versus plate subduction, Bull. Geol. Soc. Am., 124, 368–379, doi:10.1130/B30383.1. Sinclair, H. D., M. Gibson, M. Naylor, and R. G. Morris (2005), Asymmetric growth of the Pyrenees revealed through measurement and modeling of orogenic luxes, Am. J. Sci., 305, 369–406, doi:10.2475/ ajs.305.5.369. Sonnenfeld, P. (1984), Brines and evaporites, Academic Press Inc., Orlando, Florida. Souquet, P., and B. Peybernès (1987), Allochtonie des massifs primaires nord-pyrénéens des Pyrénées Centrales, C. R. Acad. Sc. Paris, 305(8), 733–739. Souquet, P., B. Peybernès, M. Bilotte, and E.-J. Debroas (1977), La chaîne alpine des Pyrénées, Géologie Alp., 53, 193–216. Souquet, P. et al. (1985), Le groupe du Flysch noir (albo-cénomanien) dans les Pyrénées, Bull. des Centres Rech. Elf Explor. Prod., 9, 183–252. Srivastava, S. P., W. R. Roest, L. C. Kovacs, G. Oakey, S. Lévesque, J. Verhoef, and R. Macnab (1990), Motion of Iberia since the Late Jurassic: Results from detailed aeromagnetic measurements in the Newfoundland Basin, Tectonophysics, 184(3–4), 229–260, doi:10.1016/0040-1951(90)90442-B. Steckler, M. S., and A. B. Watts (1978), Subsidence of the Atlantic-type continental margin of New York, Earth Planet. Sci. Lett., 41(1), 1–13, doi:10.1016/0012-821X(78)90036-5. Storti, F., and K. McClay (1995), Inluence of syntectonic sedimentation on thrust wedges in analogue models, Geology, 23(11), 999–1002, doi:10.1130/0091-7613(1995)023<0999:IOSSOT>2.3.CO;2. Storti, F., F. Salvini, and K. R. McClay (2000), Synchronous and velocity-partitioned thrusting and thrust polarity reversal in experimentally produced, doubly-vergent thrust wedges: Implications for natural orogens, Tectonics, 19(2), 378–396, doi:10.1029/1998TC001079. Suppe, B. J., and D. A. Medwedef (1990), Geometry and kinematics of fault-propagation folding, Eclogae Geol. Helv., 83(3), 409–454. Suppe, J. (1980), A retrodeformable cross section of northern Taiwan, Proc. Geol. Soc. china, 23, 46–55. Tambareau, Y., B. Crochet, J. Villatte, and J. Deramond (1995), Evolution tectono-sedimentaire du versant nord des Pyrenees centre-orientales au Paleocene et a l’Eocene inferieur, Bull. la Soc. Géologique Fr., 166(166), 375–387, doi:10.2113/gssgbull.166.4.375. Teixell, A. (1998), Crustal structure and orogenic material budget in the west central Pyrenees, Tectonics, 17(3), 395–406, doi:10.1029/98TC00561. Teixell, A., P. Labaume, and Y. Lagabrielle (2016), he crustal evolution of the west-central Pyrenees revisited: Inferences from a new kinematic scenario, Comptes Rendus - Geosci., 348(3–4), 257–267, doi:10.1016/j. crte.2015.10.010. hieulot, C. (2011), FANTOM: Two- and three-dimensional numerical modelling of creeping lows for the solution of geological problems, Phys. Earth Planet. Inter., 188(1–2), 47–68, doi:10.1016/j. pepi.2011.06.011. Turcotte, D. L., and G. Schubert (1982), Geodynamics: Applications of continuum physics to geological problems, Wiley. Ueda, K., S. D. Willett, T. Gerya, and J. B. Ruh (2015), Geomorphological-thermo-mechanical modeling: Application to orogenic wedge dynamics, Tectonophysics, 659, 12–30, doi:10.1016/j.tecto.2015.08.001. Vergés, J. (1993), Estudi geològic del vessant sud del Pirineu oriental i central. Evolució cinemàtica en 3D, University of Barcelona, Spain. Chapter 1. Introduction 31

Vergés, J., J. A. Muñoz, and A. Martínez (1992), South Pyrenean fold and thrust belt: he role of foreland evaporitic levels in thrust geometry, in hrust Tectonics, edited by K. R. McClay, pp. 255–264, Springer Netherlands, Dordrecht, the Netherlands. Vergés, J., H. Millán, E. Roca, J. A. Muñoz, M. Marzo, J. Cirés, T. Den Bezemer, R. Zoetemeijer, and S. Cloetingh (1995), Eastern Pyrenees and related foreland basins: pre-, syn- and post-collisional crustal-scale cross- sections, Mar. Pet. Geol., 12(8), 903–915, doi:10.1016/0264-8172(95)98854-X. Vergés, J., M. Marzo, T. Santaeulària, J. Serra-Kiel, D. W. Burbank, J. A. Muñoz, and J. Giménez-Montsant (1998), Quantiied vertical motions and tectonic evolution of the SE Pyrenean foreland basin, in Cenozoic foreland Basins of Western Europe, Geological Society Special Publication, vol. 134, edited by A. Mascle, C. Puigdefàbregas, H. P. Lutherbacher, and M. Fernàndez, pp. 107–134, Geological Society of London, London, United Kingdom. Vergés, J., M. Fernàndez, and A. Martínez (2002), he Pyrenean orogen: Pre-, syn-, and post-collisional evolution, J. Virtual Explor., 8, 55–74, doi:10.3809/jvirtex.2002.00058. Vissers, R. L. M., and P. T. Meijer (2012a), Iberian plate kinematics and Alpine collision in the Pyrenees, Earth- Science Rev., 114(1–2), 61–83, doi:10.1016/j.earscirev.2012.05.001. Vissers, R. L. M., and P. T. Meijer (2012b), Mesozoic rotation of Iberia: Subduction in the Pyrenees?, Earth-Science Rev., 110(1–4), 93–110, doi:10.1016/j.earscirev.2011.11.001. Walcott, R. I. (1970), Flexure of the lithosphere at Hawaii, Tectonophysics, 9(5), 435–446, doi:10.1016/0040- 1951(70)90056-9. Waltham, D. (1989), Finite diference modelling of hangingwall deformation, J. Struct. Geol., 11(4), 433–437, doi:10.1016/0191-8141(89)90020-5. Waltham, D., and S. Hardy (1995), he velocity description of deformation. paper 1: theory, Mar. Pet. Geol., 12(2), 153–163, doi:10.1016/0264-8172(95)92836-L. Watts, A. B. (1992), he efective elastic thickness of the lithosphere and the evolution of foreland basins, Basin Res., 4(3–4), 169–178, doi:10.1111/j.1365-2117.1992.tb00043.x. Whitchurch, A. L., A. Carter, H. D. Sinclair, R. A. Duller, A. C. Whittaker, and P. A. Allen (2011), Sediment routing system evolution within a diachronously upliting orogen: Insights from detrital zircon thermochronological analyses from the South-Central Pyrenees, Am. J. Sci., 311(5), 442–482, doi:10.2475/05.2011.03. Willett, S. D. (1999), Orogeny and orography: he efects of erosion on the structure of mountain belts, J. Geophys. Res. Solid Earth, 104(B12), 28957–28981, doi:10.1029/1999JB900248. Willett, S. D., C. Beaumont, and P. Fullsack (1993), Mechanical model for the tectonics of doubly vergent compressional orogens, Geology, 21(4), 371–374, doi:10.1130/0091-7613(1993)021<0371:MMFTTO>2. 3.CO. Willett, S. D., R. Slingerland, and N. Hovius (2001), Uplit , Shortening , and Steady State Topography in Active Mountain Belts, , 301, 455–485, doi:10.2475/ajs.301.4-5.455. Woodward, N. B., S. E. Boyer, and J. Suppe (1989), Balanced Geological Cross-Sections: An Essential Technique in Geological Research and Exploration, edited by M. L. Crawford and E. Padovani, American Geophysical Union, Washington D.C., United States of America. Xie, X., and P. L. Heller (2009), Plate tectonics and basin subsidence history, Bull. Geol. Soc. Am., 121(1–2), 55–64, doi:10.1130/B26398.1.

Chapter 2

Insights into the crustal-scale dynamics of a doubly vergent orogen from a quantitative analysis of its forelands: A case study of the Eastern Pyrenees

Published in Tectonics, 37(2), 450-476, doi:10.1002/2017TC004731 Arjan R. Grool1,2, Mary Ford1, Jaume Vergés3, Ritske S. Huismans2, Frédéric Christophoul4, Armin Dielforder1,* 1CRPG, UMR 7358, 15 Rue Notre Dame des Pauvres, 54501 Vandœuvre-lès-Nancy, France. 2Department of Earth Sciences, University of Bergen, Bergen, Norway 3Institute of Earth Sciences Jaume Almera, ICTJA-CSIC, Lluis Solé i Sabarís s/n, 08028 Barcelona, Spain 4GET, UMR 5563, Université de Toulouse-CNRS-IRD-OMP, 14 Avenue Edouard Belin, 31400 Toulouse, France *Now at Helmholz Centre Potsdam, GFZ German Research Centre for Geosciences, Telegrafenberg, 11473 Potsdam, Germany 34 Chapter 2. Case study of the Eastern Pyrenees

2.1 Introduction It is common in foreland basin modelling and fold-and-thrust belt studies to analyse only one side of a doubly vergent orogen [e.g., Ershov et al., 2003; Fillon et al., 2013; Decarlis et al., 2014]. However several studies, mainly based on mechanical modelling, suggest that pro- and retro-wedges interact in accommodating plate convergence [Willett et al., 1993; Sinclair et al., 2005; Hoth et al., 2007, 2008; Erdős et al., 2015]. Correlating the behaviour of the two wedges in doubly vergent orogens can therefore provide important insight into orogen dynamics. How deformation and subsidence of the pro- and retro-foreland of a natural system relate to each other, whether this relationship changes through time, and what would cause such a change remains unclear. Detailed comparisons of the two wedges and their foreland basins in terms of structure and evolution are required to investigate these issues. In this paper we investigate the Eastern Pyrenees, comparing and correlating the northern retro-foreland and southern pro-foreland along two sections directly opposite each other to better understand full orogen evolution (Figure 2.1). he South Pyrenean fold-and-thrust belt and Ebro foreland basin system is extremely well documented owing to excellent outcrop conditions and well-deined stratigraphy [e.g., Burbank et al., 1992b; Vergés et al., 1992; López-Blanco et al., 2000; Gómez-Paccard et al., 2012]. It developed on the subducting Iberian plate starting in the Late Cretaceous. Subsidence records show locally short-lived and accelerating subsidence owing to constant outward basin migration and incorporation of the proximal basin into the thrust wedge. his corresponds to simulated behaviour of kinematic and mechanical models of pro-forelands [Beaumont et al., 2000; Naylor and Sinclair, 2008; Sinclair and Naylor, 2012; Fillon et al., 2013; Erdős et al., 2014]. In contrast, owing to poorer outcrop conditions and structural and stratigraphic complexity, our understanding of the North Pyrenean retro-wedge is less well developed despite a large body of work spanning several decades [e.g., Choukroune, 1974; Souquet et al., 1977; Puigdefàbregas and Souquet, 1986; Lagabrielle et al., 2010; Clerc et al., 2015]. he retro-wedge preserves an inverted hyper- extended rit system of Early Cretaceous age, as recorded by the presence of mantle blocks and syn- extension high-temperature metamorphism in the internal, southern part of the retro-wedge [Jammes et al., 2009; Lagabrielle et al., 2010; Mouthereau et al., 2014; Clerc et al., 2015; Vacherat et al., 2016]. In general, retro-wedges can be expected to accommodate less shortening than pro-wedges, resulting in slow and long-lived foreland basin subsidence [Sinclair and Naylor, 2012]. Previous work proposes that the western Aquitaine Basin follows this predicted pattern [Desegaulx and Brunet, 1990; Naylor and Sinclair, 2008], however more recent work shows that the eastern retro-foreland basin has a more complex history that challenges these models, including a temporary stagnation in subsidence for which the cause remains unclear [Ford et al., 2016; Rougier et al., 2016]. To be able to make a detailed comparison of the pro- and retro-forelands, we studied the tectonic and stratigraphic evolution of the North Pyrenean Zone and Aquitaine foreland basin in the Eastern Pyrenees (Ariège, France; Figure 2.1). Our work is based on a structural analysis and cross section balancing that integrated newly acquired ield data, seismic relection proiles, borehole data, and published stratigraphic and structural data [e.g., Choukroune, 1974; Baby, 1988; Bilotte et al., 1988b]. his allowed us to decipher the geometry, timing, rates, and kinematics of deformation and to identify and characterize the various phases, styles, and amounts of Santonian to Oligocene shortening in the eastern Pyrenean retro-wedge and its foreland basin. We then correlated this complete retro-wedge history with its pro-wedge counterpart in the Spanish Pyrenees using the work of Vergés [1993], Vergés et al. [1995, 1998, 2002], and Vergés and García-Senz [2001]. We identiied key diferences and point out possible interactions between the pro- and retro-foreland fold belts. hen, we investigate how shortening was partitioned in time and space between pro- and retro-wedge and discuss what controls the sequence of deformation during orogeny by presenting a new crustal-scale model for the evolution of the Eastern Pyrenees. Chapter 2. Case study of the Eastern Pyrenees 35

4° 3° 2° 1° 0° 1° 2° 3° a 44° Atlantic Ocean France Bay of Biscay Massif Santander Aquitane Basin Central Toulouse San Sebastian Bilbao ECORS PYRENEES Pau Tarbes S3 Carcassonne Narbonne 43° Fig. 2 Vitoria Bay Western Pyrenees 43° Pamplona of Central Lyon and Perpignan Logroño Jaca S3b Burgos Andorra Eastern Pyrenees Tremp Huesca Ripoll 42° South Spain Central Unit Girona 42°

Iberian Range Zaragoza Lleida J3 Duero Basin Ebro Basin

Barcelona Neogene-Quaternary Catalan Coastal Ranges 41° Eocene-Miocene Tarragona Mediterranean Sea Eocene-Oligocene València Trough 41° Late Cretaceous-Paleocene thrust Early Cretaceous blind thrust 0 50 100 km Basement normal fault 4° 3° 2° 1° 0° 1° 2° 3° Tectonic Aquitaine b Ebro Basin SPZ Axial Zone NPZ Sub- zones: PZ Basin NPF NPFT -10 S Nogueres N -10

0 Orri 0 Depth (km) Rialp 10 10 20 20 European plate 30 Iberian plate 30 Depth (km) 40 Sediments 40 50 Pyrenean basement 50 Upper Crust 0 50 km Lower Crust

Figure 2.1. (a) Simpliied geologic map of the Pyrenees. Red lines locate the cross sections (Figs. 2.4, 2.5, 2.9). Black line indicates the trace of the ECORS Pyrenees deep seismic section shown in (b). Modiied ater Vergés et al. [2002] and Angrand et al. [2018]. (b) Interpretation of the ECORS Pyrenees deep seismic section, with major structural zones indicated. Modiied ater Muñoz [1992]. SPZ, South Pyrenean Zone; NPZ, North Pyrenean Zone; Sub-PZ, Sub-Pyrenean Zone; NPF, North Pyrenean Fault; NPFT, North Pyrenean Frontal hrust. 2.2 Geological Setting 2.2.1 Plate Kinematics he Pyrenees formed as part of the Alpine belt due to the collision of the Iberian microplate with the European continent. Before the collision, an Aptian to early Cenomanian oblique rit system associated with exhumation of mantle rocks formed between the two plates [Debroas, 1990; Jammes et al., 2009; Lagabrielle et al., 2010]. Riting has been related to the opening North Atlantic Ocean [Srivastava et al., 1990; Olivet, 1996; Sibuet et al., 2004]. No consensus on plate kinematics around Iberia yet exists and there are several competing models. Most models are based on reconstructions of magnetic sea-loor anomalies of the Bay of Biscay, Central, and North Atlantic Ocean [Olivet, 1996; Sibuet et al., 2004; Vissers and Meijer, 2012a, 2012b]. hese models predict diferent extensional histories for the Pyrenean realm, and, ater the onset of convergence (~84 Ma), predict diferent degrees of obliquity for N-S shortening. Pyrenean tectonic structures predominantly record a N-S shortening direction and a minor strike-slip component that was mainly focussed along the North Pyrenean Fault [Souquet et al., 1977; Choukroune and Mattauer, 1978]. 2.2.2 Tectonic Zones and Major Boundaries In the Eastern Pyrenees the orogen is divided into six tectonostratigraphic zones (Figure 2.1b). From north to south, the Aquitaine foreland basin is followed by the sub Pyrenean Zone, a folded and faulted zone north of the North Pyrenean Frontal hrust, usually delimited by a blind thrust [Souquet et al., 1977; Ford et al., 2016; Rougier et al., 2016]. South of the North Pyrenean Frontal hrust is the North Pyrenean 36 Chapter 2. Case study of the Eastern Pyrenees

Zone, a narrow, north verging fold-and-thrust belt. In the study area this zone comprises: inverted rit basins (e.g., Fougax and Roquefeuil blocks; Figure 2.2), inverted basement massifs (e.g., Saint-Barthélémy Massif; Figure 2.2), and the Metamorphic Internal Zone (Figure 2.2) [Souquet et al., 1977]. he North Pyrenean Fault separates the North Pyrenean Zone from the Axial Zone and is traditionally seen as the suture between Europe and Iberia. In the classic interpretation of the ECORS section lying some 50 km to the west (see below) [e.g., Muñoz, 1992] crustal thickening in the Axial Zone is related to a south verging, antiformal stack of Iberian crust. Interpretations in the eastern and western Pyrenees show almost no Alpine folding of Paleozoic basement rocks of the Axial Zone [Vergés et al., 2002; Teixell et al., 2016]. South of the Axial Zone lies the South Pyrenean Zone, a south verging thin-skinned fold-and-thrust belt with syn-orogenic thrust sheet-top basins. he Ebro Basin forms the pro-foreland basin of the Pyrenees. It is an asymmetric double lexural basin with a central intrabasinal high due to loading by the Pyrenees in the north and the Catalan Coastal Ranges in the south. he deep crustal structure of the Central Pyrenees is imaged by the ECORS Pyrenees deep seismic study (Figure 2.1b) [Roure et al., 1989; Muñoz, 1992; Roure and Choukroune, 1998] and more recently by seismic tomography studies [Souriau et al., 2008; Chevrot et al., 2015; Wang et al., 2016]. hese data show the Iberian Moho underthrusting the European plate, deepening northward from 35 km to 50 km depth [Beaumont et al., 2000]. Most authors show the European Moho at a fairly constant 30 km depth along the ECORS line [Beaumont et al., 2000; Díaz and Gallart, 2009] and some show a gentle southward dip [Roure et al., 1989; Roure and Choukroune, 1998]. Recent seismic tomography data show the subduction interface dipping 20° northward and show the European Moho rising southward [Chevrot et al., 2015]. he nature and extent of Alpine deformation in the Axial Zone, and the role of inherited Variscan shear zones are currently subject to renewed debate and analysis with modern tools [e.g. Laumonier, 2015; Cochelin et al., 2017]. In this paper we focus on the fold-and-thrust belts and foreland basins on both sides of the Eastern Pyrenees. As a detailed analysis of the Axial Zone is beyond the scope of this paper, we use the Axial Zone model of Vergés et al. [1995, 2002] and Vergés and García-Senz [2001], which remains valid.

2.3 Methods In order to investigate the structural evolution of the retro-foreland, we constructed two cross sections, the main section S3 and an auxiliary section S3b (for location see Figures 2.1a, 2.2). hese sections were positioned directly opposite cross section J3 of Vergés [1993] in the pro-foreland. Standard balancing techniques were used to restore the sections [Woodward et al., 1989]. While most sedimentary strata were line-length balanced, weak units with signiicant internal deformation were also area balanced. We used the restored cross sections combined with stratigraphic constraints on timing of activity to estimate average shortening rates for each map-scale structure across the pro- and retro-forelands of the Eastern Pyrenees. he minimum overall shortening rate for a given period was derived by summing the shortening rates for all unlinked structures active during that period. A complete high resolution shortening history across the whole orogen can thus be constructed. Such an analysis integrating detailed map-scale structural history is especially important in areas that record very slow and distributed shortening like the Pyrenees. his shortening history was then used to constrain the crustal scale restoration model, together with subsidence histories, thermochronology data, and deep geophysical data. he retro-foreland database includes surface geology from 1:50,000 geological maps of the French Geological Survey (BRGM; 1036 Castelnaudary [Cavaillé et al., 1975a]; 1058 Mirepoix [Cavaillé, 1976a]; 1075 Foix [Bilotte et al., 1988a]; 1076 Lavelanet [Bilotte et al., 1988b]), an unpublished geological map [Marty, 1976], publicly available borehole data (Bureau Exploration-Production des Hydrocarbures; beph.net), limited seismic relection data provided by the BRGM, a 90 m SRTM Digital Elevation Model processed and distributed by the CGIAR Consortium for Spatial Information (srtm.csi.cgiar.org) [Jarvis et al., 2008], and our own ield data. he stratigraphic framework used for the retro-foreland was established by the ANR PYRAMID research project [Ford et al., 2016; Rougier et al., 2016] and linked to the 2013/01 chronostratigraphic chart published by the International Commission on Stratigraphy [Cohen et al., Chapter 2. Case study of the Eastern Pyrenees 37

2013]. Details on stratigraphic nomenclature and equivalent BRGM units are presented in Table A.1 (Supplementary material). he litho- and chronostratigraphy for the pro-foreland are based on Vergés [1993] and Vergés et al. [1998, 2002]. Total and tectonic subsidence curves for four wells in the retro-foreland are compared with those for six wells or sections in the pro-foreland [Vergés et al., 1998], all calibrated to present-day sea level. We used standard methods of decompaction and backstripping assuming Airy isostasy [Steckler and Watts, 1978]. Decompaction calculations integrate the values for initial porosity and compaction coeicient for each lithology as used in Vergés et al. [1998]. For the backstripping process each unit was given bulk values for those constants based on the percentages of each lithology in that particular unit (Table A.2). hese percentages were determined from the highest resolution sections available, ranging in scale between 1:500 and 1:2000. Estimates for paleo-bathymetry were based on depositional environments. Continental deposits were considered to be equivalent to water-illed basins with 0 or -50 m bathymetry. We used a mantle density of 3300 kg/m3 and a density of 1000 kg/m3 for water. he original published subsidence

400 405 410 415 420

Aquitaine Basin N

Lavelanet Anticline Dr3 Dr1 NPFT Dr4 24° 70° Bx1 29°

Nalzen Block 40° Benaix Belesta 30°

65°

Fougax- Montségur et-Barrineuf 49° Fig. 6a 75° Fougax Block 82° 75° Saint-Barthélémy 83° Massif Roquefeuil Block 70° Frau Fault Fig. 6b la Frau (1925 m) 3MF (1) 59° Espezel 46° Gorges 39° Roquefeuil de la Frau3MF(2) Belcaire Fig. 6c 4740 4745 4750 4755 4760 4740 4745 4750 4755 4760 Paleozoic crystalline Camurac S3 3MF Metamorphic Internal S3b Zone Axial Zone NPF 5 km 400 405 410 415 420 Legend Plantaurel Group Black Dolomite Group borehole Carcassonne Group Petites Pyrénées Group Keuper Group village Coustouge Group Grey Flysch Group Metamorphic Internal Zone Scree Rieubach Group Black Flysch Group Basement Quaternary Aude Valley Group Mirande Limestone Group Lherzolite

South of NPF: Upper Cretaceous Lower Cretaceous Jurassic Triassic

Figure 2.2. Geologic map of the North Pyrenean Zone and sub Pyrenean Zone, compiled from the BRGM 1:50 000 geologic maps (1075 Foix and 1076 Lavelanet), Marty [1976], and de Saint Blanquat et al. [2016]. Boreholes are located between the North Pyrenean Frontal hrust and the Lavelanet Anticline. he blue lines indicate partial section traces of S3 and S3b. NPFT, North Pyrenean Frontal hrust; 3MF, 3M Fault; 3MF(1), lat of 3M Fault; 3MF(2), hanging wall shortcut of 3M Fault; NPF, North Pyrenean Fault. 38 Chapter 2. Case study of the Eastern Pyrenees calculations for the pro-foreland did not take into account eustatic sea level changes [Vergés et al., 1998]. hese data were therefore converted using the same eustatic data as used for the retro-foreland [Snedden and Liu, 2010]. In displaying total and tectonic subsidence relative to present-day sea level, many curves start above 0 m because paleo-sea level was between 160 and 210 m higher than present day and paleo- bathymetry in the forelands was relatively minor. In the case of the pro-foreland, including the long term eustatic sea level curve may introduce artefacts in the subsidence. Brief periods of low bathymetry in the Eocene are shown as tectonic uplit events in our converted curves, but the observed reduction in bathymetry may have been caused by a short-lived deviation from the long term sea level curve instead [Burbank et al., 1992b]. Errors in the estimation of bathymetry have the greatest efect on calculated subsidence. Errors in sea level and sediment ages have less impact.

2.4 North Pyrenean Foreland 2.4.1 Stratigraphy of the North Pyrenean Foreland Across the Saint-Barthélémy – Fougax study area a Mesozoic to Cenozoic succession overlies Variscan basement, that comprises Paleozoic metasediments and crystalline rocks (Figure 2.3b). Triassic to lower Cenomanian marine strata record rit-related subsidence. Upper Cenomanian to Santonian strata record post-rit subsidence. On a regional scale the syn-orogenic succession (Campanian to Oligocene) records a consistent along-strike transition from continental to marine facies to the west with facies interingering in the study area. No strata older than Cenomanian overlie Variscan basement north of the Benaix Fault implying that the foreland was largely unafected by Mesozoic riting. No strata younger than Campanian are preserved south of the thrust front.

a Distance from reference point (km) b Distance from reference point (km, doubled horiz. scale) c 0 20 40 60 80 100 120 60 40 20 0 Tectonics Chattian Continental Boreholes Dr3 T1 MVL4MRL2 MRL1 sedimentation Dr4 T2 MRL5 SPL5 30 Rupelian Phase 4 30 Oligoc. 16 15 Bx1 SPL7 Priabonian 14 Marine 13 40 Bartonian 12 11 sedimentation 40 Lutetian 10 9 8 7 CSG Phase 3

50 Eocene 6 5 50 Ypresian 2 4 3 CG Thanetian Catalan Ebro Foreland RG 60 1 60 Selandian C.R. Basin Cadí Phase 2 Danian AVG

Paleoc. thrust sheet NP Quiesc. 70 Maastricht. Stratigraphic formations 70

PG Time (Ma) 1 Tremp (Garumnian facies) Aquitaine Basin Phase 1 Campanian 2 Corones PPG 80 80 3 Sagnari Cadí Time (Ma) Time Santonian 4 Coniacian 5 Armàncies and Campdevànol 90 Stratigraphic groups Post-rift 90 Turonian 6 Penya GFG thermal Late Cretaceous CSG Carcassonne Group 7 Beuda evaporites subsidence Cenoman. 8 Bellmunt Hiatus CG Coustouge Group 100 Nalzen Block 100 9 Banyoles RG Rieubach Group 10 Tavertet AVG Aude Valley Group Albian 11 Milany PG Plantaurel Group 110 110 12 Igualada PPG Petites Pyrénées Group 13 Montserrat BFG GFG Grey Flysch Group 14 Cardona evaporites BFG Black Flysch Group 120 Aptian 120 Berga MG Mirande Group 15 Rifting 16 Solsona Barremian 130 Equivalent to Pedraforca thrust sheets 130 Hauterivian South Pyrenean North Pyrenean Fougax Block Early Cretaceous foreland foreland Valanginian MG 140 140 Berriasian

Continental facies Marine facies Fluvial sandstones and conglomerates Shallow marine clastics Carbonates, internal platform Alternating marine sandstones and shales (turbidites) Mudstone to sandstone (alluvial plain) Evaporites Carbonates, platform margin Open marine shale/marl Lacustrine limestone Carbonates, external platform (distal turbidites/hemipelagics)

Figure 2.3. Chronostratigraphic charts for the north and south Pyrenean foreland basins. (a) South Pyrenean stratigraphic chart modiied ater Vergés [1993] and Vergés et al. [2002]. (b) North Pyrenean stratigraphic chart. Boreholes Dr3, Dr4 and Bx1 are located on Figure 2.2. Note double horizontal scaling compared to (a). For detailed north Pyrenean formation names and BRGM unit codes see Table A.1. (c) Timing of main tectonic events. Phases 1 to 4 represent the Pyrenean orogeny. Chapter 2. Case study of the Eastern Pyrenees 39

South of the North Pyrenean Frontal hrust, Triassic evaporites (Keuper Group; unconstrained thickness) lie at the base of the Roquefeuil and Fougax block successions [Marty, 1976]. Jurassic limestones and dolomites occur across the northern North Pyrenean Zone (Black Dolomite Group; 100 to 200 m; Figure 2.2). Berriasian to Barremian limestones are present in the Roquefeuil block (1000 m, Mirande Group, Figure 2.3b) but absent in the Fougax and Nalzen blocks (Figures 2.4, 2.5). hese strata record littoral environments, with occasional lagoonal and distal platform deposits [Bilotte et al., 1988c]. In the Metamorphic Internal Zone (Figure 2.2) [Choukroune, 1974] protolith carbonates and siliciclastic strata are interpreted as Jurassic to Early Cretaceous in age [Golberg and Leyreloup, 1990; Lagabrielle et al., 2010] but their true age remains unclear owing to Albian to Cenomanian high-temperature, low-pressure metamorphism and strong internal deformation [Golberg and Leyreloup, 1990]. Widespread breccias predominantly comprise clasts of marble and hornfels, but also rare clasts of basement rocks and mantle lithologies [de Saint Blanquat et al., 2016]. Syn-rit deep marine shales of the Black Flysch Group (~1000 m; Aptian to Lower Cenomanian) are in stratigraphic continuity with the Mirande Group limestones on the southern Fougax block (Figures 2.2, 2.5), while on the northern side they rest unconformably on the Jurassic Black Dolomite Group [Bilotte et al., 1988b]. he term ‘lysch’ is traditionally used in the name of the Mid-Albian to Lower Cenomanian ‘Black Flysch’ Formation, despite its pre-orogenic origin [Souquet et al., 1985; Debroas, 1990]. Here we apply the name ‘Black Flysch Group’ to a wider age range due to very similar lithologies of Aptian to Lower Albian age. Black Flysch shales transition northward into rudist platform limestones (Urgonian facies) of equivalent age (Figure 2.3b) and of variable thickness (maximum ~600 m) [Bilotte et al., 1988b]. he Black Flysch Group thins westward where it is preserved in the Nalzen block north of the Saint- Barthélémy Massif (~600 m; Figure 2.2), thus recording stratigraphic continuity between the Fougax and Nalzen blocks. here, these shales contain marine conglomerates with locally sourced basement clasts (Conglomérat de Freychenet) [Bilotte et al., 1988c]. he post-rit Grey Flysch Group succession (Cenomanian to Santonian) records a deepening upward trend from platform carbonates to deep-water marls (borehole Dr4; Figure 2.3b). South of the North Pyrenean Frontal hrust, it is only preserved in the Nalzen block (Figure 2.2; 1000 to 1500 m). To the north it directly overlies a partially inverted half-graben and Variscan basement (Figure 2.5) and thins northward to pinch out north of the Lavelanet Anticline (Figure 2.4c). here, Santonian strata are either missing or indistinguishable from the overlying syn-orogenic sediments. Santonian sediments are unlikely to have been eroded completely, given the marine depositional environment. herefore, we interpret this as local non-deposition, recording southward olap of the upper Grey Flysch Group. he Petites Pyrénées Group represents the early syn-orogenic succession (Lower to Middle Campanian; Figures 2.3b, 2.4, 2.5). South of the North Pyrenean Frontal hrust the succession reaches a minimum thickness of ~500 m (Nalzen block) and records shallower and more sandy facies than further north. North of the Frontal hrust, the group thins northward from 1500 - 2500 m to pinch out in the Tréziers Anticline (Figures 2.3b, 2.4). Internal deformation is responsible for thickening of the Petites Pyrénées Group below the Lavelanet Anticline. he northern series records a shallowing upward trend from deep marine shales (with secondary siltstones and sandstones) representing delta progradation from east to west [Ricateau and Villemin, 1973; Bilotte, 1985]. It is overlain by and laterally equivalent to the luvio- deltaic Plantaurel Group (Late Campanian to middle Maastrichtian) that was supplied from the east with rare marine incursions from the west [Bilotte, 1985; Bilotte et al., 1988c]. he Plantaurel Group thickens northward from 300 m to 1300 m across the syn-depositionally inverted Benaix Fault and then gradually thins northward to 700 m in the Lavelanet Anticline, reaching 430 m in the Tréziers Anticline before pinching out around Orsans (Figures 2.3b, 2.4). North of the North Pyrenean Frontal hrust, red continental muds, sandstones, and intercalated conglomerates of the Aude Valley Group (Garumnian facies; middle Maastrichtian to hanetian) unconformably overlie the Plantaurel (Lavelanet Anticline; Figure 2.2) [Bilotte et al., 1988b] and Petites 40 Chapter 2. Case study of the Eastern Pyrenees

Pyrénées Groups (near Benaix; Figure 2.2). he Maastrichtian conglomerates contain pebbles of Upper Cretaceous limestones [Bilotte et al., 1988c], recording exhumation and erosion in the hinterland. In borehole Dr4 locally developed Danian to Selandian lacustrine limestones and marls are only ~95 m thick (Table A.3). A brief hanetian marine incursion is recorded in the southern half of the basin by internal platform limestones of the Rieubach Group, which is the western equivalent of, and interingers with, the continental Aude Valley Group [Ford et al., 2016]. Growth strata only appear in the upper hanetian portion of the succession (Figure 2.4c). Overall, the Aude Valley Group thins northward from 700 m south of the Lavelanet Anticline to pinch out just south of borehole MRL1 (Figures 2.3b, 2.4). A more signiicant but brief marine incursion in the early Ypresian (Ilerdian) is recorded across the entire eastern foreland basin at the base of the Coustouge Group [Cavaillé et al., 1975b; Cavaillé, 1976a; Bilotte et al., 1988c]. his group then records a shallowing upward trend with marine marls and limestones passing up into mudstones and nummulitic sandstones [Bilotte et al., 1988c; Tambareau et al., 1995]. hicknesses vary considerably due to syn-sedimentary deformation: from ~700 m immediately north of the Lavelenet Anticline to ~450 m in the Tréziers Anticline, ~500 m in the footwall of the Tréziers hrust, to ~100 m at the northern limit of the preserved foreland basin (borehole MRL1). he northern pinchout is not preserved (Figure 2.1a). he Late Ypresian to Rupelian Carcassonne Group covers the entire foreland north of the Lavelanet Anticline, with the main depocentre in the footwall of the Tréziers hrust (Figure 2.4). hese sediments comprise mainly luvial sandstones. Proximal conglomerates along the thrust front include clasts of crystalline basement, scapolite-bearing limestones, impure marble with garnets, and non- metamorphic Mesozoic sediments, sourced from the south [Bilotte et al., 1988c; Crochet, 1991]. 2.4.2 North Pyrenean Structure he structure of the eastern North Pyrenean Zone in the Saint-Barthélémy – Fougax area is here described along two restored cross-sections (S3 and S3b; Figures 2.4, 2.5). Restoration provides estimates of N-S shortening across the whole northern foreland of 13.3 km (S3) and 13.2 km (S3b), excluding emplacement of the Metamorphic Internal Zone (minimum ~5 km). his is lower than previous estimates of around 25- 30 km that also excluded emplacement of the Metamorphic Internal Zone [Souquet and Peybernès, 1987; Baby, 1988; Baby et al., 1988]. Most of this diference comes from a basement duplex proposed below the Saint-Barthélémy Massif that is not necessary to it our data, and thus is absent in our interpretation. In the Aquitaine Basin the sedimentary inill is gently deformed by the basement-rooted Tréziers and Orsans hrusts (Figures 2.4a, d). Both structures are the western terminations of thrusts accommodating large anticlinal basement uplits in the eastern foreland: the Mouthoumet [Cavaillé, 1976a; Bilotte et al., 1988b; Crochet et al., 1989] and Alaric thrusts respectively [Ellenberger et al., 1987; Christophoul et al., 2003]. he WNW-ESE trending Lavelanet Anticline is the most prominent surface feature of the sub Pyrenean Zone (Figure 2.2). Boreholes (Dr3, Dr4) reveal a 1150 m diference in basement depth below the southern lank (Figure 2.4c), interpreted here as due to a south-verging thrust (Lavelanet Backthrust) [Baby, 1988]. Repetition of stratigraphy in borehole Bx1 implies the existence of two other blind, north-verging thrusts (Benaix Fault and Lavelanet hrust; Figure 2.4c). he Benaix Fault is a partially inverted normal fault delimiting a Lower Cretaceous rit basin that is equivalent to the Camarade basin to the west [Ford et al., 2016]. he North Pyrenean Frontal hrust marks an abrupt southward increase in deformation intensity (Figures 2.4, 2.5). he fault is here interpreted as an inverted normal fault with the deformed Nalzen- Fougax basins in its hanging wall. he Nalzen succession (S3; Figure 2.4) forms the vertical to overturned limb of a major north verging anticline cored by the Saint-Barthélémy basement (Figure 2.4c). We interpret this structure as a basement-cored forced fold, similar to Laramide structures with a larger fault ofset [e.g., Mitra and Mount, 1998]. Several hanging wall splays of the main North Pyrenean Frontal hrust cut the Nalzen block. his basement-cored forced fold transitions into a recumbent fold with no basement ~5 km to the east, on section S3b (Figure 2.5), where the Fougax block is thrust over the overturned frontal limb. Chapter 2. Case study of the Eastern Pyrenees 41

Elevation (km) Elevation (km) Elevation (km) 4 2 0 6 4 2 0 2 0 -2 -4 -6 -8 -10 -2 -4 -6 -2 -4 20 NNE NNE 0 0 MRL1 CSG Basement Upper Cretaceous Flysch 25 South of NPF: 10 10 T1 Distance from reference point (km) Orsans Thrust T2 CSG 20 20

Treziers Thrust Treziers

PG RG

d

CG

AVG PPG SSW 30 Black Dolomite Group Keuper Group Metamorphic Internal Zone T1 syn1 NNE CG Treziers Anticline Treziers Treziers ThrustTreziers Orsans Thrust 30 30 wedging Treziers Thrust Treziers Lavelanet Backthrust parallel 40 Distance from reference point (km) CG Stratigraphy LBT Dr1 Fig. 4c Fig. 4d Grey Flysch Group Black Flysch Group Mirande Limestone Group 40 40 Lavelanet Anticline LBT Dr3 LAVELANET Distance from reference point (km) Dr4 NPFT PG Dr4 RG Nalzen Block LT FICHERE PPG RG

AVG PG 50 50 AVG Bx1 45 LT BF

Lavelanet Thrust

Fig. 6 3MF BF BFG Aude Valley Group Aude Valley Plantaurel Group Petites Pyrénées Group RF PPG Saint- Barthélémy Massif MIZ 60 60 Distance from reference point (km) Nalzen Block 13.3 km of shortening RF NPF ? a GFG SSW BDG

50

6 4 2 0 KG -2 -4 -6 Elevation (km) Elevation

MG 3MF Saint- 70 Barthélémy Benaix Fault NPFT GFG 5 km ≥ Saint- Barthélémy Massif Carcassonne Group Coustouge Group Rieubach Group + ? BFG ? MIZ thrusting c b ? SSW Section S3

4 2 0

2 0

-2 -4 -6 -8 -2 -4 Elevation (km) Elevation Elevation (km) Elevation -10

Figure 2.4. Balanced (a) and restored (b) cross section S3, located on Figures 2.1 and 2.2. (c, d) Details of the Lavelanet and Treziers Anticline located in (a). Semi-transparent areas represent extrapolation above erosion level. White lines with blue letters indicate boreholes. MIZ, Metamorphic Internal Zone; NPF, North Pyrenean Fault; 3MF, 3M Fault; RF, Roquefeuil Fault; BF, Benaix Fault; NPFT, North Pyrenean Frontal hrust; LT, Lavelanet hrust; LBT, Lavelanet Backthrust. See Figure 5 for key to structural data. See text for details. 42 Chapter 2. Case study of the Eastern Pyrenees

Elevation (km) 2 0 -2 -4 0 N

AVG Bx1 NPFT PG PPG LT N

Nalzen Block GFG KG 5

BF BDG Block

bedding (borehole) cleavage Fougax

Block BFG MG BFG N Distance from reference point (km) bedding (field) bedding (map) Bedding Cleavage

Roquefeuil Frau Fault Frau 10

Massif 3MF(2)

3MF(1) Saint-Barthélémy c S

2 0 -2 -4 Elevation (km) Elevation

Elevation (km) Elevation (km) 2 0 6 4 2 0 -2 -4 -6 -8 -2 -4 -6 N 0 0 PG PG AVG AVG LT Fig. 5c LT PPG Frau Fault PPG BF BFG

Fig. 6 NPFT BFG 10 10 SB Mirande Limestone Group Black Dolomite Group Keuper Group Metamorphic Internal Zone Basement 3MF(2) Nalzen Block GFG NPF AZ Distance from reference point (km) Stratigraphy RF 13.2 km of shortening a RF 20 20 S Fougax Block

6 4 2 0 MG -2 -4 -6 Elevation (km) Elevation Distance from reference point (km)

SB BF 3MF Roquefeuil Block Aude Valley Group Aude Valley Rieubach Group Plantaurel Group Petites Pyrénées Group Grey Flysch Group Black Flysch Group 30 ? ? b Section S3b

2 0 -2 -4 -6 -8 Elevation (km) Elevation

Figure 2.5. Balanced (a) and restored (b) cross section S3b, located on Figures 2.1 and 2.2. (c) Details of (a). Semi-transparent areas represent extrapolation above erosion level. AZ, Axial Zone; SB, Saint- Barthélémy Massif; NPF, North Pyrenean Fault; 3MF, 3M Fault; 3MF(1), lat of 3M Fault; 3MF(2), hanging wall shortcut of 3M Fault; RF, Roquefeuil Fault; NPFT, North Pyrenean Frontal hrust; BF, Benaix Fault; LT, Lavelanet hrust. See text for details. In the Fougax block, Black Dolomite and Black Flysch strata are intensely folded and cleaved (Figure 2.5) [Marty, 1976]. he E-W trending, tight, upright folds show a short wavelength of ~750 to 850 m and relatively low amplitudes of ~500 to 600 m. Using methods described in Bulnes and Poblet [1999] we use the fold geometry to infer a depth to décollement of approximately 1 to 1.5 km. his suggests that the Fougax and Nalzen successions were detached from their basement (probably along Keuper evaporites [Marty, Chapter 2. Case study of the Eastern Pyrenees 43

W Distance from reference point (km) E 0S3 6 S3b 12 a 4 Nalzen Block 4

2 2

Saint- Elevation (km) Fougax Block 0 Barthélémy 0 Massif NPFT -2 -2 Elevation (km)

-4 -4

Benaix Fault -6 -6

S3 S3b 0S3 6 S3b 12 b 4 3MF(2) MIZ 4

3MF(1) Frau Fault MIZ Fougax Block 2 2 Roquefeuil

Block Elevation (km)

0 Saint-Barthélémy Massif 0

NPFT -2 -2 Elevation (km)

-4 -4

-6 -6

S3 S3b 0S3 6S3b 12 c 4 4

3MF 2 3MF(2) MIZ 2

3MF(1) Elevation (km) 0 0 Saint-Barthélémy Massif

-2 -2 NPFT Elevation (km)

-4 -4

-6 -6

S3 S3b Stratigraphy Petites Pyrénées Group Black Dolomite Group Grey Flysch Group Keuper Group Black Flysch Group Metamorphic Internal Zone Mirande Limestone Group Basement

Figure 2.6. Lateral sections illustrating the three-dimensional structure of the Frau area. (a, b, c) Northern, middle and southern sections in Figure 2.2, respectively. Semi-transparent areas represent extrapolation above erosion level. MIZ, Metamorphic Internal Zone; NPFT, North Pyrenean Frontal hrust; 3MF, 3M Fault; 3MF(1), lat of 3M Fault; 3MF(2), hanging wall shortcut of 3M Fault; BF, Benaix Fault. 44 Chapter 2. Case study of the Eastern Pyrenees

1976]), folded and thrust northward above the North Pyrenean Frontal hrust. Baby [1988] proposes that the Nalzen basin is immediately underlain by a basement high, implying the same for the Fougax Basin. Given the diference between the estimated depth of the Fougax decollement and the observed depth of basement in boreholes further north, this model requires a throw of 3 km on an unknown north vergent thrust. he two blind thrusts observed in borehole Bx1 do not accommodate suicient shortening to produce the required displacement. In addition, our interpretation of the Nalzen Block means it was cut from the footwall of the Roquefeuil Fault, thus Black Flysch Group strata should also be present at depth in the footwall of the North Pyrenean Frontal hrust. We therefore extrapolated the foreland basin succession in borehole Bx1 southward (~4 km) to underlie the allochthonous Fougax and Nalzen blocks and overlie a partially inverted rit basin that is equivalent to the Camarade basin to the west (Figures 2.4c, 2.5c) [Ford et al., 2016]. Folds of the Roquefeuil block have a wavelength of ~1.5 to ~2.5 km (Figures 2.2, 2.5c). he southward twofold increase in fold wavelength is abrupt, and is interpreted to indicate a change in depth to the basal thrust. Because Mirande Group limestones are clearly absent below the northern Fougax block, we propose that they are limited to the north by a blind inverted extensional basin margin fault (Roquefeuil Fault; Figure 2.5b) sealed by the Black Flysch Group. he Saint-Barthélémy Massif is a large block of Paleozoic metasediments and crystalline basement [de Saint Blanquat et al., 1990]. he top of basement drops eastward to a deeper level across an east-dipping lateral ramp to lie below the Metamorphic Internal Zone [Souquet and Peybernès, 1987] and the Roquefeuil block, as evidenced by the small tectonic windows around Camurac (Figures 2.2, 2.6). South of Montségur, a syncline of the Fougax block plunges steeply eastward (Figure 2.2) [Bilotte et al., 1988b] across the same lateral ramp. his lateral ramp is interpreted here as a reactivated Cretaceous extensional structure that may also be responsible for the westward thinning of the Lower Cretaceous sedimentary cover of the Saint- Barthélémy Massif. he massif has previously been interpreted as an allochthonous unit that originally formed a basement high south of the Nalzen block [Souquet and Peybernès, 1987; Baby, 1988; Baby et al., 1988]. Apatite ission track data from the adjacent Arize Massif imply that the massifs were buried before the Eocene [Vacherat et al., 2016]. Assuming a standard geothermal gradient of 30 °C/km and closure temperature of 100-120 °C [Reiners, 2005], the massifs were buried to at least 3-4 km by Mesozoic sedimentary cover (Black Flysch Group, Grey Flysch Group, and possibly Petites Pyrénées Group). his is consistent with our restoration that places the massif below the ~4 km thick Nalzen block, which requires less shortening than earlier restorations [Baby, 1988; Baby et al., 1988]. he amount of Pyrenean shortening accommodated in the Metamorphic Internal Zone and its displacement along the 3M Fault (Figures 2.2, 2.4, 2.5, 2.6) [Marty, 1976] cannot be quantiied due to intense metamorphism and internal deformation (mainly brecciation). he 3M Fault curves around the SE corner of the Saint-Barthélémy Massif and across the lateral ramp from a steep north-verging thrust along the massif’s southern boundary (Figures 2.4, 2.5) to a sub-horizontal orientation to the east in the Camurac area (3MF(1); Figures 2.2, 2.6b) [de Saint Blanquat et al., 2016]. he sub-horizontal 3M Fault terminates northward against a north dipping tectonic boundary (Frau Fault) with the Roquefeuil succession in its hanging wall (Figures 2.5c, 2.6b). he Frau Fault branches onto the 3M Fault to the west and is cut by a north-vergent hanging wall splay of the 3M Fault (3MF(2) to the east, Figures 2.2, 2.6). he North Pyrenean Fault juxtaposes the Metamorphic Internal Zone against the Axial Zone. Following interpretations of the ECORS deep seismic line [e.g., Muñoz, 1992] (Figure 2.1b), the North Pyrenean Fault terminates downward against the North Pyrenean Frontal hrust, which carries the Axial Zone on top of European basement (Figures 2.4a, 2.5a). 2.4.3 Syn-orogenic Subsidence of the North Pyrenean Foreland he unrited Pyrenean foreland records gentle post-rit subsidence only in its proximal zone (<0.03 mm/y in Dr4; Figure 2.7c, d; Table A.3). Syn-orogenic subsidence curves display a coherent trend across the foreland with two clear periods of subsidence separated by a quiescent phase (Figure 2.7c, d). Overall syn- Chapter 2. Case study of the Eastern Pyrenees 45

Southern Pyrenees Northern Pyrenees Time (Ma) Time (Ma) 100 90 80 70 60 50 40 30 100 90 80 70 60 50 40 30 a Total subsidence c Total subsidence sea level sea level 0 0

1 1 Depth (km)

2 2 Montserrat Depth (km) Santpedor Castellfollit 3 3 Puig-Reig Dr4 Jabalí T1 Gombrén syn1 4 4 Lower Pedraforca equivalent MRL1

b Tectonic subsidence d Tectonic subsidence Depth (km) 0 0

1 1 Depth (km)

2 2 post-rift Phase 1 Ph.2 Phase 3 Ph.4 post-rift Phase 1 Ph.2 Phase 3 Ph.4 Late Cretaceous Paleoc. Eocene Olig. Late Cretaceous Paleoc. Eocene Olig.

100 90 80 70 60 50 40 30 100 90 80 70 60 50 40 30 Time (Ma) Time (Ma)

Figure 2.7. Total subsidence and tectonic subsidence curves for the southern (a, b) and northern (b, c) Pyrenees. All depths are relative to present-day sea level. (a, b) Due to uncertain dating of strata, the solid line of the Lower Pedraforca equivalent shows Maastrichtian and Paleocene subsidence as maximum and minimum rates, respectively. he dashed line shows an average subsidence rate. Other south Pyrenean subsidence data from Vergés et al. [1998]. Detailed supporting data are available as Tables A.2, A.3, and A.4 in supplementary information. orogenic tectonic subsidence is characterised by very low rates with only 0.06 mm/y on average for Dr4, which records the fastest subsidence. A sharp increase in subsidence rate at ~84 Ma marks the onset of convergence, reaching a maximum of 0.14 mm/y during the early Campanian. Subsidence slowed down ater 75 Ma to 0.02 mm/y and 0.04 mm/y in boreholes Dr4 and T1, respectively. Little or no subsidence was recorded throughout the foreland from ~66 to ~59 Ma (quiescent phase). he subsidence shown in the synthetic borehole syn1 during the irst half of this phase is probably an artefact due to poor age control. Renewed subsidence started in the more distal basin during the hanetian (~59 Ma, Figure 2.7c, d) and is characterised by lower tectonic subsidence rates than during the Late Cretaceous (0.03 mm/y and 0.09 mm/y on average, respectively). hese long-term averages are based on tectonic subsidence of 1530 m in 18 My (borehole Dr4) and 244 m in 7.3 My (borehole T1). Borehole syn1 shows a gradual slowing of subsidence towards the end of the Eocene. 2.4.4 Evolution of the North Pyrenean Foreland he model presented here for the evolution of the North Pyrenean foreland in the Saint-Barthélémy – Fougax area is based on the restorations of the cross sections S3 and S3b (Figures 2.4, 2.5), foreland basin subsidence history, and the construction of three lateral cross sections linking the two present-day cross sections (Figure 2.6). he age boundaries of the evolutionary phases correspond to rounded boundaries of international chronostratigraphic ages due to limitations in the available stratigraphic dating [Cavaillé, 1976b; Bilotte et al., 1988c]. 46 Chapter 2. Case study of the Eastern Pyrenees

2.4.4.1 Early Cretaceous Rit and Late Cretaceous Post-rit Phases he mildly metamorphic (up to 200° C; [Gouache, 2017]) North Pyrenean Zone represents the European margin of the Apto-Cenomanian hyperextended rit system. Early Cretaceous extension was focussed mainly further south, where thinning of the crust exhumed mantle between the Iberian and European plates. he Metamorphic Internal Zone is thought to have overlain the zone of exhumed mantle [Lagabrielle et al., 2010; Mouthereau et al., 2014; Tugend et al., 2015; Ford et al., 2016]. he Roquefeuil and Benaix Faults (Figures 2.5b, 2.6b) represent the proximal northern margin of the rit system and accommodated ~1.5 km of N-S extension. he Roquefeuil Fault was active from Berriasian to Barremian. In the Aptian, the rit deepened and its margin migrated north to the Benaix Fault whose footwall was eroded to basement. A N-S lateral ramp delimited the Saint-Barthélémy high to the west, although the Mirande Group may still

Southern Pyrenees Northern Pyrenees Distance from reference point (km) Distance from reference point (km) 0 20 40 60 80 100 120 60 40 20 0 a Fault Activity c Fault Activity Oligoc. 30 Phase 4 30

40 8 7 6 5 40 Time (Ma) Eocene Phase 3 50 4 50 3 7 60 2 6 60 Paleoc. Phase 2 Time (Ma) Time 70 Late fault active 70 5 Phase 1 80 Cret. fault abandoned 1 1 2 3 4 80 b Migration d Migration Oligoc. 30 53 1 1 Phase 4 30 deformation shortening AZ 40 rate (mm/y) depocentre 40 Eocene front Phase 3 50 depocentre 50 Time (Ma) 60 60 Paleoc. deformation Phase 2 70 pinchout front 70 Phase 1

Time (Ma) Time 80 Late 80 Cret. pinchout 90 post-rift 90 subsidence 100 100 Early Rifting 110 Cret. 110 0 20 40 60 80 100 120 60 40 20 0 Distance from reference point (km) Distance from reference point (km) South Pyrenean structure names North Pyrenean structure names 1 Upper Pedraforca Thrust 5 Rialp Basement Thrust 1 North Pyrenean Frontal Thrust 5 Lavelanet Anticline 2 Lower Pedraforca Thrust 6 Puig-reig Anticline 2 Benaix Fault 6 Treziers Thrust 3 Vallfogona Thrust 7 Cardona Anticline 3 Lavelanet Thrust 7 Orsans Thrust 4 Orri Basement Thrust 8 Súria Anticline 4 Lavelanet Backthrust

Figure 2.8. (a, c) Reconstruction of the distribution of shortening through time in the southern Pyrenees (J3 section; Fig. 2.9) from Vergés et al. [1993] and northern Pyrenees (S3 section; Figures 2.4 and 2.5). Each line represents either the horizontal position of the leading edge of a principal fault or the position of the hinge zone at the surface of a major anticline. Displacement and fault length are assumed to increase linearly for each fault. For emergent faults, we assumed the fault reached maximum length during deposition of youngest cut strata. Fault timings for the southern Pyrenees fromfault-strata relationships [Vergés and Martinez, 1988; Vergés et al., 1995, 1998, 2002; Ford et al., 1997; Ramos et al., 2002; Costa et al., 2010; Carrigan et al., 2016]. (b, d) Horizontal position of the deformation front, the main depocentre(s) and foreland pinchout of the basin as a function of time in the southern and northern Pyrenees, respectively. he deformation front tracks the location of the most frontal structure through time. Depocentres were measured from balanced cross sections. Pinchout locations were determined from borehole and seismic data. Shortening rates through time are shown by grey bars, derived from the fault activity in (a, c). Shortening rates for the south include internal shortening in the Axial Zone (AZ). Dashed bars have a high uncertainty for the early convergence. Chapter 2. Case study of the Eastern Pyrenees 47 have covered it. Deep marine strata (Black Flysch Group) were deposited from the Aptian to Cenomanian, thinning westward across the lateral ramp. he post-rit basin progressively deepened as its margin migrated northward, onlapping unrited basement, from Middle Cenomanian to approximately Middle Turonian times, followed by progressive olap until end Santonian (Figure 2.8d). Following Ford et al. [2016] and Vacherat et al. [2016], we propose that the post-rit Grey Flysch extended south to cover the whole rit zone. 2.4.4.2 First Orogenic Phase: 84-66 Ma, Early Inversion During early convergence, the Metamorphic Internal Zone was emplaced northward along the 3M Fault. Gentle inversion afected inherited normal faults of the European margin and began folding Mesozoic basin successions (Nalzen, Fougax). Estimated shortening rates were 0.3-0.6 mm/y, the latter value including a minimum of shortening related to emplacement of the Metamorphic Internal Zone. Early inversion of the margin is not recorded by low temperature thermochronological data from the adjacent Arize Massif [Fitzgerald et al., 1999; Vacherat et al., 2016] suggesting that early basement exhumation was limited. A deep marine lexural basin was supplied with sediment mainly from the east [Bilotte, 1985]. Tectonic subsidence was relatively rapid though still only 0.09 mm/y close to the thrust front (Table 2.1). We relate this early subsidence to emplacement of the Metamorphic Internal Zone and early inversion and crustal thickening along the European margin. Campanian strata preserved in the Nalzen block indicate that the foreland basin continued south of the thrust front implying that the early orogenic ediice had low topographic relief and remained largely submarine. Shallower water facies in the Nalzen block suggest a syn-sedimentary uplit along the North Pyrenean Frontal hrust within the subsiding basin. he foreland basin records a deepening and then shallowing upward trend, ending in luvial conditions across the entire foreland. Aude Valley Group strata contain clasts of Upper Cretaceous limestones that were either sourced from the east, the south, or from local uplits [Bilotte et al., 1988c]. 2.4.4.3 Second Phase: 66-59 Ma, Northern Quiescence During the early Paleocene, no fault activity was recorded. Tectonic subsidence slowed to near zero in the foreland with deposition of only a thin unit of ine-grained continental deposits. his quiescence is also documented further west [Ford et al., 2016; Rougier et al., 2016], suggesting it was a widespread phase in the northern Pyrenees. 2.4.4.4 hird Phase: 59-34 Ma, Slow Northern Activity Tectonic activity returned in the hanetian with distributed thick-skinned shortening across the North Pyrenean foreland. he shortening rate was lower than in the Late Cretaceous, at 0.3 mm/y on average (Table 2.1). A slight acceleration in foreland basin subsidence was associated with a minor marine incursion across the proximal foreland (Rieubach Group; Figure 2.3b). Continental facies were then deposited across the whole foreland. More widespread deformation returned in the early Ypresian (~56 Ma), with activity on the Tréziers hrust and Lavelanet Anticline. A basin-wide marine incursion can be related to a eustatic sea level rise (Coustouge Group) [Christophoul et al., 2003]. Widespread continental sedimentation (Carcassonne Group) was re-established at the end of the Ypresian and continued until at least Rupelian times (Figure 2.3b). Post-hanetian activity of the North Pyrenean Frontal hrust until at least lower Ypresian is recorded north of the Nalzen block. Uplit and erosion of the Metamorphic Internal Zone, Saint-Barthélémy Massif, and/or the Axial Zone is recorded by pebbles of metamorphosed limestone and crystalline basement in Carcassonne Group conglomerates [Crochet, 1991]. Subsidence was slower during this phase at ~0.03 mm/y near the basin depocentre. he deformation front migrated northward, progressively activating basement-involved thrusts in the foreland (Tréziers and Orsans thrusts). he basin depocentre followed, and usually lay behind the deformation front (Figure 2.8d). he youngest evidence for deformation in the North Pyrenean foreland is slight folding of the upper Carcassonne Group above the Tréziers hrust, dated as approximately Priabonian (Figure 2.4d). 48 Chapter 2. Case study of the Eastern Pyrenees

2.4.4.5 Fourth Phase: 34-28 Ma, Final Abandonment Due to Miocene uplit of the Massif Central north of this part of the Pyrenean foreland basin, the Early Oligocene record is incomplete. We have no record of any fault activity during this phase, and it appears Berga Solsona 15 16 + ~23 km AZ internal shortening Pedraforca Section J3 Orri Elevation (km) 1 3 0 -3 -6 N 3 Orri 1 Rialp 69.2 km of shortening Montserrat Cardona evaporites Lower Pedraforca Upper 4 4 13 14 2 G 4 2 Upper Pedraforca Pf Milany Igualada G 5 11 12 Orri Lower 11 6 Pedraforca 4 9 8 Cadí J Thrust 12 Banyoles Tavertet Vallfogona 7 15 9 10 Rialp Cadí P-r 5 Anticline 11 Puig-reig 8 16 Beuda evaporites Bellmunt 6 14 7 8 NNW S 10 Anticline Cardona 7 Vallfogona thrust Vallfogona 9 Stratigraphic formations 12 Distance from reference point (km) Súria Anticline 15 S C 0 20 40 60 Puig-reig Anticline Armàncies and Campdevànol 5 6 16 Keuper salt Sagnari Cadí 3 4 10 14 M Distance from reference point (km) 13 Cardona Anticline -20 Catalan Coastal Ranges Súria Anticline a SSE

3 0 Tremp (Garumnian facies) Tremp Corones Penya -3 -6 Elevation (km) Elevation b 1 2 0 20 40 60 80 100 120

Figure 2.9. Balanced and restored cross section J3 in the south Pyrenean pro-foreland. Simpliied ater Vergés [1993]. Boreholes and stratigraphic sections used for subsidence calculations are indicated as semitransparent white boxes with blue letters. M, Montserrat; C, Castellfollit; S, Santpedor; P-r, Puig-reig; J, Jabalí; G, Gombrèn (split in two); Pf, Pedraforca. Formation names are indicated by numbers in circles as in Figure 2.3a. Chapter 2. Case study of the Eastern Pyrenees 49 that the Tréziers and Orsans thrusts were abandoned at the end of the preceding phase, as they are sealed by the upper Carcassonne Group.

5 South Pyrenean Foreland We performed a similar analysis of the South Pyrenean foreland along transect J3 (Figure 2.9) [Vergés, 1993]. his 120 km long balanced cross section stretches from the Axial Zone in the north to the Catalan Coastal Ranges in the south, traversing the eastern South Pyrenean Zone and Ebro Foreland Basin (Figure 2.1). he section was constrained by borehole data, surface geology, and seismic lines. A detailed description of this area can be found in Vergés et al. [1995, 1998, 2002] and only a summary is given here. Total shortening in the South Pyrenean foreland is 69.2 km, not accounting for internal deformation in the Axial Zone (~23 km) [Vergés et al., 1995] and Catalan Coastal Ranges (Figure 2.9). 2.5.1 Stratigraphy of the South Pyrenean Foreland Upper Triassic Keuper evaporites serve as a basal detachment for the Upper and Lower Pedraforca thrust sheets in the South Pyrenean Zone but is absent further south below the Ebro Basin, except close to the Catalan Coastal Ranges (Figure 2.9). A thin Liassic limestone and dolomite succession is only found in the Pedraforca thrust sheets. Cretaceous strata are mostly found in the Pedraforca thrust sheets and thin southward over a distance of 10 km (Figure 2.9b) [Vergés, 1993]. he Upper Pedraforca thrust sheet, which restores to the most northerly position, mainly comprises a ~1500 m thick inverted Lower Cretaceous extensional basin. he Lower Pedraforca thrust sheet contains ~1700 m of post-rit and syn-orogenic Upper Cretaceous conglomerates, marls and limestones, capped by ~700 m of Maastrichtian to Paleocene red beds (Garumnian facies) of the Tremp Formation (Figure 2.3a) [Vergés and Martinez, 1988; Vergés et al., 1994]. A regional unconformity marks the base of Eocene marine sedimentation in the South Pyrenean foreland [Pujalte et al., 1994; Serra-Kiel et al., 1994; Vergés et al., 1998]. hese marine sediments overlie the Tremp Formation in the north and onlap basement further south. hey record four transgressive-regressive cycles from Ypresian to Bartonian [Puigdefàbregas and Souquet, 1986; Martinez et al., 1989; Vergés et al., 1998]. Within each cycle, terrigenous clastic deposits were deposited in front of the advancing South Pyrenean hrust Front (Corones, Bellmunt and Milany Formations, Figure 2.3a). hese depositional units grade southward into moderate to deep-water marine carbonates and calcareous mudstones (Sagnari, Armáncies, Campdevánol and Banyoles Formations; Figure 2.3a) that, in turn, grade into shallow marine platform carbonates (Cadí, Penya and Tavertet Formations; Figure 2.3a) towards the southernmost border of the foreland basin. During the fourth cycle, the basin was characterised by prodelta ofshore marls (Igualada Formation; Figure 2.3a) bounded by northward prograding fan-delta deposits sourced from the upliting Catalan Coastal Ranges in the south (Montserrat Formation; Figure 2.3a) and southward prograding Milany Formation in the north. At the end of cycles two and four, thick evaporitic units are deposited in the central marine basin, the marine Beuda and transitional Cardona Formations, respectively (Figure 2.3a). In the footwall of the Vallfogona thrust, the Beuda evaporites reach 2000 m thickness, including a 200 m thick package of salt [Martinez et al., 1989]. he Ebro foreland basin became fully continental during the deposition of the uppermost part of the Cardona evaporitic Formation at ~36 Ma (middle Priabonian) [Costa et al., 2010]. Widespread non-marine deposition was then established until the latest Oligocene comprising the alluvial and luvial Solsona Formation and equivalent lacustrine deposits in the centre of the foreland basin (Figure 2.3a) [Meigs et al., 1996; García-Castellanos et al., 2003; Valero et al., 2014; Garcia-Castellanos and Larrasoaña, 2015]. 2.5.2 Structure of the South Pyrenean Foreland he South Pyrenean foreland fold-and-thrust belt, here called the South Pyrenean Zone, lies between the antiformal Axial Zone to the north and the mildly deformed Ebro foreland basin to the south (Figure 2.1b) [Vergés et al., 2002]. In the Eastern Pyrenees, three stacked thrust sheets compose the South Pyrenean Zone, namely from top (oldest) to bottom (youngest), the Upper Pedraforca, the Lower Pedraforca and the Cadí thrust sheets (Figure 2.9). he Upper Pedraforca and Lower Pedraforca thrust sheets are mostly constituted 50 Chapter 2. Case study of the Eastern Pyrenees by Jurassic to Palaeocene units detached above the Keuper evaporites. he Vallfogona thrust transports the Cadí thrust sheet with the Upper and Lower Pedraforca thrust sheets in its hanging wall. Following Vergés [1993], and Vergés et al. [1995, 2002], the Pedraforca thrust sheets root below the Nogueres basement thrust sheet of the Axial Zone whereas the Vallfogona thrust links back to the Orri basement thrust (Figure 2.1b). Laumonier [2015] alternatively suggests that the Pedraforca thrust sheets may root north of the North Pyrenean Fault, forming part of a ‘supra-axial unit’. his model depends on the interpretation of the eastern Axial Zone as a single block. he kinematics of the supra-axial unit proposed by Laumonier [2015] implicitly require coeval closure of the exhumed mantle domain and emplacement of the Lower Pedraforca thrust sheet, which is inconsistent with our data (see sections 5.4.2 and 6.2). he Puig-reig anticline within the Ebro foreland basin is the surface expression of a thrust ramp that steps up from the Beuda evaporite horizon to the Cardona salt horizon (Figure 2.9). his thrust links back into the Rialp basal thrust in the Axial Zone. Further south, the Cardona and Súria anticlines are detached and thrust along the Cardona evaporite horizon [e.g., Vergés et al., 1992]. he southern margin of the Ebro basin was the result of the thick-skinned inversion of the Tethyan Mesozoic basin of the Catalan Coastal Ranges from Lutetian to Chattian [Anadón, 1986; López-Blanco et al., 2000; Gómez-Paccard et al., 2012]. he southern and eastern borders of the Ebro foreland basin were regionally uplited during the Cenozoic extensional phase that opened the Western Mediterranean, thus tilting the entire southern foreland [e.g., Lewis et al., 2000]. 2.5.3 Subsidence of the South Pyrenean Foreland he four boreholes and two stratigraphic sections along section J3, analysed by Vergés et al. [1998], record the main Eocene basin history (Figures 2.7a, b, 2.9; Table A.4). Of those, only the Gombrèn stratigraphic section records Paleocene subsidence. To represent the Late Cretaceous and Paleocene subsidence record for the northernmost part of the foreland basin, a new subsidence curve was constructed for the Lower Pedraforca thrust sheet, based on data from Vergés et al. [1994]. Due to uncertain dating of strata, Maastrichtian and Paleocene subsidence are shown as maximum and minimum rates, respectively (Figure 2.7a, b). If these are assumed to represent true tectonic subsidence, the Lower Pedraforca thrust sheet appears to record very slow subsidence during the Paleocene (<0.01 mm/y; Figure 2.7b; Table A.4). However, the Gombrèn stratigraphic section, now located in the Cadí thrust sheet, records slightly faster tectonic subsidence in the Paleocene (0.05 mm/y; Figure 2.7b; Table A.4). Subsidence accelerates in the Eocene, forming the classic convex-up subsidence pattern expected for pro- foreland basins [Sinclair and Naylor, 2012]. Ypresian tectonic subsidence was high in the north of the Ebro foreland basin (Gombrèn section; 0.53 mm/y around 50 Ma; Table A.4), while rates were much lower to the south (0.03 to 0.07 mm/y; Table A.4). he onset of subsidence acceleration (>0.1 mm/y) migrated south at a rate of about 10 mm/y between 50.7 Ma and 40 Ma [Vergés et al., 1998]. his outward migration is also typical of pro-foreland basins [Sinclair and Naylor, 2012]. Increases in tectonic subsidence can be correlated with tectonic activity in the Axial Zone and South Pyrenean Zone (Figure 2.8a). For example, at 43 Ma tectonic subsidence of the Jabalí and Puig-reig boreholes increased to 0.26 and 0.14 mm/y respectively in response to loading by the Orri basement thrust sheet and emplacement of the Cadí thrust sheet [Vergés et al., 1998]. Tectonic subsidence in the south (Montserrat section) sees a brief acceleration to 0.21 mm/y at 41.5 Ma that can be correlated to activity of the Catalan Coastal Ranges [Vergés et al., 1998]. Ater ~36 Ma, tectonic subsidence slowed down throughout the southern foreland basin (Figure 2.7a, b) [Gómez-Paccard et al., 2012]. 2.5.4 Evolution of the South Pyrenean Foreland he summary of South Pyrenean foreland evolution given here is based on a compilation of subsidence and stratigraphic data, and a restored cross section [Vergés, 1993]. he boundaries of each evolutionary phase were chosen to relect signiicant changes in the whole orogenic system, and therefore do not necessarily correspond perfectly to distinct changes in the South Pyrenean foreland itself. Chapter 2. Case study of the Eastern Pyrenees 51

2.5.4.1 First and Second Phase: 84 – 59 Ma, Rit Inversion he timing of the onset of shortening in the southern Pyrenees cannot be distinguished in the preserved subsidence signal and is not clear from the rest of our data. Growth strata related to the Boixols thrust, carrying the western equivalent of the Upper Pedraforca thrust sheet, place the onset of shortening around ~72 Ma [Puigdefàbregas and Souquet, 1986; Bond and McClay, 1995]. However, many authors place the onset of shortening in the southern Pyrenees at Late Santonian, or ~84 Ma [Ardèvol et al., 2000; McClay et al., 2004; Saura et al., 2016]. In the western Pyrenees, Teixell [1996] also places the onset of convergence at Late Santonian. A Late Cretaceous lexural basin is preserved in the Lower Pedraforca thrust sheet (Figures 2.7a, b, 2.9b). We interpret this subsidence as a response to early inversion of the distal Iberian margin, during closure of the exhumed mantle domain. he Tremp Formation of latest Maastrichtian to Upper Paleocene age seals the Upper Pedraforca thrust sheet [Vergés and Martinez, 1988; Vergés et al., 1995], however, tectonic subsidence appears to have continued very slowly throughout the Paleocene (Figure 2.7b; Table 2.1). herefore, while we cannot rule out a Paleocene quiescence in the southern Pyrenees, sparse data appear to indicate slow shortening during the Paleocene. he irst and second orogenic phases were characterised by inversion of rit structures. Although poorly constrained, we estimate ~10 km of shortening were accommodated between ~84 and ~56 Ma, yielding a rate of 0.4 mm/y during the Late Cretaceous and Paleocene (Figure 2.8b; Table 2.1). 2.5.4.2 hird Phase: 59 – 34 Ma, Major Convergence During early Eocene, shortening in the South Pyrenean foreland reached a maximum rate of 4.0 mm/y (Figure 2.8b). South of the rapidly advancing thrust front this was accompanied by an increase in tectonic subsidence (0.53 mm/y maximum, Gombrèn; Figure 2.7b; Table A.4), resulting in a deep marine basin (Figure 2.3a). Meanwhile, further south, tectonic subsidence rates and sediment supply were much lower (around 0.04 mm/y), allowing the growth of carbonate platforms. Emplacement of the Lower Pedraforca thrust sheet occurred during early Eocene until ~47 Ma, as shown by syn-orogenic strata [Solé Sugrañes and Clavell, 1973; Vergés and Martinez, 1988; Burbank et al., 1992b].

Table 2.1. Tectonic subsidence rates and shortening rates in the Eastern Pyreneesa Tectonic subsidence rate (mm/y) Shortening rate (mm/y) Orogenic phase Southb Northc Southb Northc 84-66 Ma 0.06 0.09 0.4 0.6 66-59 Ma 0.05 0.00 0.4 0 59-34 Ma 0.12 0.03 2.7 0.3 34-28 Ma - 0.01 2.2 0 aAll rates were averaged over their respective phases. bCombines Axial Zone, South Pyrenean Zone, and Ebro foreland basin. cCombines North Pyrenean Zone, Sub-Pyrenean Zone, and Aquitaine foreland basin.

Shortening slowed down signiicantly ater the end of the Ypresian to 1.0 mm/y (Figure 2.8b). A new basement thrust sheet (Orri) became active, as constrained by the Vallfogona thrust, marking the start of true thick-skinned frontal accretion (Figure 2.8a). Crustal thickening and exhumation in the Axial Zone started at ~50 Ma, recorded by low-temperature thermochronology data [Fitzgerald et al., 1999; Sinclair et al., 2005; Metcalf et al., 2009; Rahl et al., 2011; Whitchurch et al., 2011]. Axial Zone deformation accommodated an average of 1.1 mm/y of extra shortening until the end of convergence (~23 km since 50 Ma; Figure 2.8b) [Vergés et al., 1995]. his was accompanied by an increase in tectonic subsidence rates throughout the foreland basin to around 0.10 mm/y (Figure 2.7b). he Gombrèn subsidence record in the proximal foreland shows a much lower rate starting in the late Ypresian owing to deformation ahead of a propagating Vallfogona thrust, which reached the surface in the mid-Lutetian [Vergés et al., 1998; Ramos et al., 2002]. . Widespread growth strata within the Berga conglomerates (Figs, 2.3, 2.9) constrain the latest activity of the Vallfogona thrust to at least ~31 Ma (Figure 2.8a) [Ford et al., 1997; Vergés et al., 1998; Carrigan et al., 2016]. Most of the foreland basin remained marine to brackish, but sedimentation 52 Chapter 2. Case study of the Eastern Pyrenees

Eastern Pyrenees S Pro-forelandAxial Zone NPF Retro-foreland N -69 km -23 km -19 km Puig-Reig Cardona a 28-0 Ma Súria km D N 0 O SB 111 km R 10 20 30 Subduction (Pyrope) Moho (Pyrope) --...... 40 ...... 50 -62 km -17 km -19 km Lower Pedraforca Vallfogona NPFT km b 34 Ma LavelanetTreziers AnticlineOrsans D N 0 O SB 98 km R 10 20 30 40 50

-9 km -0 km -11 km Upper Pedraforca c 59 Ma km D 0 20 10 km R O N SB 20 30 40

-7 km -0 km -11 km Upper Pedraforca NPFT km d 66 Ma 3MF D 0 ? km 18 SB 10 km R O N 20 30

Phase 1 Phase 2 Phase 3 Phase 4 40

S Iberia (undeformed) Iberian margin Exhumed mantle Europe N 127 km 82 km ? km 80 km km e 84 Ma ｾ＠ D 0 R - --O N - - " SB// 10 7 20 ｾ＠ 30 50 km ... -- - 40 D Phase 4: Upp. Eocene - Oligocene D Phase 1: Upper Cretaceous Metamorphic Internal Zone D Phase 3: Low. - Mid. Eocene D post-rift •D basement (upper crust) D Phase 2: Paleocene • syn-rift D lower crust Figure 2.10. Balanced stepwise evolution of the Eastern Pyrenees. Shortening amounts are cumulative. Red lines indicate faults that have been active since the preceding step. hick black lines indicate abandoned faults. Dashed lines indicate faults that may have been inherited but have not yet been reactivated. (a) Moho and subduction contact are those along the ECORS line ~50 km to the west, ater Chevrot et al. [2015], projected to align along the North Pyrenean Fault. R, Rialp; O, Orri; N, Nogueres; SB, Saint-Barthélémy Massif; NPF, North Pyrenean Fault; 3MF, 3M Fault; NPFT, North Pyrenean Frontal hrust. Chapter 2. Case study of the Eastern Pyrenees 53 along the thrust front was continental from Lutetian onwards (Figure 2.3a). he Cardona salt horizon was deposited around 36 Ma as a result of the isolation from the Atlantic Ocean ater uplit of the Western Pyrenees [Burbank et al., 1992b; Vergés and Burbank, 1996; Costa et al., 2010]. Growth strata on the lanks of the Puig-reig anticline date the onset of folding to around the same time [Vergés et al., 2002]. his was related to the Rialp basement thrust and increased the shortening rate to 1.6 mm/y (Figure 2.8b). 2.5.4.3 Fourth Phase: 34 – 28 Ma, Final Deformation During this inal phase, the shortening rate in the foreland decreased to 0.6 mm/y (Figure 2.8b). Most shortening in the foreland was accommodated by detachment folding above the Cardona evaporites and the inal activity of the Vallfogona thrust [Vergés et al., 1992; Carrigan et al., 2016]. he deformation front thus rapidly migrated to the south along the Cardona evaporite horizon, while the depocentre remained approximately stationary (Figure 2.8b). We have no subsidence records for this phase, but sedimentation in the foreland remained continental (Solsona Formation and equivalents; Figure 2.3a). In this part of the southern foreland, the youngest preserved sediments (Rupelian) are deformed, suggesting deformation continued until at least ~28 Ma. Shortening in the Central Pyrenees is estimated to have ended around 30 Ma based on Apatite ission track data from the Axial Zone [Fitzgerald et al., 1999], or ~24.7 Ma based on magnetostratigraphic dating of the South Central Unit and its footwall [Meigs et al., 1996].

2.6 Discussion 2.6.1 Crustal-scale Model of Orogenic Evolution Geodynamic models for the Pyrenees have evolved over the last 30 years, as have estimates of total N-S shortening. Most crustal-scale restorations have focussed on the Central Pyrenees along the Ariège ECORS deep seismic line. Using upper crustal imbricates in the Axial Zone and restoring to normal crustal thickness, Muñoz [1992] proposes 147 km of N-S shortening while Beaumont et al. [2000] prefer 165 km over several other possible geometries. More recent reconstructions restore to the pre-orogenic rit system [e.g., Mouthereau et al., 2014; Teixell et al., 2016], most notably integrating the hypothesis that the Iberian and European plates were separated by a zone of exhumed mantle due to Aptian-Cenomanian hyper-extension [e.g., Jammes et al., 2009; Lagabrielle et al., 2010]. However, the original width of this zone and hence the convergence necessary to subsequently close it are unknown. Proposed widths vary from 15 km in the Western Pyrenees [Teixell et al., 2016] to 50 km of exhumed mantle in the Central Pyrenees [Mouthereau et al., 2014] or several hundred km of oceanic crust [Vissers and Meijer, 2012b]. However, deep seismic tomography imaging below the Pyrenees rules out oceanic subduction prior to collision, as no subducted oceanic slab can be identiied in the mantle [Chevrot et al., 2014]. Using full crustal imbrication, as proposed by Roure et al. [1989], Mouthereau et al. [2014] obtain 92 km of total shortening along the ECORS line, excluding closure of the exhumed mantle domain. his markedly lower value is principally related to a revised shortening estimate for the southern Pyrenees compared to Beaumont et al. [2000]. In the western Pyrenees along ECORS-Arzacq, Teixell [1996, 1998] estimates a total shortening of about 80 km while more recently, Teixell et al. [2016] propose 99 km along a section 15 km to the east of ECORS-Arzacq, excluding closure of the exhumed mantle domain. In the Eastern Pyrenees, Vergés et al. [1995] integrate surface geology with the deep structure of the ECORS line projected onto the same section line as this paper. By using the work of Baby et al. [1988] for the northern Pyrenees, assuming imbrication of Iberian upper crust only and subduction of Iberian lower crust, restoring to normal crustal thickness, and integrating estimates of erosion and sedimentation during main orogenic phases, they ind a minimum total shortening of 125 km. Vergés and García-Senz [2001] revise the restoration (but not the estimated shortening) by proposing two possible geometries for the rited crust. We further update this reconstruction by integrating the detailed deformation history presented above (Figure 2.8) with subsidence (Figure 2.7), thermochronology [e.g., Metcalf et al., 2009], and deep geophysical data [e.g., Chevrot et al., 2015]. his allowed us to build a new and more detailed evolutionary model for the Eastern Pyrenees in four phases, chosen to relect signiicant changes in the 54 Chapter 2. Case study of the Eastern Pyrenees

dynamics of the orogenic system (Figure 2.10). he main diferences compared to Vergés et al. [1995, 2002] and Vergés and García-Senz [2001] are summarised in the following paragraphs. First, the deep structure of the present day section integrates major lithospheric boundaries identiied by recent P wave tomography and migration of converted P waves along the ECORS line [Chevrot et al., 2014, 2015]. Chevrot et al. [2015] identify the interface between the top of the Iberian crust and the European mantle, dipping 20° N. heir Iberian Moho (at 32 km depth) corresponds well with ECORS data, but when projected to our section and aligned along the North Pyrenean Fault, the Moho inlexion to dip north at 20° is positioned below the Puig-reig anticline (Figure 2.10a). his Moho geometry leads to an overthickened Iberian crust (>50 km) below the Axial Zone, which can be neither isostatically justiied nor volumetrically balanced. We therefore continue the Iberian lower crust northward to an inlexion point below the southern Axial Zone (as imaged on the ECORS proile), where it dips 20°N into the mantle. his creates a slight mismatch between the projected Moho and our model, a result of the slightly narrower Axial Zone along our section. he apparent southward thinning of the European crust observed by Chevrot et al. [2015] is integrated into our model. Second, in line with recent indings, the composite fully restored section (end Santonian; Figure 2.10e) shows the rited margins separated by a zone of exhumed mantle with a hypothetical width of 50 km. he basins are illed with a syn-rit to post-rit succession that is ~4 km thick above mantle. Following Lagabrielle et al. [2010], we position the metamorphosed syn-rit succession (future Metamorphic Internal Zone) directly above mantle. he geometry of the pre-orogenic extensional fault system is still a matter of much debate as discussed above. Restored geometries on the Iberian plate follow those of Vergés and García-Senz [2001] and Vergés et al. [2002] while the European margin is based on new work in this paper. he inferred geometries of the Iberian and European rited margins are constrained by the observed remnants of extensional Mesozoic basins, preserve the area needed for a precise crustal balancing through time, are isostatically stable, and are guided by new seismic interpretations of the Atlantic Iberian margins [e.g., Sutra et al., 2013]. hird, the lower estimate of shortening for the European foreland (~20 km) leads to a new minimum overall shortening of 111 km. All estimations of shortening given here exclude closure of the exhumed mantle domain. We cannot constrain the original volume of the Metamorphic Internal Zone nor its tectonic and erosional history. he largest uncertainties associated with the amount and distribution of shortening are therefore for the irst phase. 2.6.1.1 First Phase: 84-66 Ma, Rit Inversion During the Late Cretaceous (Figure 2.10e to d) slow and distributed shortening (~18 km) was accommodated by inversion of normal faults across the whole rit system while the mantle domain closed. An overall shortening rate of ~1 mm/y is estimated with deformation distributed roughly equally across the two margins (Figure 2.8b, d; Table 2.1). As this estimate excludes closure of the mantle zone, true convergence rates would be considerably higher. For example, assuming a 50 km wide mantle zone implies a convergence rate of 3.8 mm/y. he Metamorphic Internal Zone, of unknown dimensions, was thrust northward and probably also southward if we assume that it covered the whole zone of exhumed mantle (Figure 2.10d) [Mouthereau et al., 2014; Teixell et al., 2016]. Marine lexural foreland basins developed on both plates and may have formed a continuous basin across the submarine, low relief orogen. Depocentres became overilled and continental in the Maastrichtian (Figure 2.3). Average tectonic subsidence rates are 0.09 mm/y and 0.06 mm/y for the north and south, respectively. Low temperature thermochronological data to the east suggest the onset of uplit of the North Pyrenean Massifs during this early phase [Ternois et al., 2017]. 2.6.1.2 Second Phase: 66-59 Ma, Slowdown and Northern Quiescence he total shortening rate slowed to ~0.4 mm/y during the Danian and Selandian (Figure 2.10c). Subduction polarity was established during this phase, the asymmetry of which resulted in a temporary stop in foreland subsidence and deformation of the upper plate (Figures 2.7d, 2.8d; Tables 2.1, A.3). However, very slow Chapter 2. Case study of the Eastern Pyrenees 55 tectonic subsidence (0.05 mm/y) and shortening (0.4 mm/y) continued on the lower plate, accommodating some 2 km of shortening. Continental sedimentation prevailed in both foreland basins. 2.6.1.3 hird Phase: 59-34 Ma, Main Collision and Southern Dominance he principal phase of collision (approximately 78 km of shortening) took place between ~59 and ~34 Ma with shortening and tectonic subsidence rates revealing a strong asymmetry, being much higher in the south (2.7 mm/y and 0.12 mm/y on average, respectively) than in the north (0.3 mm/y and 0.03 mm/y on average, respectively; Figures 2.7, 2.8; Table 2.1). he peak in overall shortening rate (~4.5 mm/y) began at ~56 Ma and was followed by a gradual deceleration to 2.3 mm/y until ~36 Ma, when accretion of a new basement thrust sheet (Rialp) increased the overall rate to 2.9 mm/y (Figures 2.8a, A1c). he South Pyrenean wedge accommodated ~70 km of shortening by frontal accretion and thickening of the Axial Zone. In the north, shortening was accommodated along several thrusts simultaneously, with the North Pyrenean Frontal hrust accounting for the majority of shortening (Figure 2.8c). Ater the early Ypresian marine incursion, continental conditions were quickly re-established in the northern foreland while the south sustained a carbonate platform and deep marine basin until ~36 Ma (Figure 2.3) [Vergés et al., 1998; Costa et al., 2010]. Low-temperature thermochronology records uplit and exhumation of the Axial Zone from ~50 Ma [e.g., Metcalf et al., 2009], which is corroborated by alluvial sedimentation along the thrust fronts of both wedges. 2.6.1.4 Fourth Phase: 34-28 Ma, Final Shortening In the inal stage, ater ~34 Ma, orogenic processes gradually died out. he eastern northern foreland appears to have been abandoned at around 34 Ma but this is unclear due to later uplit and erosion (Fig 2.8c; Table 2.1). In contrast, during the Oligocene ~13 km of shortening was accommodated in the south by detachment anticlines above the Cardona salt horizon (~7 km) and by internal deformation and uplit of the Axial Zone (~6 km; Figures 2.8a, b, 2.10a). While the exact timing of the end of convergence is not well known, we estimate an average shortening rate of ~2.2 mm/y (Table 2.1) up to at least ~28 Ma. Post- orogenic exhumation of the southern Axial Zone continued into the Miocene [Rushlow et al., 2013]. 2.6.2 Improved detail in orogen deformation history Crustal deformation histories previously proposed for the Pyrenees are similar in broad terms to our results, with inversion of the rit followed by pro-wedge propagation and basement stacking in the Axial Zone [e.g., Sinclair et al., 2005]. However, along our section in the Eastern Pyrenees, the distribution of displacement on faults through time (Figure 2.8) has revealed a more detailed deformation history. While recent models include the closure of an exhumed mantle domain, how this event relates to inversion of the rited margins remains diicult to constrain. Mouthereau et al. [2014] show the mantle domain closing irst, followed by thick-skinned inversion of rit structures ater ~75-70 Ma. We propose instead that closure of the mantle domain took place progressively and simultaneously with thick-skinned inversion of the rited margins, based on timing of the North Pyrenean Frontal hrust as constrained by stratigraphy. his is corroborated by the foreland tectonic subsidence signal that we link to emplacement of the Metamorphic Internal Zone and loading of the margins. None of the published crustal-scale restorations recognise the Paleocene slowdown described in this paper, either because it is below the temporal resolution of the sequential restoration steps used [Muñoz, 1992; Beaumont et al., 2000; Teixell et al., 2016], or because it cannot be detected by time-temperature models from apatite ission track data used to constrain vertical motions [Mouthereau et al., 2014]. Furthermore, previous attempts at estimating convergence rates through time have given conlicting results. In the Central Pyrenees, Beaumont et al. [2000] show an overall acceleration, in particular at 36 Ma, related to late stacking of crustal nappes in the Axial Zone. In contrast, Mouthereau et al. [2014] propose a high initial rate of convergence (~3 mm/y) in the Late Cretaceous and Early Paleocene, followed by gradual deceleration until the Miocene. For the Western Pyrenees, Teixell et al. [2016] estimate very low rates (<1 mm/y) during the Late Cretaceous and Paleocene, followed by the principal peak of convergence in the Eocene (3-2.5 mm/y). Along our section in the Eastern Pyrenees, we show three peaks of relatively rapid 56 Chapter 2. Case study of the Eastern Pyrenees

shortening (83, 56, and 36 Ma), each followed by a period of slower and decelerating shortening (Figures 2.8b, d, A1; Table 2.1). We discuss the main diferences in the following paragraphs. he highest shortening rates in the Eastern Pyrenees, according to our reconstruction, occurred during the early Eocene, which can be considered as the onset of main collision. his coincides with relatively rapid tectonic subsidence of the Aquitaine Basin starting at ~55 Ma (Figure 2.7) [Desegaulx and Brunet, 1990; Ford et al., 2016; Rougier et al., 2016], and the onset of crustal stacking in the Axial Zone around 50 Ma [e.g., Metcalf et al., 2009]. Teixell et al. [2016] show that the main pulse of shortening in the Western Pyrenees also occurred during the Eocene. In the Central Pyrenees however, neither Beaumont et al. [2000] nor Mouthereau et al. [2014] place the highest shortening rates during the Eocene. he initial high shortening rate of Mouthereau et al. [2014] is based on Late Cretaceous to Paleocene cooling detected by detrital low- temperature thermochronology, from foreland basin sediments that the authors assume were derived from the Pyrenean orogen. However, paleocurrent and facies analyses [Bilotte, 1985] show that these sediments were sourced from an area to the east of the Pyrenees. herefore, these data do not constrain exhumation rate in the Pyrenees along the ECORS line. Furthermore, the gradually decelerating convergence proposed by Mouthereau et al. [2014] cannot be reconciled with the increased Eocene tectonic subsidence and shortening rates for the forelands. Beaumont et al. [2000] place the highest shortening rates ater 36 Ma, in accordance with rapid exhumation in the Axial Zone, as shown by numerous thermochronological studies [Morris et al., 1998; Fitzgerald et al., 1999; Jolivet et al., 2007; Maurel et al., 2008; Metcalf et al., 2009; Fillon and van der Beek, 2012; Rushlow et al., 2013]. However, thrusting along the eastern border of the South Central Unit (~25 km east of ECORS; Figure 2.1a) appears to have slowed down ater 36 Ma [Burbank et al., 1992a]. Furthermore, foreland tectonic subsidence slows down around 36 Ma in both foreland basins. his is well constrained in the south by magnetostratigraphic dating (Figure 2.7b) [Vergés et al., 1998]. However, the inal deceleration is less well constrained in the north as rates were already very low and the youngest strata are poorly dated and preserved (Figure 2.7d) [Ford et al., 2016]. Although the peak in exhumation rate of the Axial Zone may have occurred around 36 Ma, there is evidence for exhumation of the Axial Zone since ~50 Ma [Fitzgerald et al., 1999; Sinclair et al., 2005; Metcalf et al., 2009; Rahl et al., 2011; Whitchurch et al., 2011]. Because exhumation rate cannot be directly related to shortening rate, we assume a constant shortening rate for internal shortening of the Axial Zone since ~50 Ma, thus yielding a lower shortening rate ater 36 Ma than shown by Beaumont et al. [2000] and providing a better it with foreland evolution. 2.6.3 Linked Pro- and Retro-wedge Dynamics As described in section 2.6.1, Late Cretaceous shortening was distributed across both the Iberian and European rited margins, before the convergent system evolved into an asymmetrical, pro-wedge-dominant orogen (Figure 2.8b, d). Models of doubly vergent orogens with distinct lower and upper plates typically show a pro-wedge-dominant strain partitioning [Beaumont et al., 2000; Sinclair et al., 2005; Hoth et al., 2007; Cruz et al., 2008; Duerto and McClay, 2009; Hardy et al., 2009; Jammes and Huismans, 2012; Erdős et al., 2014, 2015; Jammes et al., 2014]. However at the onset of Pyrenean convergence, no such distinction can be made. Instead, the template comprises two opposing rited margins separated by exhumed mantle [Jammes et al., 2009, 2014; Lagabrielle et al., 2010; Mouthereau et al., 2014; Clerc et al., 2015; Vacherat et al., 2016]. Numerical models that include rited margins show early inversion that is roughly symmetrical, until full collision and onset of subduction of continental mantle lithosphere forces a change to an asymmetric geometry at depth [Jammes and Huismans, 2012; Erdős et al., 2014; Jammes et al., 2014]. Whether this change to an asymmetric deep geometry in the Pyrenees occurred during closure of the exhumed mantle domain or at the beginning of main collision is unclear. Our model suggests that the transition to a pro- wedge-dominant strain partitioning and related abandonment of the retro-wedge around 66 Ma occurred shortly ater closure of the exhumed mantle domain. he change in strain partitioning may thus record the change from early rit inversion to main collision (i.e. continental subduction) in the Pyrenees, implying this major change does not necessarily rely on external forcing. In the Western Pyrenees, Teixell et al. [2016] infer that closure of the mantle domain did not occur until the Early Eocene. his suggests that Chapter 2. Case study of the Eastern Pyrenees 57 early convergence was less signiicant in the west than in the east, and/or that the mantle domain was wider in the west than in the east as proposed by Jammes et al. [2009]. he Paleocene retro-foreland quiescence must be explained on a regional scale, as it has also been recognised in the Central Pyrenees [Ford et al., 2016; Rougier et al., 2016]. One potential cause for this north Pyrenean quiescence is a pause in plate convergence. While some plate reconstructions based on magnetic sea loor anomalies show a complete stagnation of convergence in the Paleocene [Roest and Srivastava, 1991; Rosenbaum et al., 2002], others show continuous convergence [Vissers and Meijer, 2012a]. Other potential contributing factors for the north Pyrenean quiescence can be related to redistribution of shortening within the surrounding plates [e.g., Mouthereau et al., 2014], or a change in the deep structure of the orogen as it evolves to full collision, as discussed below. Late Cretaceous subsidence in both the pro- and retro-forelands was mainly driven by loading related to emplacement of the Metamorphic Internal Zone and onset of inversion of the rited margins. hanetian to Oligocene pro-foreland behaviour can be directly related to crustal thickening and loading during main collision. We here propose that contemporaneous retro-foreland subsidence was driven by backthrusting of the thickened pro-wedge onto the European plate, similar to predictions from numerical and analogue models [e.g., Hoth et al., 2007; Erdős et al., 2014]. he deep structure imaged by the ECORS line indicates that the northern part of the Axial Zone overlies the European plate (Figure 2.1b) [Roure et al., 1989; Muñoz, 1992; Teixell, 1998; Teixell et al., 2016]. hermochronological data generally place the onset of exhumation in the Axial Zone at ~50 Ma [Fitzgerald et al., 1999; Sinclair et al., 2005; Metcalf et al., 2009; Rahl et al., 2011; Whitchurch et al., 2011], but in the eastern Pyrenees this may have started as early as ~55- 60 Ma [Maurel et al., 2008]. Tectonic subsidence of the retro-foreland resumed around 59 Ma, suggesting that the period between the onset of pro-dominant strain partitioning (~66 Ma) and the onset of Axial Zone backthrusting (~60 Ma at the earliest) may have contributed to the Paleocene quiescence observed in the retro-foreland. Using subsidence analysis and detailed tectonic reconstructions to compare the evolution on both sides of a natural system has given insight into crustal-scale dynamics of doubly vergent orogens. Changes in the strain distribution over time may be intrinsic to the system and independent of external factors. So far, these changes have not been suggested by models [e.g. Sinclair et al., 2005; Hoth et al., 2007]. Previous foreland basin subsidence models simplify the orogen into a uniformly growing double wedge and therefore do not predict multiple periods of retro-foreland subsidence, nor a quiescent phase [Sinclair et al., 2005; Naylor and Sinclair, 2008; Sinclair and Naylor, 2012].

2.7 Conclusions We present new stratigraphic and tectonic data for the north Pyrenean fold-and-thrust belt and Aquitaine Basin in the Eastern Pyrenees, which, when integrated with published data, provide new insight into retro- wedge evolution. By integrating these new results with equivalent data on a section through the southern Pyrenees and other published data (thermochronology, seismic tomography data) we derive an updated full crustal history of the East Pyrenean double-wedge orogen. Calculated minimum shortening for the East Pyrenean retro-foreland fold-and-thrust belt and Aquitaine Basin was ~20 km during the Pyrenean orogeny. Late Cretaceous shortening and tectonic subsidence rates in the northern Pyrenees (0.6 mm/y and 0.09 mm/y, respectively) were followed by a Paleocene quiescence and lower rates during the Eocene (0.3 mm/y and 0.03 mm/y). A minimum overall N-S shortening of ~111 km was estimated for the Eastern Pyrenees, not including closure of an inferred exhumed mantle domain of unknown width. he evolution of the double wedge orogen is divided into four phases: (1) Late Cretaceous shortening was characterised by closure of the exhumed mantle domain and inversion of the Iberian and European margins. An overall shortening rate of ~1 mm/y on average was distributed roughly equally between the two margins. Signiicant early foreland basins developed on both margins well before the creation of signiicant 58 Chapter 2. Case study of the Eastern Pyrenees

topography. (2) In the Paleocene slow shortening continued in the southern Pyrenees (~0.4 mm/y on average) while the north was temporarily inactive. (3) he highest overall shortening rate was reached in the Eocene (~3.1 mm/y on average) recording main collision. his was predominantly accommodated by shortening in the southern Pyrenees and Axial Zone. However, continued thickening of the Axial Zone then led to backthrusting onto the upper plate, reactivating the retro-wedge and its foreland basin. (4) Overall shortening slowed in the Oligocene to ~2.1 mm/y on average. he northern Pyrenees were probably inactive in this phase. According to our revised evolutionary model for the Eastern Pyrenees, the evolution from rit inversion to main collision caused a change from equal shortening in both rited margins to a pro-wedge dominant strain distribution, accompanied by a temporary quiescence of the retro-wedge. Comparative analysis of pro- and retro-wedge behaviour reveals the self-organisation of a complex orogenic system. Major changes or events (change in strain distribution) may be due to internal intrinsic thresholds being reached (e.g. onset of continental collision) and therefore may not rely on external forcing (plate kinematics). We believe that the application of the same detailed approach in other orogens can potentially identify similar self-organisation and intrinsic thresholds involved in orogenesis.

Acknowledgements his study was funded by the ANR (France) PYRAMID research project. he French-Norwegian Foundation (13-06 PYR-FFTP; Sedimentary basin and North Pyrenean foreland fold and thrust belt formation) supported study visits to the University of Bergen, Norway. Collaboration with CSIC Barcelona, Spain was funded by the project ALPIMED (PIE-CSIC-201530E082). A stratigraphic correlation of the BRGM 1:50,000 geological maps supporting Figure 2.3 is available as Table A.1 in the supplementary information. Detailed subsidence data supporting Figure 2.7 are available as Tables A.2, A.3, and A.4 in the supplementary information. Field data used to constrain the structure in the northern Pyrenees (Figures 2.2, 2.4, 2.5 and 2.6) are available as Table A.5 in the supplementary information. We thank the PYRAMID team, in particular M. de Saint Blanquat for fruitful discussions on the structure and nature of the Metamorphic Internal Zone in the Saint-Barthélémy area. We thank Andrew Leier and one anonymous reviewer for their insightful comments that helped to signiicantly improve the manuscript.

References Anadón, P. (1986), Las facies lacustres del olioceno de campins (Vallès Oriental, provincia de Barcelona), Cuad. Geol. Ibérica, 10, 271–294. Angrand, P., M. Ford and A.B. Watts (2018). Subsidence origin, lateral variation in foreland evolution, and lexure of a rited continental margin: the Aquitaine foreland basin (SW France), Tectonics, 37(2), 430–449, doi:10.1002/2017TC004670 Ardèvol, L., J. Klimowitz, J. Malagón, and P. J. C. Nagtegaal (2000), Depositional sequence response to foreland deformation in the upper Cretaceous of the Southern Pyrenees, Spain, Am. Assoc. Pet. Geol. Bull., 84(4), 566–588, doi:10.1306/C9EBCE55-1735-11D7-8645000102C1865D. Baby, P. (1988), Chevauchements dans une zone à structure complexe: La zone nord-pyrénéenne ariégeoise, Université Paul Sabatier Toulouse III, Toulouse, France. Baby, P., G. Crouzet, M. Specht, J. Deramond, M. Bilotte, and E.-J. Debroas (1988), Rôle des paléostructures albo- cénomaniennes dans la géométrie des chevauchements fronaux nord-pyrénéens, C. R. Acad. Sci. Paris, 306(2), 307–313. Beaumont, C., J. A. Muñoz, J. Hamilton, and P. Fullsack (2000), Factors controlling the Alpine evolution of the central Pyrenees inferred from a comparison of observations and geodynamical models, J. Geophys. Res. Solid Earth, 105(B4), 8121–8145, doi:10.1029/1999JB900390. Bilotte, M. (1985), Le Crétacé supérieur des plates-formes est-pyrénéennes, Université Paul-Sabatier, Toulouse, France. Bilotte, M., M. Casteras, B. Peybernès, J. Rey, J.-C. Soula, and F. Taillefer (1988a), Carte géologique de la France au 1/50 000: Foix (feuille №1075), BRGM, Orléans, France. Chapter 2. Case study of the Eastern Pyrenees 59

Bilotte, M., J. Cosson, B. Crochet, B. Peybernès, J. Roche, F. Taillefer, Y. Tambareau, Y. Ternet, and J. Villatte (1988b), Carte géologique de la France au 1/50 000: Lavelanet (feuille №1076), BRGM, Orléans, France. Bilotte, M., J. Cosson, B. Crochet, B. Peybernès, J. Roche, F. Taillefer, Y. Tambareau, Y. Ternet, and J. Villatte (1988c), Notice explicative de la feuille Lavelanet à 1/50 000 (Feuille №1076), BRGM, Orléans, France. Bond, R. M. G., and K. R. McClay (1995), Inversion of a Lower Cretaceous extensional basin, south central Pyrenees, Spain, in Basin Inversion, Geological Society Special Publication, vol. 88, edited by J. G. Buchanan and P. G. Buchanan, pp. 415–431, Geological Society of London, London, United Kingdom. Bulnes, M., and J. Poblet (1999), Estimating the detachment depth in cross sections involving detachment folds, Geol. Mag., 136(4), S0016756899002794, doi:10.1017/S0016756899002794. Burbank, D. W., J. Verges, J. A. Munoz, and P. Bentham (1992a), Coeval hindward- and forward-imbricating thrusting in the south- central Pyrenees, Spain: timing and rates of shortening and deposition, Geol. Soc. Am. Bull., 104(1), 3–17, doi:10.1130/0016-7606(1992)104<0003:CHAFIT>2.3.CO;2. Burbank, D. W., C. Puigdefàbregas, and J. A. Muñoz (1992b), he chronology of the Eocene tectonic and stratigraphic development of the eastern Pyrenean foreland basin, northeast Spain, Geol. Soc. Am. Bull., 104(9), 1101–1120, doi:10.1130/0016-7606(1992)104<1101:TCOTET>2.3.CO;2. Carrigan, J. H., D. J. Anastasio, K. P. Kodama, and J. M. Parés (2016), Fault-related fold kinematics recorded by terrestrial growth strata, Sant Llorenç de Morunys, Pyrenees Mountains, NE Spain, J. Struct. Geol., 91, 161–176, doi:10.1016/j.jsg.2016.09.003. Cavaillé, A. (1976a), Carte géologique de la France au 1/50 000: Mirepoix (Feuille № 1058), BRGM, Orléans, France. Cavaillé, A. (1976b), Notice explicative de la feuille Mirepoix à 1/50 000 (Feuille №1058), Cavaillé, A., P. Debat, and G. Calas (1975a), Carte géologique de la France au 1/50 000: Castelnaudary (feuille №1036), BRGM, Orléans, France. Cavaillé, A., P. Debat, and G. Calas (1975b), Notice explicative de la feuille Castelnaudary à 1/50 000 (Feuille №1036), BRGM, Orléans, France. Chevrot, S. et al. (2014), High-resolution imaging of the Pyrenees and Massif Central from the data of the PYROPE and IBERARRAY portable array deployments, J. Geophys. Res. Solid Earth, 119(8), 6399–6420, doi:10.1002/2014JB010953. Chevrot, S. et al. (2015), he Pyrenean architecture as revealed by teleseismic P-to-S converted waves recorded along two dense transects, Geophys. J. Int., 200(1), 1096–1107, doi:10.1093/gji/ggu400. Choukroune, P. (1974), Structure et evolution tectonique de la zone nord-Pyrénéenne. Analyse de la déformation dans une portion de chaine a schistosité sub-verticale., Université des Sciences et Techniques du Languedoc, Montpellier, France. Choukroune, P., and M. Mattauer (1978), Tectonique des plaques et Pyrénées: sur le fonctionnement de la faille transformante nord-Pyrénéenne; comparaisons avec des modèles actuels, Bull. la Soc. Geol. Fr., 7 t. XX(5), 689–700. Christophoul, F., J.-C. Soula, S. Brusset, B. Elibana, M. Roddaz, G. Bessiere, and J. Deramond (2003), Time, place and mode of propagation of foreland basin systems as recorded by the sedimentary ill: examples of the Late Cretaceous and Eocene retro-foreland basins of the north-eastern Pyrenees, in Tracing Tectonic Deformation Using the Sedimentary Record. Geological Society, London, Special Publications, vol. 208, edited by T. McCann and A. Saintot, pp. 229–252, Geological Society of London, London, United Kingdom. Clerc, C., A. Lahid, P. Monié, Y. Lagabrielle, C. Chopin, M. Poujol, P. Boulvais, J.-C. Ringenbach, E. Masini, and M. de Saint Blanquat (2015), High-temperature metamorphism during extreme thinning of the continental crust: A reappraisal of the North Pyrenean passive paleomargin, Solid Earth, 6(2), 643–668, doi:10.5194/se-6-643-2015. Cochelin, B., B. Lemirre, Y. Denèle, M. de Saint Blanquat, A. Lahid, and S. Duchêne (2017), Structural inheritance in the Central Pyrenees: the Variscan to Alpine tectonometamorphic evolution of the Axial Zone, J. Geol. Soc. London. Cohen, K. M., S. Finney, and P. L. Gibbard (2013), International Chronostratigraphic Chart 2013/01, Costa, E., M. Garcés, M. López-Blanco, E. Beamud, M. Gómez-Paccard, and J. C. Larrasoaña (2010), Closing and continentalization of the South Pyrenean foreland basin (NE Spain): magnetochronological constraints, Basin Res., 22(6), 904–917, doi:10.1111/j.1365-2117.2009.00452.x. Crochet, B. (1991), Molasses syntectoniques du versant nord des Pyrénées: la série de Palassou, BRGM, Orléans, France. 60 Chapter 2. Case study of the Eastern Pyrenees

Crochet, B., J. Villatte, Y. Tambareau, M. Bilotte, R. Bousquet, A. Kuhfuss, J. P. Bouillin, J. P. Gelard, G. Bessiere, and J. P. Paris (1989), Carte géologique de la France au 1/50 000: Quillan (Feuille №1077), BRGM, Orléans, France. Cruz, L., C. Teyssier, L. Perg, A. Take, and A. Fayon (2008), Deformation, exhumation, and topography of experimental doubly-vergent orogenic wedges subjected to asymmetric erosion, J. Struct. Geol., 30(1), 98–115, doi:10.1016/j.jsg.2007.10.003. Debroas, E. J. (1990), Le lysch noir albo-cénomanien témoin de la structuration albienne à sénonienne de la Zone nord-pyrénéenne en Bigorre (Hautes-Pyrénées, France), Bull. la Soc. Geol. Fr., VI(2), 273–285, doi:10.2113/gssgbull.VI.2.273. Decarlis, A., M. Maino, G. Dallagiovanna, A. Lualdi, E. Masini, S. Seno, and G. Toscani (2014), Salt tectonics in the SW Alps (Italy-France): From riting to the inversion of the European continental margin in a context of oblique convergence, Tectonophysics, 636, 293–314, doi:10.1016/j.tecto.2014.09.003. Desegaulx, P., and M.-F. Brunet (1990), Tectonic subsidence of the Aquitaine Basin since Cretaceous times, Bull. la Soc. Geol. Fr., VI(2), 295–306, doi:10.2113/gssgbull.VI.2.295. Díaz, J., and J. Gallart (2009), Crustal structure beneath the Iberian Peninsula and surrounding waters: A new compilation of deep seismic sounding results, Phys. Earth Planet. Inter., 173(1–2), 181–190, doi:10.1016/j.pepi.2008.11.008. Duerto, L., and K. R. McClay (2009), he role of syntectonic sedimentation in the evolution of doubly vergent thrust wedges and foreland folds, Mar. Pet. Geol., 26(7), 1051–1069, doi:10.1016/j. marpetgeo.2008.07.004. Ellenberger, F., P. Freytet, J. Plaziat, G. Bessiere, P. Viallard, G.-M. Berger, and J.-P. Marchal (1987), Carte géologique de la France au 1/50 000: Capendu (Feuille №1060), BRGM, Orléans, France. Erdős, Z., R. S. Huismans, P. van der Beek, and C. hieulot (2014), Extensional inheritance and surface processes as controlling factors of mountain belt structure, J. Geophys. Res. Solid Earth, 119, 9042–9061, doi:10.1002/2014JB011408. Erdős, Z., R. S. Huismans, and P. van der Beek (2015), First-order control of syntectonic sedimentation on crustal- scale structure of mountain belts, J. Geophys. Res. Solid Earth, 120, 1–16, doi:10.1002/2014JB011785. Ershov, A. V., M.-F. Brunet, A. M. Nikishin, S. N. Bolotov, B. P. Nazarevich, and M. V. Korotaev (2003), Northern Caucasus basin: hermal history and synthesis of subsidence models, Sediment. Geol., 156(1–4), 95–118, doi:10.1016/S0037-0738(02)00284-1. Fillon, C., and P. van der Beek (2012), Post-orogenic evolution of the southern Pyrenees: Constraints from inverse thermo-kinematic modelling of low-temperature thermochronology data, Basin Res., 24(4), 418–436, doi:10.1111/j.1365-2117.2011.00533.x. Fillon, C., R. S. Huismans, P. van der Beek, and J. A. Muñoz (2013), Syntectonic sedimentation controls on the evolution of the southern Pyrenean fold-and-thrust belt: Inferences from coupled tectonic-surface processes models, J. Geophys. Res. Solid Earth, 118(10), 5665–5680, doi:10.1002/jgrb.50368. Fitzgerald, P. G., J. A. Muñoz, P. J. Coney, and S. L. Baldwin (1999), Asymmetric exhumation across the Pyrenean orogen: Implications for the tectonic evolution of a collisional orogen, Earth Planet. Sci. Lett., 173, 157–170, doi:10.1016/S0012-821X(99)00225-3. Ford, M., E. A. Williams, A. Artoni, J. Vergés, and S. Hardy (1997), Progressive evolution of a fault-related fold pair from growth strata geometries, Sant Llorenç de Morunys, SE Pyrenees, J. Struct. Geol., 19(3–4), 413–441, doi:10.1016/S0191-8141(96)00116-2. Ford, M., L. Hemmer, A. Vacherat, K. Gallagher, and F. Christophoul (2016), Retro-wedge foreland basin evolution along the ECORS line, eastern Pyrenees, France, J. Geol. Soc. London., (173), 419–437, doi:10.1144/ jgs2015-129. Garcia-Castellanos, D., and J. C. Larrasoaña (2015), Quantifying the post-tectonic topographic evolution of closed basins: he Ebro basin (northeast Iberia), Geology, 43(8), 663–666, doi:10.1130/G36673.1. García-Castellanos, D., J. Vergés, J. Gaspar-Escribano, and S. Cloetingh (2003), Interplay between tectonics, climate, and luvial transport during the Cenozoic evolution of the Ebro Basin (NE Iberia), J. Geophys. Res., 108(B7), 2347, doi:10.1029/2002jb002073. Golberg, J. M., and A. F. Leyreloup (1990), High temperature-low pressure Cretaceous metamorphism related to crustal thinning (Eastern North Pyrenean Zone, France), Contrib. to Mineral. Petrol., 104(2), 194–207, doi:10.1007/BF00306443. Gómez-Paccard, M., M. López-Blanco, E. Costa, M. Garcés, E. Beamud, and J. C. Larrasoaña (2012), Tectonic and climatic controls on the sequential arrangement of an alluvial fan/fan-delta complex (Montserrat, Eocene, Ebro Basin, NE Spain), Basin Res., 24(4), 437–455, doi:10.1111/j.1365-2117.2011.00532.x. Chapter 2. Case study of the Eastern Pyrenees 61

Gouache, C. (2017), Etudes tectono-métamorphiques des brèches de la zone interne métamorphique des Pyrénées ariégeoises., Université de Lorraine. Hardy, S., K. R. McClay, and J. A. Muñoz (2009), Deformation and fault activity in space and time in high- resolution numerical models of doubly vergent thrust wedges, Mar. Pet. Geol., 26(2), 232–248, doi:10.1016/j.marpetgeo.2007.12.003. Hoth, S., A. Hofmann-Rothe, and N. Kukowski (2007), Frontal accretion: An internal clock for bivergent wedge deformation and surface uplit, J. Geophys. Res. Solid Earth, 112(January), B06408, doi:10.1029/2006JB004357. Hoth, S., N. Kukowski, and O. Oncken (2008), Distant efects in bivergent orogenic belts — How retro-wedge erosion triggers resource formation in pro-foreland basins, Earth Planet. Sci. Lett., 273(1–2), 28–37, doi:10.1016/j.epsl.2008.05.033. Jammes, S., and R. S. Huismans (2012), Structural styles of mountain building: Controls of lithospheric rheologic stratiication and extensional inheritance, J. Geophys. Res. Solid Earth, 117(B10), B10403, doi:10.1029/2012JB009376. Jammes, S., G. Manatschal, L. Lavier, and E. Masini (2009), Tectonosedimentary evolution related to extreme crustal thinning ahead of a propagating ocean: Example of the western Pyrenees, Tectonics, 28(4), TC4012, doi:10.1029/2008TC002406. Jammes, S., R. S. Huismans, and J. A. Muñoz (2014), Lateral variation in structural style of mountain building: controls of rheological and rit inheritance, Terra Nov., 26, 201–207, doi:10.1111/ter.12087. Jarvis, A., H. I. Reuter, A. Nelson, and E. Guevara (2008), Hole-illed SRTM for the globe Version 4, available from the CGIAR-CSI SRTM 90m Database, Jolivet, M., P. Labaume, P. Monié, M. Brunel, N. Arnaud, and M. Campani (2007), hermochronology constraints for the propagation sequence of the south Pyrenean basement thrust system (France-Spain), Tectonics, 26, TC5007, doi:10.1029/2006TC002080. Lagabrielle, Y., P. Labaume, and M. de Saint Blanquat (2010), Mantle exhumation, crustal denudation, and gravity tectonics during Cretaceous riting in the Pyrenean realm (SW Europe): Insights from the geological setting of the lherzolite bodies, Tectonics, 29(4), TC4012, doi:10.1029/2009TC002588. Laumonier, B. (2015), he Alpine southeastern Pyrenees (France, Spain); a synthesis, Rev. Géologie pyrénéenne, 2(1), 44 p. Lewis, C. J., J. Vergés, and M. Marzo (2000), High mountains in a zone of extended crust: Insights into the Neogene-Quaternary topographic development of northeastern Iberia, Tectonics, 19(1), 86–102, doi:10.1029/1999tc900056. López-Blanco, M., M. Marzo, D. W. Burbank, J. Vergés, E. Roca, P. Anadón, and J. Piña (2000), Tectonic and climatic controls on the development of foreland fan deltas: Montserrat and Sant Llorenç del Munt systems (Middle Eocene, Ebro Basin, NE Spain), Sediment. Geol., 138(1–4), 17–39, doi:10.1016/S0037- 0738(00)00142-1. Martinez, A., J. Vergés, E. Clavell, and J. Kennedy (1989), Stratigraphic framework of the thrust geometry and structural inversion in the southeastern Pyrenees : La Garrotxa area, Geodin. Acta, 3(3), 185–194, doi:10. 1080/09853111.1989.11105185. Marty, F. (1976), Relations géologiques entre le massif du Saint-Barthélémy et les séries post-hercyniennes du pays de Sault (Pyrénées ariégeoises), Université Paul Sabatier, Toulouse, France. Maurel, O., P. Monié, R. Pik, N. Arnaud, M. Brunel, and M. Jolivet (2008), he Meso-Cenozoic thermo-tectonic evolution of the Eastern Pyrenees: An 40Ar/39Ar ission track and (U-h)/He thermochronological study of the Canigou and Mont-Louis massifs, Int. J. Earth Sci., 97, 565–584, doi:10.1007/s00531-007- 0179-x. McClay, K. R., J. A. Muñoz, and J. García-Senz (2004), Extensional salt tectonics in a contractional orogen: A newly identiied tectonic event in the Spanish Pyrenees, Geology, 32(9), 737–740, doi:10.1130/G20565.1. Meigs, A. J., J. Vergés, and D. W. Burbank (1996), Ten-million-year history of a thrust sheet, Bull. Geol. Soc. Am., 108(12), 1608–1625, doi:10.1130/0016-7606(1996)108<1608:TMYHOA>2.3.CO;2. Metcalf, J. R., P. G. Fitzgerald, S. L. Baldwin, and J. A. Muñoz (2009), hermochronology of a convergent orogen: Constraints on the timing of thrust faulting and subsequent exhumation of the Maladeta Pluton in the Central Pyrenean Axial Zone, Earth Planet. Sci. Lett., 287(3–4), 488–503, doi:10.1016/j.epsl.2009.08.036. Mitra, S., and V. S. Mount (1998), Foreland basement-involved structures, Am. Assoc. Pet. Geol. Bull., 82, 70–109, doi:10.1306/1D9BC39F-172D-11D7-8645000102C1865D. Morris, R. G., H. D. Sinclair, and A. J. Yelland (1998), Exhumation of the Pyrenean Orogen: implications for sediment discharge, Basin Res., 10, 69–85, doi:10.1046/j.1365-2117.1998.00053.x. 62 Chapter 2. Case study of the Eastern Pyrenees

Mouthereau, F., P.-Y. Filleaudeau, A. Vacherat, R. Pik, O. Lacombe, M. G. Fellin, S. Castelltort, F. Christophoul, and E. Masini (2014), Placing limits to shortening evolution in the Pyrenees: Role of margin architecture and implications for the Iberia/Europe convergence, Tectonics, 33(DECEMBER 2014), 2283–2314, doi:10.1002/2014TC003663. Muñoz, J. A. (1992), Evolution of a continental collision belt: ECORS-Pyrenees crustal balanced cross-section, in hrust Tectonics, edited by K. R. McClay, pp. 235–246, Springer Netherlands, Dordrecht, the Netherlands. Naylor, M., and H. D. Sinclair (2008), Pro- vs. retro-foreland basins, Basin Res., 20(3), 285–303, doi:10.1111/ j.1365-2117.2008.00366.x. Olivet, J. L. (1996), La cinématique de la plaque Ibérique, Bull. des Centres Rech. Elf Explor. Prod., 20, 191–195. Puigdefàbregas, C., and P. Souquet (1986), Tecto-sedimentary cycles and depositional sequences of the Mesozoic and Tertiary from the Pyrenees, Tectonophysics, 129(1–4), 173–203, doi:10.1016/0040-1951(86)90251-9. Pujalte, V., J. I. Baceta, A. Payros, X. Orue-Etxebarria, and J. Serra-Kiel (1994), Late Cretaceous-Middle Eocene sequence stratigraphy and biostratigraphy of the SW and W Pyrenees (Pamplona and Basque basins, Spain), Libr. des Excursions du Prem. Congrés Français Stratigr., 1–118. Rahl, J. M., S. H. Haines, and B. A. van der Pluijm (2011), Links between orogenic wedge deformation and erosional exhumation: Evidence from illite age analysis of fault rock and detrital thermochronology of syn-tectonic conglomerates in the Spanish Pyrenees, Earth Planet. Sci. Lett., 307(1–2), 180–190, doi:10.1016/j.epsl.2011.04.036. Ramos, E., P. Busquets, and J. Vergés (2002), Interplay between longitudinal luvial and transverse alluvial fan systems and growing thrusts in piggyback basin (SE Pyrenees), Sediment. Geol., 146(1–2), 105–131, doi:10.1016/S0037-0738(01)00169-5. Reiners, P. W. (2005), Zircon (U-h)/He hermochronometry, Rev. Mineral. Geochemistry, 58(1936), 151–179, doi:10.2138/rmg.2005.58.6. Ricateau, R., and J. Villemin (1973), Evolution au Crétacé supérieur de la pente séparant le domaine de plate-forme du sillon sous-pyrénéen en Aquitaine méridionale, Bull. la Soc. Geol. Fr., 7(1), 30–39. Roest, W. R., and S. P. Srivastava (1991), Kinematics of the plate boundaries between Eurasia, Iberia, and Africa in the North Atlantic from the Late Cretaceous to the present, Geology, 19(6), 613–616. Rosenbaum, G., G. S. Lister, and C. Duboz (2002), Relative motions of Africa, Iberia and Europe during Alpine orogeny, Tectonophysics, 359(1–2), 117–129, doi:10.1016/S0040-1951(02)00442-0. Rougier, G., M. Ford, F. Christophoul, and A.-G. Bader (2016), Stratigraphic and tectonic studies in the central Aquitaine Basin, northern Pyrenees: Constraints on the subsidence and deformation history of a retro- foreland basin, Comptes Rendus Geosci., 348(3–4), 224–235, doi:10.1016/j.crte.2015.12.005. Roure, F., and P. Choukroune (1998), Contribution of the ECORS seismic data to the Pyrenean geology: crustal architecture and geodynamic evolution of the Pyrenees, Mémoires la Soc. Geol. Fr., 173, 37–52. Roure, F., P. Choukroune, X. Berastegui, J. A. Muñoz, A. Villien, P. Matheron, M. Bareyt, M. Seguret, P. Camara, and J. Deramond (1989), Ecors deep seismic data and balanced cross sections: Geometric constraints on the evolution of the Pyrenees, Tectonics, 8(1), 41–50, doi:10.1029/TC008i001p00041. Rushlow, C. R., J. B. Barnes, T. A. Ehlers, and J. Vergés (2013), Exhumation of the southern Pyrenean fold-thrust belt (Spain) from orogenic growth to decay, Tectonics, 32(4), 843–860, doi:10.1002/tect.20030. de Saint Blanquat, M., J. M. Lardeaux, and M. Brunel (1990), Petrological arguments for high-temperature extensional deformation in the Pyrenean Variscan crust (Saint Barthélémy Massif, Ariège, France), Tectonophysics, 177(1–3), 245–262, doi:10.1016/0040-1951(90)90284-F. de Saint Blanquat, M., F. Bajolet, A. Grand’Homme, A. Proietti, M. Zanti, A. Boutin, C. Clerc, Y. Lagabrielle, and P. Labaume (2016), Cretaceous mantle exhumation in the central Pyrenees: New constraints from the peridotites in eastern Ariège (North Pyrenean zone, France), Comptes Rendus - Geosci., 348(3–4), 268–278, doi:10.1016/j.crte.2015.12.003. Saura, E., L. Ardèvol I Oró, A. Teixell, and J. Vergés (2016), Rising and falling diapirs, shiting depocenters, and lap overturning in the Cretaceous Sopeira and Sant Gervàs subbasins (Ribagorça Basin, southern Pyrenees), Tectonics, 35(3), 638–662, doi:10.1002/2015TC004001. Serra-Kiel, J., J. I. Canudo, J. Dinares, E. Molina, N. Ortiz, J. O. Pascual, J. M. Samsó, and J. Tosquella (1994), Cronoestratigrafía de los sedimentos marinos del Terciario inferior de la Cuenca de Graus-Trem (Zona Central Surpirenaica), Rev. la Soc. Geol. España, 7(3–4), 273–295. Sibuet, J.-C., S. P. Srivastava, and W. Spakman (2004), Pyrenean orogeny and plate kinematics, J. Geophys. Res., 109(B8), B08104, doi:10.1029/2003JB002514. Chapter 2. Case study of the Eastern Pyrenees 63

Sinclair, H. D., and M. Naylor (2012), Foreland basin subsidence driven by topographic growth versus plate subduction, Bull. Geol. Soc. Am., 124, 368–379, doi:10.1130/B30383.1. Sinclair, H. D., M. Gibson, M. Naylor, and R. G. Morris (2005), Asymmetric growth of the Pyrenees revealed through measurement and modeling of orogenic luxes, Am. J. Sci., 305, 369–406, doi:10.2475/ ajs.305.5.369. Snedden, J. W., and C. Liu (2010), A Compilation of Phanerozoic Sea-Level Change , Coastal Onlaps and Recommended Sequence Designations, Am. Assoc. Pet. Geol. Search Discov., Article 40594. Solé Sugrañes, L., and E. Clavell (1973), Nota sobre la edad y posición tectónica de los conglomerados eocenos de Queralt (Prepirineo oriental, Prov. de Barcelona), Acta Geol. Hisp., 8, 1–6. Souquet, P., and B. Peybernès (1987), Allochtonie des massifs primaires nord-pyrénéens des Pyrénées Centrales, C. R. Acad. Sc. Paris, 305(8), 733–739. Souquet, P., B. Peybernès, M. Bilotte, and E.-J. Debroas (1977), La chaîne alpine des Pyrénées, Géologie Alp., 53, 193–216. Souquet, P. et al. (1985), Le groupe du Flysch noir (albo-cénomanien) dans les Pyrénées, Bull. des Centres Rech. Elf Explor. Prod., 9, 183–252. Souriau, A., S. Chevrot, and C. Olivera (2008), A new tomographic image of the Pyrenean lithosphere from teleseismic data, Tectonophysics, 460(1–4), 206–214, doi:10.1016/j.tecto.2008.08.014. Srivastava, S. P., W. R. Roest, L. C. Kovacs, G. Oakey, S. Lévesque, J. Verhoef, and R. Macnab (1990), Motion of Iberia since the Late Jurassic: Results from detailed aeromagnetic measurements in the Newfoundland Basin, Tectonophysics, 184(3–4), 229–260, doi:10.1016/0040-1951(90)90442-B. Steckler, M. S., and A. B. Watts (1978), Subsidence of the Atlantic-type continental margin of New York, Earth Planet. Sci. Lett., 41(1), 1–13, doi:10.1016/0012-821X(78)90036-5. Sutra, E., G. Manatschal, G. Mohn, and P. Unternehr (2013), Quantiication and restoration of extensional deformation along the Western Iberia and Newfoundland rited margins, Geochemistry, Geophys. Geosystems, 14(8), 2575–2597, doi:10.1002/ggge.20135. Tambareau, Y., B. Crochet, J. Villatte, and J. Deramond (1995), Evolution tectono-sedimentaire du versant nord des Pyrenees centre-orientales au Paleocene et a l’Eocene inferieur, Bull. la Soc. Géologique Fr., 166, 375–387, doi:10.2113/gssgbull.166.4.375. Teixell, A. (1996), he Anso transect of the southern Pyrenees: basement and cover thrust geometries, J. Geol. Soc. London., 153(2), 301–310, doi:10.1144/gsjgs.153.2.0301. Teixell, A. (1998), Crustal structure and orogenic material budget in the west central Pyrenees, Tectonics, 17(3), 395–406, doi:10.1029/98TC00561. Teixell, A., P. Labaume, and Y. Lagabrielle (2016), he crustal evolution of the west-central Pyrenees revisited: Inferences from a new kinematic scenario, Comptes Rendus - Geosci., 348(3–4), 257–267, doi:10.1016/j. crte.2015.10.010. Ternois, S., A. Vacherat, R. Pik, M. Ford, and B. Tibari (2017), Zircon ( U-h )/ He evidence for pre-Eocene orogenic exhumation of eastern North Pyrenean massifs , France, in Geophysical Research Abstracts, vol. 19, p. 17649. Tugend, J., G. Manatschal, N. J. Kusznir, and E. Masini (2015), Characterizing and identifying structural domains at rited continental margins: application to the Bay of Biscay margins and its Western Pyrenean fossil remnants, Geol. Soc. London, Spec. Publ., 413(1), 171–203, doi:10.1144/SP413.3. Vacherat, A. et al. (2016), Rit-to-collision transition recorded by tectonothermal evolution of the northern Pyrenees, Tectonics, 35, 907–933, doi:10.1002/2015TC004016. Valero, L., M. Garcés, L. Cabrera, E. Costa, and A. Sáez (2014), 20 Myr of eccentricity paced lacustrine cycles in the Cenozoic Ebro Basin, Earth Planet. Sci. Lett., 408, 183–193, doi:10.1016/j.epsl.2014.10.007. Vergés, J. (1993), Estudi geològic del vessant sud del Pirineu oriental i central. Evolució cinemàtica en 3D, University of Barcelona, Spain. Vergés, J., and D. W. Burbank (1996), Eocene-Oligocene hrusting and Basin Coniguration in the Eastern and Central Pyrenees (Spain), in Tertiary Basins of Spain. he Stratigraphic Record of Crustal Kinematics, edited by P. F. Friend and C. J. Dabrio, pp. 120–133, Cambridge University Press, Cambridge, United Kingdom. Vergés, J., and J. García-Senz (2001), Mesozoic evolution and Cainozoic inversion of the Pyrenean rit, in Peri- Tethys Memoir 6: Peri-Tethyan Rit/Wrench Basins and Passive Margins, vol. 186, edited by P. A. Ziegler, W. Cavazza, A. H. F. Robertson, and S. Crasquin-Soleau, pp. 187–212, Mémoires du Muséum national d’histoire naturelle, Paris, France. 64 Chapter 2. Case study of the Eastern Pyrenees

Vergés, J., and A. Martinez (1988), Corte compensado del Pirineo oriental: Geometría de las cuencas de antepaís y edades de emplazamiento de los mantos de corrimiento, Acta Geol. Hisp., 23200106,(2), 95–106. Vergés, J., J. A. Muñoz, and A. Martínez (1992), South Pyrenean fold and thrust belt: he role of foreland evaporitic levels in thrust geometry, in hrust Tectonics, edited by K. R. McClay, pp. 255–264, Springer Netherlands, Dordrecht, the Netherlands. Vergés, J., A. Martínez-Ríuz, F. Domingo, J. A. Muñoz, M. Losantos, J. Fleta, and J. Gisbert (1994), Mapa Geológico de España. Plan Magna a escala 1:50 000. Hoja de La Pobla de Lillet (255), Instituto Tecnológico Geominero de España, Madrid, Spain. Vergés, J., H. Millán, E. Roca, J. A. Muñoz, M. Marzo, J. Cirés, T. Den Bezemer, R. Zoetemeijer, and S. Cloetingh (1995), Eastern Pyrenees and related foreland basins: pre-, syn- and post-collisional crustal-scale cross- sections, Mar. Pet. Geol., 12(8), 903–915, doi:10.1016/0264-8172(95)98854-X. Vergés, J., M. Marzo, T. Santaeulària, J. Serra-Kiel, D. W. Burbank, J. A. Muñoz, and J. Giménez-Montsant (1998), Quantiied vertical motions and tectonic evolution of the SE Pyrenean foreland basin, in Cenozoic foreland Basins of Western Europe, Geological Society Special Publication, vol. 134, edited by A. Mascle, C. Puigdefàbregas, H. P. Lutherbacher, and M. Fernàndez, pp. 107–134, Geological Society of London, London, United Kingdom. Vergés, J., M. Fernàndez, and A. Martínez (2002), he Pyrenean orogen: Pre-, syn-, and post-collisional evolution, J. Virtual Explor., 8, 55–74, doi:10.3809/jvirtex.2002.00058. Vissers, R. L. M., and P. T. Meijer (2012a), Iberian plate kinematics and Alpine collision in the Pyrenees, Earth- Science Rev., 114(1–2), 61–83, doi:10.1016/j.earscirev.2012.05.001. Vissers, R. L. M., and P. T. Meijer (2012b), Mesozoic rotation of Iberia: Subduction in the Pyrenees?, Earth-Science Rev., 110(1–4), 93–110, doi:10.1016/j.earscirev.2011.11.001. Wang, Y. et al. (2016), he deep roots of the western Pyrenees revealed by full waveform inversion of teleseismic P waves, Geology, 44(6), 475–478, doi:10.1130/G37812.1. Whitchurch, A. L., A. Carter, H. D. Sinclair, R. A. Duller, A. C. Whittaker, and P. A. Allen (2011), Sediment routing system evolution within a diachronously upliting orogen: Insights from detrital zircon thermochronological analyses from the South-Central Pyrenees, Am. J. Sci., 311(5), 442–482, doi:10.2475/05.2011.03. Willett, S. D., C. Beaumont, and P. Fullsack (1993), Mechanical model for the tectonics of doubly vergent compressional orogens, Geology, 21(4), 371–374, doi:10.1130/0091-7613(1993)021<0371:MMFTTO>2. 3.CO. Woodward, N. B., S. E. Boyer, and J. Suppe (1989), Balanced Geological Cross-Sections: An Essential Technique in Geological Research and Exploration, edited by M. L. Crawford and E. Padovani, American Geophysical Union, Washington D.C., United States of America. Chapter 3

Rit inheritance, surface processes, and décollement control on orogenic mountain belt structure explained by dynamic models

in preparation for publication in Geology Arjan R. Grool1,2, Ritske S. Huismans2, Mary Ford1 1CRPG, UMR 7358, 15 Rue Notre Dame des Pauvres, 54501 Vandœvre-lès-Nancy, France 2Department of Earth Science, University of Bergen, Bergen N-5007, Norway 66 Chapter 3. Numerical modelling of orogen crustal structure

3.1 Introduction he distribution of shortening between the pro-wedge (lower plate) and retro-wedge (upper plate) varies greatly between orogens around the world. Distributions range from the highly asymmetric Zagros [Molinaro et al., 2005; Mouthereau et al., 2007], to the symmetric High Atlas [Beauchamp et al., 1999; Arboleya et al., 2004]. Shortening distribution can also vary strongly along strike, as in the Alps [Rosenberg and Kissling, 2013]. he factors that control the distribution of shortening and how it evolved through time are currently poorly understood. However, models have shown qualitatively that rit inheritance can inluence orogen crustal structure, promoting propagation of deformation into the retro-wedge [e.g., Jammes and Huismans, 2012; Erdős et al., 2014]. Determining whether this alters the distribution of shortening between the pro- and retro-wedge requires further quantitative analysis. Another poorly understood aspect of the crustal structure of orogens is the formation of crustal scale antiformal stacks: long crustal thrust sheets with large ofsets that are stacked and folded into a crustal scale anti-form, such as the Lepontine dome in the Alps [e.g., Schmid and Kissling, 2000] and Axial Zone in the Pyrenees [e.g., Muñoz, 1992]. Proposed mechanisms include increasing the critical taper through strength changes along the mid-crustal décollement [Beaumont et al., 2000], or triggering underplating by reducing taper with erosion of the internal thick-skinned pro-wedge [Rushlow et al., 2013] and/or by sediment blanketing of the external thick-skinned pro-wedge [Sinclair et al., 2005]. Considering that surface processes can change the distribution of loads within an orogenic ediice, can ongoing thin- skinned deformation inluence the deeper thick-skinned structure? Although thin-skinned deformation is generally understood to be driven by thick-skinned deformation, the potential for a feedback efect on the crustal structure warrants investigation. Here we build on previous work [Erdős et al., 2014] that documents the efect of rit inheritance and investigate the combined roles of rit inheritance, a highly mobile cover and surface processes on the interaction between thin- and thick-skinned deformation across a doubly vergent orogen. We use high- resolution 2D thermo-mechanical numerical models to study the efect of fault- to lithosphere-scale dynamics on the crustal-scale shortening distribution including the formation of a crustal scale antiformal stack. We quantify the shortening distribution through time and compare our results with observations from the Pyrenees.

3.2 Methods We use a modiied 2D version of FANTOM, a high resolution, thermo-mechanically coupled, arbitrary Lagrangian-Eulerian, inite element code [hieulot, 2011]. he high resolution allowed us to model deformation at the lithospheric scale while still resolving good detail in the fold-and-thrust belts and associated sedimentary basins. Localisation of deformation is incorporated through strain weakening of frictional-plastic materials. Surface processes were included with simple algorithms for moderate elevation dependent erosion and sedimentation (up to 0 m base level) [e.g., Erdős et al., 2014; Appendix B]. Erosion and deposition are not mass balanced. Model 1 and supplementary models DR1 and DR2 consist of a laterally homogeneous continental crust, comprising 3 km of pre-orogenic sediments, underlain by a 1 km thick décollement layer, 21 km of upper crust and 10 km of strong lower crust. he lithospheric mantle extends to 120 km depth and sub-lithospheric mantle until 600 km depth (see Appendix B for details). Convergence was achieved by imposing 5 km/My velocity boundary conditions on the let and right sides for a total convergence rate of 10 km/My. Initial localisation of deformation is seeded by a small weak seed at the top of the strong lower crust (see Appendix B). Rit inheritance (Models 1 and DR1) was generated by irst extending for 50 km before starting convergence [e.g., Jammes and Huismans, 2012]. Model 1 has a highly mobile cover with a weak viscous décollement layer representing salt with efective viscosity = 1019 Pa∙s. Supplementary models DR1 and DR2 have a moderately mobile cover with a weak frictional-plastic décollement layer representing shale (efective angle of internal friction = 2°). hese models are similar to earlier work [Erdős et al., 2014, 2015] and are included in Appendix B for comparison with Model 1 and for quantitative Chapter 3. Numerical modelling of orogen crustal structure 67 analysis of the strain distribution between the pro- and retro-wedge, achieved by tracking the horizontal position of a column of points originating in the centre of the model (see Appendix B for more detail).

Model 1: inheritance + salt + surface processes

a t = 5 My, ∆xc = 0 km pre-orogenic sediments 0 ｣ｾ 350°ﾷＭ ﾷ ｟ ﾷ ｟ﾷ ｟ Ｍ C｟ Ｎ ｟ Ｎ ｊｦｭｔｦｭｾＱｾｉｾＡｦｾｾｾｾＺＺ ｭ ＺＺＭ Ｍ［［￪Ｚ［Ｍ saltＭＮＭ ｲ Ｎｲ｟ upperﾷ ｟ ﾷ｟ｾ Ｍ Ｍｾ ﾷcrustＭ ﾷ｟ ﾷ ＺＺ｟ ﾷ＠ strain rate ｾﾷﾷﾷﾷﾷﾷﾷﾷﾷﾷﾷﾷﾷﾷﾷﾷ ＺＺＺＺＺＺＺ＠ ...... :- -; ""'\":-,.,. ::.·.-.·.-.•ｾ［［ ﾷﾷﾷﾷﾷﾷﾷﾷﾷﾷﾷﾷﾷＡ＠ 10-13 ＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺＺｾＺＺＺＺＺｾ＠ . ..:;::::=:::::: lower crust : 40 550° C km lithospheric mantle s-1

80 0

b t = 7 My, ∆xc = 23 km 0 350° C

40 550° C km

80

c t = 9 My, ∆xc = 38 km 0 350° C

40

km 550° C

80

d t = 15 My, ∆xc = 95 km syn-orogenic sediments 0 350° C

40

km 550° C

80

e t = 28 My, ∆xc = 225 km 0 350° C

40 km 550° C

80 200 250 300 350 400 450 500 550 600 km

Figure 3.1. Evolution of model 1, with extensional inheritance and a salt décollement level. Δxc is total convergence. Plot windows were chosen so the upper plate appears stationary. he white line in the keystone block marks the column of points tracked for measuring the shortening distribution. Points on this line that are above topography (red) are ignored. Strain rate plots (let) show second invariant of the strain rate tensor. Grey dashed lines in strain rate plots indicate fully strain weakened areas. (a) rit phase. (b) Symmetric inversion. (c) Onset of collision and asymmetry, retro-wedge abandoned. (d) Retro-wedge reactivated and wide thin-skinned pro-foreland fold-and-thrust belt. (e) Final coniguration. Well-developed retro-wedge, narrow thick-skinned pro-wedge and wide thin-skinned pro-foreland fold- and-thrust belt. 68 Chapter 3. Numerical modelling of orogen crustal structure

3.3 Results 3.3.1 Model 1: Efects of extensional inheritance, salt, and surface processes. Model 1 is characterised by a three-phase evolution. Phase 0 is the pre-orogenic rit phase, where at t = 5 My, 50 km of extension results in a narrow, roughly symmetrical rit bounded by two conjugate crustal scale frictional-plastic shear zones that root in the weak mid-crust (Figure 3.1a). Two conjugate frictional- plastic shear zones in the lithospheric mantle accommodate extension in the mantle. he sedimentary cover has gravitationally slid of the rit margins. No sediments were deposited in the rit. During Phase 1 initial lithospheric shortening leads to symmetric inversion of the rit zone. Inherited extensional shear zones are preferentially reactivated during the irst 25 km of convergence, restoring crustal thickness, followed by uplit of a symmetric central block (Figure 3.1b). he strain rate distribution shows that both conjugate shear zones in the upper crust and upper mantle lithosphere are simultaneously active during this irst inversion phase. Phase 2 comprises the development of an asymmetric crustal scale orogeny (Figure 3.1c, d, e). Following symmetric inversion, strain localisation on a single large-scale shear zone initiates asymmetric subduction of lower crust and mantle lithosphere and abandonment of other, previously active shear zones (Figure 3.1c). Initial narrow thin-skinned fold and thrust belts root back into both the deep shear zones and the cover sliding of the upliting keystone block. At t = 15 My and 95 km of convergence, syn-tectonic sedimentation in a wedge-top basin promotes the creation of wide thin-skinned thrust sheets in the external pro-wedge (Figure 3.1d). Below and at the same time, thick-skinned deformation propagates outward in the pro-wedge to accrete a new upper crustal thrust sheet. Over-steepening of the antiformal stack trailing edge reactivates the retro-wedge (Figure 3.1d). Ater 230 km of convergence (28 My; Figure 3.1e) eicient decoupling along the salt décollement together with wedge-top sedimentation facilitate outward propagation of the thin-skinned pro-wedge fold-and- thrust belt. he internal antiformal stack continues to grow by basal accretion of crustal scale thrust sheets and the upliting internal zone is continually eroded. he contrasting behaviour resulting from a less eicient décollement horizon and absence of extensional inheritance is illustrated in models DR1 and DR2 [Appendix B; see also Erdős et al., 2014, 2015]. Model DR1 includes extensional inheritance and surface processes, but has a stronger shale décollement horizon as compared to Model 1. It exhibits a clear coupling between shallow and deep structures. As deformation migrates outward in the pro-wedge, thin-skinned leading edge structures root back into an active crustal scale shear zone. hese thin-skinned structures are then carried passively above the next crustal imbricate as deformation propagates forward. he décollement level is therefore active only at its leading edge at any one time, linking back to a single crustal structure. In contrast to Model 1 an antiformal stack does not form, however, retro-wedge development is very similar to Model 1 (Figure B.2). Supplementary model DR2 combines absence of rit inheritance with surface processes and a less eicient shale décollement. Subduction asymmetry is created with the irst shear zone, resulting in a highly asymmetric crustal scale pro-wedge orogen and negligible retro-wedge shortening (Figure B.3). he pro-wedge deformation style is the same as in DR1. 3.3.2 Shortening distribution Quantitative analysis of shortening distribution in the pro- and retro-wedges reveal characteristic diferences between Model 1 and supplementary models DR1 and DR2. Models 1 and DR1 (with rit inheritance) show initial symmetric inversion during Phase 1 with 50% of total shortening in the future retro-wedge. During asymmetric orogenic growth in Phase 2, the models with extensional inheritance exhibit ~18-20 % shortening in the retro-wedge (Figure 3.2a). Retro-wedge shortening rates show a similar pattern with an initial high, on average 50% shortening rate corresponding to symmetric inversion during Phase 1, a short phase during which the retro-wedge is abandoned, followed by reactivation of the retro- wedge at 20% of the total shortening rate (Figure 3.2b). In contrast, model DR2 without rit inheritance Chapter 3. Numerical modelling of orogen crustal structure 69 exhibits only minor retro-wedge shortening stabilising at about 6% of total shortening (Figure 3.2a) and correspondingly low retro-wedge shortening rate 3-9% ater 50 km of shortening (Figure 3.2b).

a Shortening distribution 70 30

60 Eastern 40 Pyrenees 50 50

40 60

30 Central 70 Pyrenees

20 SD1 80 pro-wedge (%) retro-wedge (%) 10 Model 1 90 SD2 0 100 0 100 200 300 convergence (km)

b Shortening rate distribution 100 0 phase 2 phase 1 80 20

60 40

40 60

20 80 pro-wedge (%) retro-wedge (%) 0 100 0 100 200 300 convergence (km)

Figure 3.2. (a) Distribution of shortening between the retro-wedge and pro-wedge, plotted against total convergence to eliminate efects of variable convergence rates. Central Pyrenean data (black box) from Beaumont et al. [2000]. East Pyrenean data (red line) from Grool et al. [2018]. (b) Distribution of shortening rates, plotted against total convergence. East Pyrenean data from Grool et al. [2018].

3.4 Discussion he models presented here identify three main factors that control the structural style of mountain belt formation: 1) rit inheritance, 2) eiciency of decoupling between thin-skinned and thick-skinned deformation, and 3) syn-tectonic erosion and sedimentation. Rit inheritance promotes more shortening in the retro-wedge and generates two phases, each with a distinct distribution of shortening. Phase 1 is characterised by reactivation of inherited upper crustal extensional frictional-plastic shear zones. In the model presented here the pre-existing rit symmetry results in symmetric inversion. Phase 1 structural styles of crustal shortening will depend on inherited extensional crustal geometries. Lithosphere scale asymmetric subduction and accretion is the primary characteristic of Phase 2 mountain building. Without extensional inheritance asymmetry at depth is immediately established and an asymmetric crustal orogen focuses most shortening on the pro-wedge with limited retro-wedge activity. Extensional inheritance leads to a two-phase shortening history with early inversion across both wedges. Further shortening in the thick-skinned retro-wedge leads to transport of crustal material onto the retro-wedge (Figure 3.3a). Although décollement rheology has a negligible direct impact on the distribution of shortening between the pro- and retro-wedge, it does provide a primary control on deformation style. A weak salt décollement allows very eicient decoupling leading to the formation of a broad low taper thin-skinned fold and thrust belt in the pro-wedge (Figure 3.3a). his efectively reduces the pro-wedge taper to subcritical and thus promotes thick-skinned underthrusting in the core of the orogenic wedge, resulting in a narrower inner orogen zone. he impact of erosion and sedimentation on the crustal structure of orogens is well-known [e.g., Willett et al., 1993; Braun and Yamato, 2010; Erdős et al., 2014, 2015]. Erosion is most eicient in the core of the 70 Chapter 3. Numerical modelling of orogen crustal structure orogen, reducing critical taper and resulting in a narrower orogen (Figure 3.3a). Sedimentation increases the load in the foreland (Figure 3.3a), reduces critical taper and promotes longer thin and thick-skinned thrust sheets with greater ofset along each thrust [e.g., Erdős et al., 2015]. he efects of a highly mobile cover and surface processes work in tandem to create a wide pro-wedge fold-and-thrust belt and promote a crustal scale antiformal stack (Figure 3.3a), without needing additional complexities such as stacked crustal weakness zones or lateral strength variations of the mid-crustal décollement [Beaumont et al., 2000]. 3.4.1 Pyrenean double vergent orogen

a Model 1: inheritance + salt + surface processes erosion gravitational moves onto sedimentation retro-wedge sliding upper plate reactivated 0 Legend

basal accretion pro retro 40 50 km syn-orogenic sediment 80

pre-orogenic b Central Pyrenees (ECORS Pyrenees) sediment S Ebro Basin South Pyrenean Axial Zone North Pyrenean Aquitaine N Zone Zone Basin upper crust 0 Iberian plate European plate lower crust

50 50 km

Figure 3.3. (a) Cartoon of Model 1 ater 225 km of convergence, showing the combined efects of rit inheritance, an eicient décollement layer and surface processes. (b) cartoon of Central Pyrenees, modiied ater Muñoz [1992] and Beaumont et al. [2000].

We next describe the characteristics of the Central Pyrenean type example of a mountain belt that resulted from inversion of a salt rich, narrow hyper extended rit (Figure 3.3a) [e.g., Muñoz, 1992; Jammes et al., 2009]. Features that require explanation include: (1) a two stage mountain belt evolution with initial symmetric inversion, followed by largely asymmetric shortening with about 80 % in the pro-wedge and 20% in the northern retro-wedge; (2) formation of a narrow crustal scale pro-wedge antiformal stack in the core of the orogen; (3) a broad pro-wedge fold and thrust belt with minor thick-skinned thrusting at depth and (4) a narrow thick-skinned retro-wedge. Previous work suggests that complex inherited weak zones and highly eicient erosion played a key role in the evolution of these geometries [Beaumont et al., 2000]. he numerical models presented here demonstrate that these key features can be explained by the combined efects of eicient erosion of the elevated internal zone of the orogen and a weak décollement layer at the base of the thin-skinned fold and thrust belt, promoting basal accretion of thick-skinned basement thrust sheets at depth in the core of the orogen, and a wide low taper pro-wedge basin. Deposition of material across the low taper foreland basin areas reinforces the interaction and feedback between thin and thick- skinned deformation. Extensional inheritance is an essential element that allows for shortening in both pro- and retro-wedge. Notably the models do not require complex inherited weaknesses. Similarly, the combined efects of orogenic erosion and low taper thin-skinned fold and thrust belts with syn-tectonic wedge top sedimentation may explain the antiformal structure of the Lepontine dome in the Central Alps and the adjacent wide foreland basin in North Alpine foreland [Schmid et al., 1996]; and comparable Chapter 3. Numerical modelling of orogen crustal structure 71 structures in the frontal zone of the Himalayan mountain belt [Beaumont et al., 2001; Searle et al., 2003; DeCelles et al., 2016].

3.5 Conclusions We have shown how forward dynamic models provide a general explanation for the evolution of Alpine type mountain belts as exempliied by the Central Pyrenean type example with formation of an crustal scale antiformal stack in the axial zone, adjacent wide low taper fold and thrust belts decoupled on weak salt, and eicient syn-orogenic erosion and deposition. We have also shown that rit inheritance promotes the two-phase evolution with early symmetric inversion followed by asymmetric mountain building. Rit inheritance is also an essential element for forming a doubly vergent orogenic wedge that accommodates >10% of total convergence in the retro-wedge. he models show that formation of an antiformal stack of crustal scale thrust sheets and adjacent low taper thin-skinned fold and thrust belts can be explained by the interaction and feedback between: (1) a weak décollement layer decoupling thin and thick-skinned deformation and resulting in low taper foreland basin deformation, (2) eicient erosion of the internal thick-skinned pro-wedge, and 3) wedge top deposition of sediments in the pro-wedge.

Acknowledgements Supplementary methods and model descriptions are available in Appendix B. Computing hours are part of SIGMA2 high performance computing allocation project NN4704K: 3D forward modeling of lithosphere extension and inversion. Ritske Huismans thanks and is indebted to Josep Anton Muñoz for sharing his insights into Pyrenean geology and for highly fruitful discussions.

References Arboleya, M.L., Teixell, A., Charroud, M., and Julivert, M., 2004, A structural transect through the High and Middle Atlas of Morocco: Journal of African Earth Sciences, v. 39, p. 319–327, doi: 10.1016/j. jafrearsci.2004.07.036. Beauchamp, W., Allmendinger, R.W., Barazangi, M., Demnati, A., El Alji, M., and Dahmani, M., 1999, Inversion tectonics and the evolution of the High Atlas Mountains, Morocco, based on a geological-geophysical transect: Tectonics, v. 18, p. 163–184, doi: 10.1029/1998TC900015. Beaumont, C., Jamieson, R.A., Nguyen, M.H., and Lee, B., 2001, Himalayan tectonics explained by extrusion of a low-viscosity crustal channel coupled to focused surface denudation: Nature, v. 414, p. 738–742, doi: 10.1038/414738a. Beaumont, C., Muñoz, J.A., Hamilton, J., and Fullsack, P., 2000, Factors controlling the Alpine evolution of the central Pyrenees inferred from a comparison of observations and geodynamical models: Journal of Geophysical Research: Solid Earth, v. 105, p. 8121–8145, doi: 10.1029/1999JB900390. Braun, J., and Yamato, P., 2010, Structural evolution of a three-dimensional, inite-width crustal wedge: Tectonophysics, v. 484, p. 181–192, doi: 10.1016/j.tecto.2009.08.032. DeCelles, P.G., Carrapa, B., Gehrels, G.E., Chakraborty, T., and Ghosh, P., 2016, Along-strike continuity of structure, stratigraphy, and kinematic history in the Himalayan thrust belt: he view from Northeastern India: Tectonics, v. 35, p. 2995–3027, doi: 10.1002/2016TC004298. Erdős, Z., Huismans, R.S., and van der Beek, P., 2015, First-order control of syntectonic sedimentation on crustal- scale structure of mountain belts: Journal of Geophysical Research: Solid Earth, v. 120, p. 1–16, doi: 10.1002/2014JB011785. Erdős, Z., Huismans, R.S., van der Beek, P., and hieulot, C., 2014, Extensional inheritance and surface processes as controlling factors of mountain belt structure: Journal of Geophysical Research: Solid Earth, v. 119, p. 9042–9061, doi: 10.1002/2014JB011408. Grool, A.R., Ford, M., Vergés, J., Huismans, R.S., Christophoul, F., and Dielforder, A., 2018, Insights Into the Crustal-Scale Dynamics of a Doubly Vergent Orogen From a Quantitative Analysis of Its Forelands: A Case Study of the Eastern Pyrenees: Tectonics, v. 37, p. 450–476, doi: 10.1002/2017TC004731. Jammes, S., and Huismans, R.S., 2012, Structural styles of mountain building: Controls of lithospheric rheologic stratiication and extensional inheritance: Journal of Geophysical Research: Solid Earth, v. 117, p. B10403, doi: 10.1029/2012JB009376. 72 Chapter 3. Numerical modelling of orogen crustal structure

Jammes, S., Manatschal, G., Lavier, L., and Masini, E., 2009, Tectonosedimentary evolution related to extreme crustal thinning ahead of a propagating ocean: Example of the western Pyrenees: Tectonics, v. 28, p. TC4012, doi: 10.1029/2008TC002406. Molinaro, M., Leturmy, P., Guezou, J.C., Frizon de Lamotte, D., and Eshraghi, S. a., 2005, he structure and kinematics of the southeastern Zagros fold-thrust belt, Iran: From thin-skinned to thick-skinned tectonics: Tectonics, v. 24, p. 1–19, doi: 10.1029/2004TC001633. Mouthereau, F., Tensi, J., Bellahsen, N., Lacombe, O., De Boisgrollier, T., and Kargar, S., 2007, Tertiary sequence of deformation in a thin-skinned/thick-skinned collision belt: he Zagros Folded Belt (Fars, Iran): Tectonics, v. 26, p. n/a-n/a, doi: 10.1029/2007TC002098. Muñoz, J.A., 1992, Evolution of a continental collision belt: ECORS-Pyrenees crustal balanced cross-section, in McClay, K.R. ed., hrust Tectonics, Dordrecht, the Netherlands, Springer Netherlands, p. 235–246, doi: 10.1007/978-94-011-3066-0_21. Rosenberg, C.L., and Kissling, E., 2013, hree-dimensional insight into central-alpine collision: Lower-plate or upper-plate indentation? Geology, v. 41, p. 1219–1222, doi: 10.1130/G34584.1. Rushlow, C.R., Barnes, J.B., Ehlers, T.A., and Vergés, J., 2013, Exhumation of the southern Pyrenean fold-thrust belt (Spain) from orogenic growth to decay: Tectonics, v. 32, p. 843–860, doi: 10.1002/tect.20030. Schmid, S.M., and Kissling, E., 2000, he arc of the western Alps in the light of geophysical data on deep crustal structure: Tectonics, v. 19, p. 62, doi: 10.1029/1999TC900057. Schmid, S.M., Pifner, O.A., Froitzheim, N., Schönborn, G., and Kissling, E., 1996, Geophysical-geological transect and tectonic evolution of the Swiss-Italian Alps: Tectonics, v. 15, p. 1036–1064, doi: 10.1029/96TC00433. Searle, M.P., Simpson, R.L., Law, R.D., Parrish, R.R., and Waters, D.J., 2003, he structural geometry, metamorphic and magmatic evolution of the Everest massif, High Himalaya of Nepal-South Tibet: Journal of the Geological Society, v. 160, p. 345–366, doi: 10.1144/0016-764902-126. Sinclair, H.D., Gibson, M., Naylor, M., and Morris, R.G., 2005, Asymmetric growth of the Pyrenees revealed through measurement and modeling of orogenic luxes: American Journal of Science, v. 305, p. 369–406, doi: 10.2475/ajs.305.5.369. hieulot, C., 2011, FANTOM: Two- and three-dimensional numerical modelling of creeping lows for the solution of geological problems: Physics of the Earth and Planetary Interiors, v. 188, p. 47–68, doi: 10.1016/j. pepi.2011.06.011. Willett, S.D., Beaumont, C., and Fullsack, P., 1993, Mechanical model for the tectonics of doubly vergent compressional orogens: Geology, v. 21, p. 371–374, doi: 10.1130/0091-7613(1993)021<0371:MMFTTO> 2.3.CO. Chapter 4

Strain partitioning in doubly vergent orogens: controls of rit inheritance, surface processes and décollement using numerical models

in preparation for publication in Journal of Geophysical Research: Solid Earth Arjan R. Grool1,2, Ritske S. Huismans2, Mary Ford1 1CRPG, UMR 7358, 15 Rue Notre Dame des Pauvres, 54501 Vandœuvre-lès-Nancy, France. 2Department of Earth Sciences, University of Bergen, Bergen N-5007, Norway 74 Chapter 4. Numerical modelling of strain partitioning

4.1 Introduction Collisional orogens can accommodate convergence by shortening their lower plate (pro-wedge) and shortening their upper plate (retro-wedge) [e.g., Willett et al. 1993]. he strain partitioning between the pro- and retro-wedge varies greatly between natural systems. An orogen can be singly vergent, accommodating all shortening in the pro-wedge, such as the Zagros [e.g., Molinaro et al., 2005; Sherkati et al., 2006; Allen et al., 2013], and most of the Greater Caucasus [e.g., Saintot et al., 2006]. Alternatively, orogens can be doubly vergent, accommodating shortening in both the pro- and retro-wedge, such as the Pyrenees [e.g., Muñoz, 1992], and Tien Shan [e.g., Loury et al., 2015]. In doubly vergent orogens, the pro-retro strain partitioning can even vary along strike, such as in the Alps [Rosenberg and Kissling, 2013], and in time, as in the Pyrenees [Grool et al., 2018]. What controls this variation in space and time is currently poorly understood. Most investigations into the controlling factors of this variation using forward models have been qualitative, focusing only on the distinction between singly and doubly vergent orogens. First order controlling factors that enable double vergence are rit inheritance [Jammes and Huismans, 2012; Erdős et al., 2014; see also Chapter 3], strength contrasts between the upper and lower plate [Vogt et al., 2017b], and decoupling of the upper and lower crust of the lower plate [Vogt et al., 2017a]. Other factors have been found to control the crustal structure but not orogen vergence, such as sedimentation and erosion resulting in a wider and narrower thick-skinned orogen, respectively [Erdős et al. 2015]. In combination with these surface processes, the crustal structure can also be inluenced by thin-skinned deformation, which depends on the décollement rheology (see Chapter 3). hese factors that inluence the crustal structure but cannot themselves control orogen vergence may still inluence the pro-retro strain partitioning within a doubly vergent orogen. So far, this inluence has not been quantiied for these factors individually. Especially the inluence of thin-skinned deformation is currently poorly understood. Décollement rheology and distribution provide major controls on the thin-skinned deformation, as understood through critical taper theory, models, and natural examples [e.g., Dahlen, 1990; Ford, 2005; Buiter, 2012]. herefore, testing the inluence of décollement rheology and distribution on the thick-skinned structure and pro-retro strain partitioning may shed light on the role of thin-skinned deformation. Testing combinations of multiple factors may shed light on their interaction. he pro-retro strain partitioning may also be inluenced by interaction of the pro- and retro-wedge themselves. How the pro- and retro-wedge interact is not clear, and highly dependent on one’s deinition of what constitutes the boundary between them. For practical purposes, we deine the pro- and retro-wedge as all material derived from the lower and upper plate respectively, rather than placing the boundary at the topographic drainage divide as in the deinition of Willett et al. [1993]. Using our deinition, there can be no material transfer between wedges. However, some studies suggest there is a link in the timing of deformation of both wedges [Sinclair et al. 2005; Hoth et al. 2007; see also Chapter 3]. We test various combinations of rit inheritance, décollement rheology, décollement distribution, and surface processes. We individually quantify the inluence of these factors on the pro-retro strain partitioning and how they interact. he quantiied strain partitioning also allows us to investigate the interaction between the pro- and retro-wedge through time. We then verify our quantiied pro-retro strain partitioning results by comparing to natural systems.

4.2 Methods For our experiments we used a modiied version of FANTOM, an arbitrary Lagrangian-Eulerian inite- element code [hieulot, 2011]. We ran high-resolution experiments featuring 2D, thermo-mechanically coupled, visco-plastic low. Flow is governed by the semi-static Stokes equation, assuming incompressibility and negligible inertial forces. In the viscous regime, materials follow a power-law rheology: Chapter 4. Numerical modelling of strain partitioning 75

Table 4.1. Mechanical and thermal material properties. Parameter sediment shale salt upper crust lower crust lithospheric astheno- (units) mantle sphere Flow law wet qtz ∙ 1 wet qtz ∙ 1 wet qtz ∙ 1 wet qtz ∙ 1 wet qtz ∙ 100 dry olivine wet olivine Type visco-plastic visco-plastic viscous visco-plastic visco-plastic visco-plastic visco-plastic Plastic rheology c (MPa)a 20 → 4 2 - 20 → 4 20 → 4 20 → 4 20 → 4 ϕa 15° → 2° 2° → 1° - 15° → 2° 15° → 2° 15° → 2° 15° → 2° Viscous rheology n 4 4 4 4 4 3.5 3.0 η (Pa∙s) power-law power-law 1019 power-law power-law power-law power-law A (Pa-n∙s-1)b 8.574 ∙ 10-28 8.574 ∙ 10-28 - 8.574 ∙ 10-28 8.574 ∙ 10-36 2.4168 ∙ 10-15 1.393 ∙ 10-14 Q (kJ mol-1) 222.815 ∙ 103 222.815 ∙ 103 - 222.815 ∙ 103 222.815 ∙ 103 540.41 ∙ 103 429.83 ∙ 103 V (cm3 mol-1) 0 0 - 0 0 25 ∙ 10-6 15 ∙ 10-6 Density & thermal parameters ρ (kg m-3) 2300 2300 2300 2800 2800 3360 3300 α (K-1) 3.1 ∙ 10-5 3.1 ∙ 10-5 3.1 ∙ 10-5 3.1 ∙ 10-5 3.1 ∙ 10-5 0 0

T0 (°C) 273.15 273.15 273.15 273.15 273.15 273.15 273.15 -1 -1 cp (J kg K ) 803.57 803.57 803.57 803.57 803.57 681.82 681.82 k (W m-1) 2.25 2.25 2.25 2.25 2.25 2.25 2.25 H (µW m-3) 0.8 ∙ 10-6 0.8 ∙ 10-6 0.8 ∙ 10-6 0.8 ∙ 10-6 0.8 ∙ 10-6 0 0 ahe arrow indicates the efect of strain weakening, going from the initial value to maximum weakening. b -n A scaling factor f is included in the pre-exponential material constant. he conversion is A=f Abase.

(4.1) where is the strain rate tensor, deined as the spatial derivatives of the velocity ield, is the pre- exponential material constant, the viscous creep stress, the exponent, the activation energy, is pressure, the activation volume, the universal gas constant and is temperature. Material constants , , , and are determined from laboratory data for wet quartzite, wet olivine and dry olivine rheologies [Karato and Wu, 1993; Gleason and Tullis, 1995]. In the plastic regime, the material’s strength follows the Drucker-Prager yield criterion, equivalent to Coulomb failure in 2D. his is given by the yield stress (4.2) where is the second invariant of the deviatoric stress tensor , is cohesion, and is the efective angle of internal friction. For visco-plastic materials, the model automatically chooses the appropriate low regime by computing an efective viscosity with the lowest available stress ( or ):

(4.3) where is the second invariant of the strain rate tensor. he thermal structure is governed by the heat equation

(4.4) where is density, the heat capacity, is time, the spatial dimensions, the thermal conductivity, and is radiogenic heat production. Density variations due to temperature are taken into account. 76 Chapter 4. Numerical modelling of strain partitioning

Figure 4.1 shows the initial coniguration of our models. he model domain is 1200 km wide and 600 km deep, and contains 2400x1200 Eulerian cells. he resolution is varied along the vertical axis to achieve a resolution of 500x200 m in the top 25 km. Pre-orogenic sediment occupies the top 3 km, underlain by a 1 km thick décollement layer and 21 km thick upper crust basement. he lower crust is 10 km thick, placing the Moho at 35 km depth. he Lithospheric mantle lies below until 120 km depth. Material properties for each layer are listed in Table 4.1. A 6x6 km weak seed of pre-strain weakened material is placed in the lower crust at the centre of the model. Velocity boundary conditions are imposed on both sides, introducing lithosphere material at 5 mm/y on each side while material is extracted below to keep the volume constant. he top is a free surface and the bottom allows free slip. he initial thermal proile is laterally uniform, increasing parabolically from = 0° C at the surface to = 550° C at the Moho, and linearly to = 1330° C at the base of the Lithosphere. he top and bottom boundaries are kept at = 0° C and = 1330° C, respectively while the side boundaries are thermally insulated. he initial coniguration of our models is completely symmetrical unless speciied otherwise, and the plate polarity is random. In our igures we have mirrored some models so all appear with the same polarity for ease of comparison.

a 6 x 6 km weak seed temperature (°C) 3 km centered around 28 km depth 0 550 1330 0.5 cm/y 0 1 km Crust 35 Lithospheric mantle 21 km 120 Sublithospheric mantle

10 km Depth (km)

600 0 600 1200 Horizontal scale (km) Materials b Strain softening c Strength 0 Predeformation sediment 15° Wet Quartz, ƒ=1

Décollement layer ) ε

( Upper crust

Upper crust eff

φ Wet Lower crust Lower crust Quartz, ƒ=100 2° Depth (km) Lithospheric mantle 0.5ε 1.5 Sublithospheric mantle Dry Olivine, ƒ=1 20 Lithospheric mantle

4 Cohesion (MPa) 120 0.5ε 1.5 0Stress (MPa) 400

Figure 4.1. (a) Initial geometry and boundary conditions. (b) Implementation of strain sotening of frictional materials. (c) Initial strength proile and rheological layering. Ater Erdős et al. [2014]. Chapter 4. Numerical modelling of strain partitioning 77

In our models we have tested the inluence of rit inheritance, sedimentation and erosion, décollement rheology and décollement distribution (Table 4.2). Rit inheritance was implemented by inverting the velocity boundary conditions during the irst 5 My, leading to 50 km of extension [Jammes and Huismans, 2012]. We used a simple sedimentation algorithm that ills all available space up to a base level of 0 m, starting at 10 My [e.g., Erdős et al., 2014]. Erosion is altitude dependent, following , where is altitude, is time and is the inverse erosional timescale (in s-1). was chosen so that a 4 km high topography erodes by 1 km in 2 My [Erdős et al., 2014]. Eroded and deposited volumes were not linked. he inluence of décollement rheology was tested by using two possible décollement rheologies: a reference shale-like visco-plastic rheology with an efective angle of internal friction = 2° (M1-M5), or a weak salt-like linear viscous rheology with an efective viscosity = 1019 Pa∙s (M6-M8), similar to estimates for halite [e.g., van Keken et al., 1993; Chemia et al., 2008]. We tested the inluence of the décollement distribution by removing the décollement from half the model (M5), replacing it with regular pre-orogenic sediment. We quantify the strain partitioning between the pro- and retro-wedge by tracking the horizontal position of a column of points that originate at the centre of the model (see also Chapter 3). his column is displaced towards the side that accommodates the most shortening, and the amount of displacement is a measure for the ratio of shortening accommodated on either side. Note that this does not distinguish between the pro- and retro-wedge as originally deined by Willett et al. [1993], where the boundary between the pro- and retro-wedge is a topographic drainage divide. Instead, our method tracks a tectonic boundary that originates at the plate boundary or S-point. his facilitates comparison with natural systems, where the position of this boundary can be reconstructed.

Table 4.2. List of models Décollement Model Extension Rheologya Distribution Sedimentationb Erosionc Set 1: Shale décollement M1d,e,f No Shale Full width No No M2d Yes Shale Full width No No M3d Yes Shale Full width Yes No M4d,f Yes Shale Full width Yes Yes M5 Yes Shale Let half No No Set 2: Salt décollement M6 Yes Salt Full width No No M7 Yes Salt Full width Yes No M8c Yes Salt Full width Yes Yes Supplementary modelsg S1 Yes Weak salt Full width No No S2 Yes Weak Salt Full width Yes No S3 Yes Weak Salt Full width Yes Yes a 19 18 Shale ϕef = 2°, Salt ηef = 10 Pa∙s, Weak salt ηef = 10 Pa∙s. bbase level = 0 m, start at 10 My. c -15 -1 Ef = 4.35∙10 s . dErdős et al., [2014]. eErdős et al., [2015]. fSee Chapter 3. gSee Appendix C for details, igures, and descriptions. 78 Chapter 4. Numerical modelling of strain partitioning

4.3 Results Here we describe our model results in two sets (Table 4.2). Set 1 contains all models with a shale décollement (M1 to M5), and set 2 contains all models with a salt décollement (M6 to M8). Some of these models are identical or very similar to previously published models, and are used as reference against the new models to provide insight into the inluence of the décollement rheology and distribution. Amounts of convergence are given as ‘convergence since irst shear zone’ for M1, or ‘convergence beyond full inversion of the extensional shear zones’ for M2 to M8. 4.3.1 Model descriptions 4.3.1.1 Set 1: Shale décollement models Model M1 accommodates early convergence by distributed deformation. At 6 My, a shear zone forms through the crust and lithospheric mantle, creating an asymmetry that determines the subduction polarity. Ater 35 km of further convergence a new shear zone is formed through the upper crust, forward of the irst shear zone, forming the basal thrust of the irst thick-skinned pro-wedge thrust sheet. More thick- a M1: no extension + shale pre-orogenic t = 36 My; x = 300 km ∆ C shale sediments 0 350° C upper crust lower crust 40 km 550° C lithospheric mantle 80 b M2: extension + shale

t = 36 My; ∆xC = 300 km 0 350° C

40 km 550° C

80 c M3: extension + shale + sedimentation syn-orogenic sediments t = 36 My; ∆xC = 300 km 0 350° C

40

km 550° C

80 d M4: extension + shale + sedimentation + erosion

t = 36 My; ∆xC = 300 km 0 350° C

40 km 550° C

80 0 100 200 300 400 500 600 km

Figure 4.2. Model results of set 1 ater 36 My and 300 km of convergence. Convergence (ΔxC) was measured starting from formation of the irst shear zone in M1, or starting from restored crustal thickness in models with extensional inheritance. he white line indicates the tracked points that measure the distribution of shortening. Points above topography are ignored and coloured red. Chapter 4. Numerical modelling of strain partitioning 79 skinned pro-wedge thrust sheets are accreted by in-sequence formation of new shear zones every ~31 km of additional convergence, forming a wide thick-skinned pro-wedge (Figure 4.2a). In contrast, only a single, poorly developed retro-ward shear zone that accommodates very little shortening is formed. his gives the orogen a highly asymmetric, one-sided appearance. hin-skinned deformation is mostly active in the footwall of the frontal thick-skinned thrust, and abandoned once a newer, more distal thick-skinned thrust is formed. During the irst 5 My in M2, extension has resulted in a narrow, symmetric rit bounded by a conjugate pair of shear zones in the upper crust and lithospheric mantle. Inversion occurs symmetrically along these inherited shear zones, upliting the central block and over-inverting the shear zones past 6.25 My. A new shear zone is formed on one side ater 15 km of convergence beyond full inversion, creating an asymmetry that determines the plate polarity. he retro-wedge is abandoned and all shortening is accommodated by subduction of the lower plate. Ater 65 km of convergence, a new pro-wedge thrust sheet is accreted, upliting the older, upper pro-wedge thrust sheet and reactivating the retro-wedge. More pro-wedge thrust sheets are accreted in-sequence every ~33 km, while older, upper thrust sheets slowly migrate onto the upper plate. Most thick-skinned pro-wedge thrusts have a forelandward vergence, but one backthrust is created at 97.5 km of convergence. Continued convergence creates a doubly-vergent orogen with a wide thick-skinned pro-wedge and well-developed retro-wedge (Figure 4.2b). he hinterland of the pro-wedge rests on top of the upper plate. Like in M1, thin-skinned deformation is only active in the footwall of the frontal thick-skinned thrust. he irst 10 My of M3 are identical to M2, except the plate polarity happens to be reversed. he rit is inverted symmetrically, followed by an asymmetric phase with an abandoned retro-wedge ater 15 km of convergence beyond full inversion. Sedimentation is started at 10 My, illing the pro- and retro-foreland with ~4 and ~1 km of sediment, respectively. Ater 65 km of convergence, a new thick-skinned pro-wedge thrust sheet is accreted in sequence every ~34 km. Older, upper thrust sheets migrate onto the upper plate, reactivating the retro-wedge. Four out of a total of 10 thick-skinned pro-wedge thrusts develop with a hindward vergence, at 100, 120, 230, and 287.5 km of convergence. Ater 300 km of convergence, a doubly- vergent orogen is formed, with a wide, thick-skinned pro-wedge overlain by a thick sedimentary cover and a well-developed retro-wedge (Figure 4.2c). hrust sheets, both thick-skinned and thin-skinned, appear longer than in M2 but this cannot be measured with conidence due to the high variability in thrust sheet length within each model. hin-skinned deformation is only active in the footwall of the frontal thick- skinned thrust and is mostly abandoned once a new thick-skinned frontal thrust is formed. Model M4 adds erosion. he evolution is similar to previous models, with symmetric rit inversion and the onset of asymmetry at 15 km of convergence beyond full inversion. At 10 My, the sedimentary cover of the uplited central block is almost fully eroded, and sedimentation ills the pro- and retro-foredeep with ~4 and ~1 km of sediment, respectively. From 67.5 km of convergence onwards, pro-wedge thrust sheets are accreted in sequence every ~36 km of convergence. Older-upper pro-wedge thrust sheets are uplited and migrate onto the upper plate, reactivating the retro-wedge, while also eroding slowly. Ater 300 km of convergence, this creates a narrower doubly vergent orogen than M2 and M3. Basement is exposed in the centre of the orogen. he thin-skinned wedges are also smaller. M5 is the only asymmetric model, with a shale décollement on the let side and no décollement on the right side. he efects of this asymmetry are already apparent in the riting stage, where the rit becomes slightly asymmetric, accommodating more extension along the let-hand (décollement side) shear zone. Ater symmetry is almost fully restored at 6.25 My during inversion, the other shear zone (non-décollement side) becomes dominant. At 10 km of convergence beyond full inversion, a new shear zone is formed on the décollement side, creating the deinitive asymmetry with a décollement in the pro-wedge (Figure 4.3a) and no décollement in the retro-wedge (Figure 4.3b). Further shortening is accommodated by subduction of the lower plate while the retro-wedge extends slightly. Ater 70 km of convergence, a new pro-wedge thrust sheet is accreted every ~36 km, as the older, upper pro-wedge thrust sheets slowly migrate onto the upper plate and the retro-wedge is very slowly reactivated (Figure 4.3c). At 237.5 km, a new thick-skinned 80 Chapter 4. Numerical modelling of strain partitioning

M5: extension + asymmetric shale pre-orogenic 6 6 shale a b sediments 0 0 km km

14 14 400 470 525 600 c t = 19 My; ∆xC = 125 km 0 350° C upper crust lower crust 40 550° C km lithospheric mantle 80

d t = 36 My; ∆xC = 300 km 0 350° C

40 km 550° C

80 0 100 200 300 400 500 600 km

Figure 4.3. Model results of M5, which has an asymmetric shale décollement layer. (a) and (b) are zooms

of (c). Convergence (ΔxC) was measured from restored crustal thickness onwards. he white line indicates the tracked points that measure the distribution of shortening. retro-wedge thrust is formed (Figure 4.3d). Ater 300 km of convergence, this creates a doubly vergent orogen with frontal accretion in both the pro- and retro-wedge. hin-skinned deformation is not present in the retro-wedge, owing to the lack of a décollement layer. 4.3.1.2 Set 2: Salt décollement models he salt décollement in M6 allows the sedimentary cover to slide into the rit during the irst 5 My, exposing the rit shoulders. he rit is then symmetrically inverted along the two inherited shear zones, past full inversion at 6.5 My. As the keystone is uplited, the cover slides of towards both sides under the inluence of gravity. At 15 km beyond full inversion, a new shear zone is created, establishing an asymmetry and abandoning the retro-wedge (Figure 4.4a). All shortening is accommodated by subduction of the lower plate until 80 km beyond full inversion. During this period, the sedimentary cover of the lower plate is stacked in the pro-foreland. A new thrust sheet is accreted into the thick-skinned pro-wedge at 80 km beyond full inversion, upliting the upper, older thrust sheet and keystone block, and reactivating the retro-wedge. he stack of thin-skinned thrust sheets slides of into the pro-foreland as the new thrust sheet is uplited (Figure 4.4b). As the keystone and upper pro-wedge thrust sheets slowly migrate onto the upper plate, thin-skinned thrust sheets are stacked in front of the basement blocks in the retro-wedge (Figure 4.4c). he process of new pro-wedge thrust sheets upliting and the cover sliding into the foreland repeats every ~43 km of convergence, creating long thick-skinned pro-wedge thrust sheets with large ofsets (Figure 4.4d). Two new thick-skinned retro-wedge thrusts are formed at 245 and 285 km of convergence, respectively. Ater 300 km of convergence, a doubly vergent orogen with a narrow thick-skinned pro- wedge and a wide thick-skinned retro-wedge is formed. he highly mobile cover has allowed wide and thick thin-skinned fold-and-thrust belts to form in both forelands, the pro-foreland belt being the largest. During the irst 10 My, M7 develops very similar to M6. he rit is symmetrically inverted, with the sedimentary cover sliding of the keystone block ater reaching full inversion of the inherited shear zones at 6.25 My. At 17.5 km beyond full inversion, a new thrust is formed, creating an asymmetry and abandoning the retro-wedge. he sedimentary cover of the pro-wedge slides into the pro-foreland. However, as soon Chapter 4. Numerical modelling of strain partitioning 81

M6: extension + salt a t = 10 My; ∆xC = 35 km 0 350° C upper crust lower crust 40 550° C km lithospheric mantle

80 b c pre-orogenic 4 4 salt sediments 0 0 km km

14 14 750 680 619 544 d t = 19 My; ∆xC = 125 km 0 350° C

40

km 550° C

80

e t = 36 My; ∆xC = 300 km 0 350° C

40

km 550° C

80 0 100 200 300 400 500 600

Figure 4.4. Model results of M6, which has a viscous salt décollement. (b) and (c) are zooms of (d).

Convergence (ΔxC) was measured from restored crustal thickness onwards. he white line indicates the tracked points that measure the distribution of shortening. as sedimentation is started at 10 My (37.5 km), the pro-foreland thin-skinned deformation front jumps outward to a more distal position, forming a wide wedge-top basin. Ater 80 km of convergence beyond full inversion, a new pro-wedge thrust is formed, upliting the older, upper thrust sheet. he retro-wedge is reactivated and the sedimentary cover of the pro-wedge slides downwards into the pro-foreland. his is associated with another forelandward jump of the pro-foreland thin-skinned deformation front, creating a new wedge-top basin (Figure 4.5a). he upper pro-wedge thrust sheet and keystone block migrate onto the upper plate, forming a narrow thin-skinned retro-foreland fold-and-thrust belt in front (Figure 4.5b). he new pro-wedge thrust sheet at 80 km is quickly followed by another at 97.5 km of convergence. Gravitational sliding has exposed basement in the core of the orogen (Figure 4.5c). New pro-wedge thrusts form in pairs, alternating between a long interval (>60 km) and a short interval (<20 km), forming an average of ~45 km of convergence between thrusts. his forms a doubly vergent orogen with a thick-skinned pro-wedge that consists of alternating wide and narrow thrust sheets, and a well-developed retro-wedge. A very wide thin- skinned pro-foreland fold-and-thrust belt is characterised by wide wedge-top basins separated by narrow zones where thin-skinned deformation is concentrated (Figure 4.5d). Erosion is included in model M8. he early evolution is almost identical to other models with a salt décollement: symmetric inversion of the rit, reaching full inversion at 6.25 My. he cover slides of the keystone block as it is uplited, and at 17.5 km beyond full inversion a new thrust is created that determines the asymmetry and abandons the retro-wedge. At 10 My (37.5 km) sedimentation is started, illing the pro- foreland and triggering the thin-skinned deformation front to jump outward in the pro-foreland, creating a wide wedge-top basin between two zones of intense deformation (Figure 4.5e). At 80 km of convergence, 82 Chapter 4. Numerical modelling of strain partitioning

M7: extension + salt + sedimentation a 6 pre-orogenic syn-orogenic b 6 salt sediments sediments 0 0 km km

14 14 310 380 550 625 c t = 19 My; ∆xC = 125 km 0 350° C upper crust lower crust 40 550° C km lithospheric mantle 80

d t = 36 My; ∆xC = 300 km 0 350° C

40 km 550° C

80 0 100 200 300 400 500 600

M8: extension + salt + sedimentation + erosion e 6 f 6

0 0 km km

14 14 380 450 550 625 g t = 19 My; ∆xC = 125 km 0 350° C

40 km 550° C

80

h t = 36 My; ∆xC = 300 km 0 350° C

40 km 550° C

80 0 100 200 300 400 500 600

Figure 4.5. Model results of M7 and M8, which are part of set 2 with a viscous salt décollement. (a)

and (b) are zooms of (c). (e) and (f) are zooms of (g). Convergence (ΔxC) was measured from restored crustal thickness onwards for both models. he white line indicates the tracked points that measure the distribution of shortening. Points above topography are ignored and coloured red. a new thick-skinned pro-wedge thrust is formed. he retro-wedge is reactivated as the older, upper pro- wedge thrust sheet and keystone block migrate onto the upper plate and are gradually eroded (Figure 4.5f). A new pro-wedge thrust follows quickly at 97.5 km of convergence (Figure 4.5g). Further pro-wedge thrust sheets are accreted in the same alternating pattern of long interval (>60 km), short interval (<20 km), etc. Chapter 4. Numerical modelling of strain partitioning 83

As each pro-wedge thrust sheet is uplited, the cover slides down into the pro-foreland and upper thrust sheets are partially eroded. At 252.5 km of convergence, a new thick-skinned thrust is formed in the retro- wedge. 300 km of convergence have created a narrower orogen compared to M7, with a relatively narrow thick-skinned pro-wedge with alternating long and short thrust sheets, and a well-developed retro-wedge (Figure 4.5h). A wide pro-foreland fold-and-thrust belt has wide basins separated by narrow zones of intense deformation. Erosion has exposed basement along a signiicant portion of the central orogen.

a 70 Strain distribution M1 no surface processes M2 50 High Atlas M5 M6 Central Alps

20 Pyrenees Western Alps

retro-wedge shortening (%) 0 0 100 200 300 350

∆xc (km) b 70 Strain distribution M1 with surface processes M3 50 High Atlas M4 start M7 sedimentation Central Alps M8 Pyrenees 20 Western Alps

retro-wedge shortening (%) 0 0 100 200 300 350

∆xc (km) c 100 Shortening rate distribution (M4)

80

pro-wedge

50

retro-wedge

20 fraction of shortening rate (%)

timing of new pro-wedge thrusts 0

0 100 200 300 350

∆xc (km)

Figure 4.6. (a) plots of the distribution of shortening between the lower (pro) and upper (retro) plate for models with no surface processes. Strain distributions for the High Atlas, Pyrenees, Western Alps and Central Alps from [Teixell et al., 2003; Beaumont et al., 2000; Schmid and Kissling, 2000; Schmid et al., 1996]. (b) plots of the distribution of shortening between the lower and upper plate for models with surface processes, and M1 as reference. (c) Distribution of shortening rates as percentage of the total convergence rate between the lower and upper plate. his plot shows the distribution for M4 as an example. All models feature a similar evolution of the distribution through time. 84 Chapter 4. Numerical modelling of strain partitioning

4.3.2 Strain distribution analysis Here we describe the quantiied strain distribution between the pro- and retro-wedge in our models. hese results are plotted against convergence, not time, to eliminate any diferences in convergence rates when comparing to natural systems later on. Our irst observation is that the strain distribution varies through time, but eventually tends to stabilise ater 75 to 125 km of convergence (Figure 4.6a). he shortening distribution in M1, without rit inheritance, rapidly shits in favour of the pro-wedge, stabilising at only ~6% of shortening accommodated in the retro-wedge (Figure 4.6a). Every model that includes rit inheritance has accommodated more shortening in the retro-wedge than M1. his conirms that rit inheritance is essential to promoting retro-wedge shortening. he asymmetric décollement distribution we tested (M5) inluences the polarity of subduction, favouring a décollement in the lower plate. As a result, the retro-wedge does not have a décollement layer and shortening is decreased to ~11% of total convergence. his is lower than other models that also include rit inheritance. When surface processes are not included, the décollement rheology makes a signiicant diference in the shortening distribution. A shale décollement results in the retro-wedge accommodating ~18% of total convergence, while a salt décollement allows accommodating ~24% of total convergence in the retro-wedge (Figure 4.6a). However, when sedimentation and erosion are included, the diference in strain distribution between a shale décollement and a salt décollement is greatly reduced (Figure 4.6b). he strain distribution varies a few percent between these models, but they all accommodate 16-20% of total convergence in the retro-wedge. Erosion appears to systematically result in slightly lower retro-wedge shortening, but this diference is small and may be within the error margin. We have also quantiied the shortening rates of the pro- and retro-wedge in our models (Figure 4.6c). Shortening rates of the pro- and retro-wedge are not constant, even during the later stages of orogeny when the shortening distribution has stabilised. On a larger timescale, the shortening rates reveal a two- phase evolution in pro-retro strain partitioning: During the symmetric rit inversion phase, shortening is partitioned roughly 50/50 between both sides. he retro-wedge is then temporarily abandoned during the transition to the asymmetric collision phase. Later in the asymmetric collision phase, the retro-wedge is reactivated but the shortening rate remains low. Shortening rates also vary on a shorter timescale: Every time a new pro-wedge thrust is formed, the pro-wedge shortening rate is increased briely. he retro-wedge shortening rate lowers accordingly at these times. Events in the pro-wedge thus inluence the shortening rate of the retro-wedge and vice versa. he two are inextricably linked.

No extensional inheritance + shale

thin-skinned a

thick-skinned UC pro-wedge LC LM asymmetric orogenic wedge

Extensional inheritance Extensional inheritance Shale décollement Salt décollement

b d

thick-skinned retro-wedge wide bivergent wedge narrow bivergent wedge + gravitational sliding

Inheritance + Shale + Erosion + Deposition Inheritance + Salt + Erosion + Deposition E E c S S e S S

narrow bivergent wedge efficient decoupling of FFTB + narrow bivergent wedge

Figure 4.7. Summary of model behaviour as a result of the tested factors. Chapter 4. Numerical modelling of strain partitioning 85

4.4 Discussion 4.4.1 Individual factors Rit inheritance enables an orogen to become doubly vergent, by creating a weak shear zone through the retro-wedge that facilitates shortening [Jammes and Huismans, 2012; Erdős et al., 2014; see also Chapter 3]. Our quantiied results conirm that rit inheritance has a irst order inluence on creating a doubly vergent orogen. he inherited retro-shear zone is easily reactivated during the inversion phase, and later easily reactivated again during the collision phase. his results in a doubly vergent orogen, rather than a singly vergent orogen without rit inheritance (Figure 4.7a). Our quantiied results show that the décollement rheology can make a diference in the pro-retro strain partitioning and crustal structure (Figures 4.6a and 4.7). A shale décollement enables thin-skinned foreland fold-and-thrust belts to form. However, because the shale décollement does not allow gravitational sliding, the sedimentary cover remains on top of the thick-skinned thrust sheets, promoting a wide thick-skinned orogen (Figure 4.7b). In contrast, a salt décollement promotes gravitational sliding of the sedimentary cover. his unloads the core of the orogen and increases overburden over the distal pro- and retro-wedge, resulting in a narrower thick-skinned orogen (Figure 4.7c). he salt décollement also reduces friction along the retro shear zone, which, combined with the weight of the pro-foreland fold-and-thrust belt results in more shortening being accommodated in the retro-wedge (Figure 4.6a). For maximum efect, the décollement must be weak enough for gravitational sliding to occur, but not so weak that the resulting fold- and-thrust belt becomes very wide and thin. his is conirmed by models with a weaker salt décollement ( = 1018 Pa∙s) that we describe in Appendix C. An asymmetric décollement distribution inluences the plate polarity, preferentially placing the décollement in the lower plate (Figure 4.3). his can be explained by comparing the integrated strength of the crust on both sides of the model. he décollement lowers the integrated strength of the crust. hus, all else being equal, when a new shear zone is formed ater inversion of the rit, it is highly likely that it forms on the décollement side. his new shear zone is what determines the plate polarity and the décollement thus ends up in the lower plate. he lack of a décollement in the upper plate results in increased friction along retro- wedge shear zones compared to the pro-wedge, afecting the pro-retro strain partitioning (Figure 4.6a). Without a décollement the structure of the retro-wedge is diferent, with reduced fault throws and earlier frontal accretion compared to a retro-wedge with a décollement (Figure 4.3). With these results, we can speculate about the efects of other possible décollement distributions. he plate polarity is only inluenced by the décollement distribution around the plate boundary. A laterally limited décollement will alter the properties of both wedges once they reach the edge of the décollement, potentially inluencing the pro- retro strain partitioning, favouring the side with the largest proportion of décollement. Because thickness of a viscous décollement is a key parameter in determining the cover mobility, a thicker décollement will further facilitate gravitational sliding and reduce taper of the resulting thin-skinned wedge while a thinner décollement will have the opposite efect. he inluence of surface processes on the crustal structure is well-documented [e.g., Buiter, 2012; Erdős et al., 2014, 2015]. Sedimentation results in longer thrust sheets by increasing the overburden of the distal pro-wedge and erosion unloads the core. Together these promote shortening of the hinterland, resulting in a narrower thick-skinned orogen (Figure 4.7c, d) [Erdős et al., 2015]. 4.4.2 Interactions Although surface processes do not appear to inluence the pro-retro strain partitioning when combined with a shale décollement, they do have an inluence when combined with a salt décollement. In fact, combining surface processes and a salt décollement results in a very similar pro-retro strain partitioning as a shale décollement model. Surface processes thus efectively counteract the efect of a salt décollement on the pro-retro strain partitioning. he results show that sedimentation without erosion already has this efect. Sedimentation stabilises the sedimentary cover in the salt décollement models while still allowing gravitational sliding (Figure 4.5). One would expect the greater mass of sediment in the pro-foreland to have a greater efect on the strain partitioning than the relatively little amount of sediment in the retro- 86 Chapter 4. Numerical modelling of strain partitioning

foreland, but this is not the case. It is unclear why sedimentation results in a reduction of retro-wedge shortening relative to a salt décollement without sedimentation instead. he evidence that the thick-skinned structure can be inluenced by surface processes changing and redistributing the loads on top of an orogenic wedge has been steadily growing [e.g., Sinclair et al., 2005; Fillon et al., 2013]. hin-skinned deformation can also redistribute signiicant loads, but the relationship between thin-skinned and thick-skinned deformation is usually considered a one-way street, with thick- skinned deformation driving thin-skinned deformation without any feedback [e.g., Bergh et al., 1997; Molinaro et al., 2005]. Our models show that the décollement rheology and resulting thin-skinned structure have a clear efect on thick-skinned structure. hin-skinned deformation is still driven by uplit of thick-skinned thrust sheets, but if the décollement is weak enough to enable signiicant gravitational sliding, thin-skinned deformation can inluence the thick-skinned structure. Gravitational sliding of the sedimentary cover results in a narrower thick-skinned pro-wedge. Our results allow us to link the evolution of the pro- and retro-wedge in two ways. According to the original deinition of the pro- and retro-wedge [Willett et al., 1993], material migrates from the pro-wedge to the retro-wedge. Using our deinition instead, the hinterland of the pro-wedge migrates onto the upper plate during the asymmetric collision phase. his migration reactivates the retro-wedge. hus, with either deinition, growth of the pro-wedge drives retro-wedge shortening during the asymmetric collision phase, but not the other way around. Shortening rates of both wedges are also linked, but this link can work in both directions. he shortening rates for the pro- and retro-wedge are complementary, thus if the shortening rate of one wedge increases, the shortening rate of the opposing wedge decreases by an equal amount (Figure 4.6c). In our models, the most common cause for this is creation of a new pro-wedge thrust. Every new pro-wedge thrust causes a temporary increase in the pro-wedge shortening rate and an equal decrease in the retro-wedge shortening rate. A similar relationship has been observed in analogue models of bivergent wedges [Hoth et al., 2007]. In addition to new pro-wedge thrusts, new retro-wedge thrusts can have the same efect. Phrased more generally, any event or condition that afects the shortening rate of one wedge has an immediate, equal and opposite efect on the shortening rate of the opposing wedge (assuming shortening rate is measured as a percentage of the total convergence rate). his relationship is a natural consequence of the force balance that must always exist within the orogen as a whole. 4.4.3 Natural systems Our models consistently reproduce a two-phase evolution when rit inheritance is involved. his allows us to propose a generic evolutionary model for inverted rit systems: Phase 1 is the roughly symmetric inversion of the rit. Phase 2 is the asymmetric main collision phase, where continental subduction forces ~80% of total convergence to be accommodated in the pro-wedge. Here we compare several natural systems to this proposed model. 4.4.3.1 High Atlas he High Atlas is an ENE-trending intracontinental mountain belt in North Africa. he Middle Atlas branches of the High Atlas in a NE direction. Here we will discuss a cross section through the Central High Atlas to the east of this branch intersection (Fig. 4.8a) [Arboleya et al., 2004; Ayarza et al., 2005]. he High Atlas evolved from a rit that opened during the Mesozoic and was inverted during several Cretaceous and Cenozoic compressional phases [e.g., Mattauer et al., 1977; Jacobshagen et al., 1988; Beauchamp et al., 1996; Frizon de Lamotte et al., 2000; Teixell et al., 2003]. Overall shortening is very low, only ~26 km, accommodated on several partially inverted normal faults that dip towards the centre of the orogen [Teixell et al., 2003]. Using a boundary to separate north-verging from south-verging structures, shortening is partitioned equally between the north and the south (~13 km each). Given that the High Atlas is an inverted rit system, and shortening is very low and partitioned equally, we propose that the High Atlas is a natural analogue for phase 1 of our evolutionary model. Chapter 4. Numerical modelling of strain partitioning 87

a High Atlas

~13 km ~13 km

SSE South North NNW High Atlas Middle Atlas HA front HA front 0 0

50 ? ? 50 50 km

b Central Pyrenees (ECORS Pyrenees)

~128 km ~37 km

Axial Zone North S Pyrenean Aquitaine N Ebro basin Zone Basin 0 0 Iberian plate European plate

50 50 50 km

c Western Alps (ECORS CROP)

~102 km ~22 km

Penninic Briançonnais NW Molasse Chaînes front terrane Sesia SE Jura Insubric line basin subalpines 0 0 M. Blanc Apulian Valaisan European plate Ivrea plate mantle 50 50 km 50

d Central Alps (NFP-20 East)

~108 km ~56 km

Engadine line Insubric Valaisan Lecco N line Coltignone S Helvetic thrust nappes thrust 0 0 Aar Massif Milan thrust European plate Apulian plate

50 50 km 50

Legend pro retro Syn-tectonic sediments Upper crust (cover) Upper crust (basement) Lower crust

Figure 4.8. Summary of natural systems. he lower plate is shown in green and the upper plate in blue. Distributions of shortening were obtained with a basic area balancing that assumes a constant-thickness crust before orogenesis. (a) Central Pyrenees ECORS Pyrenees section, ater Muñoz [1992] and Beaumont et al. [2000]. (b) High and Middle Atlas, ater Arboleya et al. [2004] and Ayarza et al. [2005]. (c) Western Alps ECORS CROP section, ater Schmid and Kissling [2000]. (d) Central Alps NFP-20 East section, ater Schmid et al. [1996]. 88 Chapter 4. Numerical modelling of strain partitioning

4.4.3.2 Pyrenees he Pyrenees is a WNW-ESE trending mountain belt that formed due to the Alpine collision of the Iberian microplate and the European continent. his Cenomanian to Oligocene collision inverted a hyperextended rit system that exhumed mantle between the two plates [e.g., Jammes et al., 2009; Lagabrielle et al., 2010]. he present-day structure of the Pyrenees consists of a narrow northern retro-wedge, a south verging antiformal stack of Iberian basement thrust sheets, and a zone of south Pyrenean cover thrust sheets (Fig. 4.8b). Estimates of shortening for the Central Pyrenees vary, from ~90 km to ~165 km [Beaumont et al., 2000; Mouthereau et al., 2014] Of this higher estimate, ~128 km is accommodated in the pro-wedge and ~37 km in the retro-wedge. Estimates for the Eastern Pyrenees are lower, but show a similar strain partitioning: ~111 to ~125 km overall, with the lower estimate partitioned as ~92 km in the pro-wedge and ~19 km in the retro-wedge [Vergés et al., 1995, 2002; Grool et al., 2018]. A recent reconstruction of the pro-retro strain partitioning for the Eastern Pyrenees revealed a roughly symmetrical rit inversion phase, followed by an asymmetric main collision phase [Grool et al., 2018]. hese observations correspond very well to a transition from phase 1 to phase 2 of our evolutionary model. 4.4.3.3 Alps he Alps formed due to southward oceanic subduction of the Tethys Ocean followed by collision of the Apulian (south) and European (north) plates. Total convergence is estimated to be ~600 km over several phases [Schmid et al., 1996; 2008; Handy et al., 2010]. Here we compare only the post-suture or collisional phase (since ~35 Ma) in the Western and Central Alps. For a compilation of shortening estimates in the Alps, see Figure 4.9. In terms of overall structure, the Western Alps show a wide thick-skinned pro-wedge and a narrow thick-skinned retro-wedge (Fig. 4.7c) [Schmid and Kissling, 2000]. he Ivrea mantle has reached a very shallow position between the upper and lower plate. he Jura mountains form a distal thin-skinned pro- foreland fold-and-thrust belt detached along an evaporitic décollement. Shortening ater 35 Ma is ~124 km, partitioned as ~102 km in the pro-wedge and ~22 km in the retro-wedge. hese numbers appear to it with the strain partitioning predicted by our model. However, alternative shortening estimates result in a less compatible strain partitioning: One such alternative north of the Penninic front and to the north of this transect is 76-96 km [Burkhard and Sommaruga, 1998], but it is unclear how much of this shortening is the result of pre-35 Ma convergence. Even further northeast along the Alpine chain, Rosenberg and Kissling [2013] ind signiicantly lower ‘post-nappe’ shortening of 71 km and 17 km north and south of the Insubric line. he Central Alps comprise a large stack of basement-involved thrust sheets north of the Insubric line that largely predates the collision phase (Figure 4.8d) [Schmid et al., 1996] he southern half of this stack rests on top of the lower crust of the Apulian plate. his corresponds to the hinterland of the pro-wedge that migrated onto the upper plate in our models. he Apulian upper crust forms a relatively narrow thick-skinned stack with thin-skinned thrusts of the Milan thrust system linking back into it. he strain partitioning does not correspond to those predicted by any of our models. Shortening since 40 Ma is partitioned as ~108 km and ~56 km north and south of the Insubric line, respectively [Schmid et al., 1996]. When accounting only for shortening since 32 Ma, shortening is partitioned as ~63 km and ~56 km north and south of the Insubric line [Schmid et al., 1996]. An alternative estimate of ‘post nappe’ shortening is ~41 km and ~56 km, respectively [Rosenberg and Kissling, 2013]. None of these estimates are close to the 80/20 split predicted by our models. In fact, it appears that more recent shortening was distributed equally. Based on these estimates, we conclude that our evolutionary model for inverted rit systems cannot be applied to the Alps. he high convergence, multiple orogenic phases, and pre-existing asymmetry due to oceanic subduction inluence the pro-retro strain partitioning to such a degree that a diferent evolutionary model is required to explain the strain partitioning of the Alps. Other, higher shortening estimates for the area around this section exist [Roeder, 1992; Schönborn, 1992; Pifner et al., 2000], but either the timing of these estimates is unclear or the estimates only cover one half of the section (Fig. 4.9). Chapter 4. Numerical modelling of strain partitioning 89 46° 47° 48° 17° 16° 16° 15° 15° 14° 14° 13° Periadriatic line Periadriatic 13° 12° 12°

11° 11°

mid-late Miocene mid-late

100 km 100 Laubscher 1990 Laubscher

(post nappe) (post (post nappe) (post

Rosenberg & Kissling, 2013 Kissling, & Rosenberg

10-12 km 10-12

87 km 87 44°

10° km 70-100

Roeder 1992 Roeder 10°

>62 km (incl L Cret) L (incl km >62

Schönborn 1992 Schönborn

63 km (post 32 Ma) 32 (post km 63 56 km (post 32 Ma) 32 (post km 56

Schmid et al. 1996 al. et Schmid

(post-nappe) (post nappe) (post

9° km 41 56 km 56 Po basin Po Insubric line

17 km (post-nappe)

8° Schmid & Kissling, 2000 71 km Rosenberg & Kissling, 2013 (post-nappe)

22.5 km Penninic front Penninic (post 35 Ma)

40-60 km 7°

Rosenberg & Kissling, 2013Molasse basin >10 km

26 km

6° 101.5 km (post 35 Ma)

Burkhard & Sommaruga, 1998 Jura 6° 7° 8° 9° 5° 47° 48° 46° 45° 44°

Figure 4.9. Summary of available shortening estimates in the Western and Central Alps [Laubscher, 1990; Roeder, 1992; Schönborn, 1992; Schmid et al., 1996; Burkhard and Sommaruga, 1998; Schmid and Kissling, 2000; Rosenberg and Kissling, 2013]. 90 Chapter 4. Numerical modelling of strain partitioning

4.4.4 Limitations he models presented in this paper are simpliied systems that have some inherent limitations. Firstly, all models with extensional inheritance have a very simple rit structure. Natural rit systems generally have more complexity such as asymmetry or a higher number of normal faults that may inluence the orogenic structure when inverted. Secondly, our models do not feature full crustal breakup during the riting phase. However, earlier tests have shown that this has no noticeable efect on the orogenic structure [Erdős et al., 2014]. hirdly, our models feature no precursor subduction. As discussed above, this limits the applicability to orogens that did have oceanic subduction before collision. Fourthly, the décollement distribution and crustal strength are completely uniform in our models. his is a necessary simpliication to eliminate some variables and emphasise the fundamental behaviour and mechanisms. However, it prevents the models from exactly duplicating the structures and evolution of particular natural systems. hese models are a tool that helps us understand how natural systems work. Due to these limitations in the models, comparing them to natural systems also brings some limitations. For the Alps, we were only able to compare post-suture shortening. However, during oceanic subduction, the geometry of the Alpine domain already evolved to a geometry that is signiicantly diferent from our models [e.g., Schmid et al., 1996]. In addition, oblique convergence may limit the amount of reactivation of inherited structures that are not optimally oriented. Our 2D models cannot account for this.

4.5 Conclusions We have used high resolution numerical models to investigate the inluence of several factors, décollement rheology in particular, on the general crustal structure and strain partitioning between the pro-and retro- wedge. From our results we draw the following conclusions:

• he evolutionary pattern of an inverted rit system predicted by our models can be applied to natural orogens that also started as inverted rit systems. he High Atlas represents phase 1, roughly symmetrical inversion. he Pyrenees represent phase 2, asymmetric collision.

• he rheology of a décollement layer in the sedimentary cover inluences the crustal structure of the orogen, and may also inluence the strain distribution between the pro- and retro-wedge.

• he distribution of a décollement layer inluences the plate polarity and promotes shortening of the side where the décollement is situated (or thicker). A décollement that is far away from the area of initial deformation does probably not inluence this, as such a décollement would remain uninvolved until ater the plate polarity is established.

• he absence of a décollement altogether in the retro-wedge promotes early thick-skinned propagation of deformation into the upper plate.

• Sedimentation and erosion tend to negate the efects of the décollement rheology on the pro-retro strain partitioning, but not the efects on the crustal structure (width of the thick-skinned orogen).

• he pro- and retro-wedge are inextricably linked. Events and conditions on one side have an immediate efect on the other side, as a result from the force balance that must exist in the system.

References Allen, M. B., C. Saville, E. J.-P. Blanc, M. Talebian, and E. Nissen (2015), Orogenic plateau growth: Expansion of the Turkish-Iranian Plateau across the Zagros fold-and-thrust belt, Tectonics, 32(2), 171–190, doi:10.1002/tect.20025. Arboleya, M. L., A. Teixell, M. Charroud, and M. Julivert (2004), A structural transect through the High and Middle Atlas of Morocco, J. African Earth Sci., 39(3–5), 319–327, doi:10.1016/j.jafrearsci.2004.07.036. Chapter 4. Numerical modelling of strain partitioning 91

Ayarza, P., F. Alvarez-Lobato, A. Teixell, M. L. Arboleya, E. Tesón, M. Julivert, and M. Charroud (2005), Crustal structure under the central High Atlas Mountains (Morocco) from geological and gravity data, Tectonophysics, 400(1–4), 67–84, doi:10.1016/j.tecto.2005.02.009. Beauchamp, W., M. Barazangi, A. Demnati, and M. El Alji (1996), Intracontinental riting and inversion: Missour basin and Atlas mountains, Morocco, Am. Assoc. Pet. Geol. Bull., 80(9), 1459–1482, doi:10.1306/64ED9A60-1724-11D7-8645000102C1865D. Beaumont, C., J. A. Muñoz, J. Hamilton, and P. Fullsack (2000), Factors controlling the Alpine evolution of the central Pyrenees inferred from a comparison of observations and geodynamical models, J. Geophys. Res. Solid Earth, 105(B4), 8121–8145, doi:10.1029/1999JB900390. Bergh, S. G., A. Braathen, and A. Andresen (1997), Interaction of basement-involved and thin-skinned tectonism in the Tertiaru Fold-and-thrust belt of central Spitsbergen Svalbard, Am. Assoc. Pet. Geol. Bull., 4(4). Buiter, S. J. H. (2012), A review of brittle compressional wedge models, Tectonophysics, 530–531(40), 1–17, doi:10.1016/j.tecto.2011.12.018. Burkhard, M., and A. Sommaruga (1998), Evolution of the western Swiss Molasse basin: structural relations with the Alps and the Jura belt, Geol. Soc. London, Spec. Publ., 134(1), 279 LP-298. Chemia, Z., H. Koyi, and H. Schmeling (2008), Numerical modelling of rise and fall of a dense layer in salt diapirs, Geophys. J. Int., 172(2), 798–816, doi:10.1111/j.1365-246X.2007.03661.x. Dahlen, F. a. (1990), Critical Taper Model of Fold-And-hrust Belts and Accretionary Wedges, Annu. Rev. Earth Planet. Sci., 18(1), 55–99, doi:10.1146/annurev.ea.18.050190.000415. Erdős, Z., R. S. Huismans, P. van der Beek, and C. hieulot (2014), Extensional inheritance and surface processes as controlling factors of mountain belt structure, J. Geophys. Res. Solid Earth, 119, 9042–9061, doi:10.1002/2014JB011408. Erdős, Z., R. S. Huismans, and P. van der Beek (2015), First-order control of syntectonic sedimentation on crustal- scale structure of mountain belts, J. Geophys. Res. Solid Earth, 120, 1–16, doi:10.1002/2014JB011785. Fillon, C., R. S. Huismans, P. van der Beek, and J. A. Muñoz (2013), Syntectonic sedimentation controls on the evolution of the southern Pyrenean fold-and-thrust belt: Inferences from coupled tectonic-surface processes models, J. Geophys. Res. Solid Earth, 118(10), 5665–5680, doi:10.1002/jgrb.50368. Ford, M. (2004), Depositional wedge tops: Interaction between low basal friction external orogenic wedges and lexural foreland basins, Basin Res., 16(3), 361–375, doi:10.1111/j.1365-2117.2004.00236.x. Frizon de Lamotte, D., B. Saint Bezar, R. Bracène, and E. Mercier (2000), he two main steps of the Atlas building and geodynamics of the western Mediterranean, Tectonics, 19(4), 740–761. Gleason, G. C., and J. Tullis (1995), A low law for dislocation creep of quartz aggregates determined with the molten salt cell, Tectonophysics, doi:10.1016/0040-1951(95)00011-B. Grool, A.R., Ford, M., Vergés, J., Huismans, R.S., Christophoul, F., and Dielforder, A., 2018, Insights Into the Crustal-Scale Dynamics of a Doubly Vergent Orogen From a Quantitative Analysis of Its Forelands: A Case Study of the Eastern Pyrenees: Tectonics, v. 37, p. 450–476, doi: 10.1002/2017TC004731. Handy, M. R., S. M. Schmid, R. Bousquet, E. Kissling, and D. Bernoulli (2010), Reconciling plate-tectonic reconstructions of Alpine Tethys with the geological-geophysical record of spreading and subduction in the Alps, Earth-Science Rev., 102, 121–158, doi:10.1016/j.earscirev.2010.06.002. Hoth, S., A. Hofmann-Rothe, and N. Kukowski (2007), Frontal accretion: An internal clock for bivergent wedge deformation and surface uplit, J. Geophys. Res. Solid Earth, 112(January), B06408, doi:10.1029/2006JB004357. Jacobshagen, V., R. Brede, M. Hauptmann, W. Heinitz, and R. Zylka (1988), Structure and post-Palaeozoic evolution of the central High Atlas, Atlas Syst. Morocco SE - 15, 15, 245–271, doi:10.1007/BFb0011596. Jammes, S., and R. S. Huismans (2012), Structural styles of mountain building: Controls of lithospheric rheologic stratiication and extensional inheritance, J. Geophys. Res. Solid Earth, 117(B10), B10403, doi:10.1029/2012JB009376. Jammes, S., G. Manatschal, L. L. Lavier, and E. Masini (2009), Tectonosedimentary evolution related to extreme crustal thinning ahead of a propagating ocean: Example of the western Pyrenees, Tectonics, 28(4), TC4012, doi:10.1029/2008TC002406. Karato, S. -i., and P. Wu (1993), Rheology of the Upper Mantle: A Synthesis, Science, 260(5109), 771–778, doi:10.1126/science.260.5109.771. van Keken, P. E., C. J. Spiers, A. P. van den Berg, and E. J. Muyzert (1993), he efective viscosity of rocksalt: implementation of steady-state creep laws in numerical models of salt diapirism, Tectonophysics, 225(4), 457–476, doi:10.1016/0040-1951(93)90310-G. 92 Chapter 4. Numerical modelling of strain partitioning

Lagabrielle, Y., P. Labaume, and M. de Saint Blanquat (2010), Mantle exhumation, crustal denudation, and gravity tectonics during Cretaceous riting in the Pyrenean realm (SW Europe): Insights from the geological setting of the lherzolite bodies, Tectonics, 29(4), TC4012, doi:10.1029/2009TC002588. Laubscher, H. P. (1990), he problem of the deep structure of the Southern Alps: 3-D material balance considerations and regional consequences, Tectonophysics, 176(1–2), 103–121, doi:10.1016/0040- 1951(90)90261-6. Loury, C., Y. Rolland, S. Guillot, A. V Mikolaichuk, P. Lanari, O. Bruguier, and D. Bosch (2015), Crustal-scale structure of South Tien Shan: implications for subduction polarity and Cenozoic reactivation, Geol. Evol. Cent. Asian Basins West. Tien-Shan Range, 427(March), doi:10.1144/SP427.4. Mattauer, M., P. Tapponnier, and F. Proust (1977), Sur les mecanismes de formation des chaines intracontinentales; l’exemple des chaines atlasiques du Maroc, Bull. la Soc. Geol. Fr., S7–XIX(3), 521–526, doi:10.2113/ gssgbull.S7-XIX.3.521. Molinaro, M., P. Leturmy, J. C. Guezou, D. Frizon de Lamotte, and S. a. Eshraghi (2005), he structure and kinematics of the southeastern Zagros fold-thrust belt, Iran: From thin-skinned to thick-skinned tectonics, Tectonics, 24(3), 1–19, doi:10.1029/2004TC001633. Mouthereau, F., P.-Y. Filleaudeau, A. Vacherat, R. Pik, O. Lacombe, M. G. Fellin, S. Castelltort, F. Christophoul, and E. Masini (2014), Placing limits to shortening evolution in the Pyrenees: Role of margin architecture and implications for the Iberia/Europe convergence, Tectonics, 33(DECEMBER 2014), 2283–2314, doi:10.1002/2014TC003663. Muñoz, J. A. (1992), Evolution of a continental collision belt: ECORS-Pyrenees crustal balanced cross-section, in hrust Tectonics, edited by K. R. McClay, pp. 235–246, Springer Netherlands, Dordrecht, the Netherlands. Pifner, O. A., S. Ellis, and C. Beaumont (2000), Collision tectonics in the Swiss Alps: Insight from geodynamic modeling, Tectonics, 20(2), 288. Roeder, D. (1992), hrusting and wedge growth, Southern Alps of Lombardia (Italy), Tectonophysics, 207(1–2), 199–243, doi:10.1016/0040-1951(92)90478-O. Rosenberg, C. L., and E. Kissling (2013), hree-dimensional insight into central-alpine collision: Lower-plate or upper-plate indentation?, Geology, 41(12), 1219–1222, doi:10.1130/G34584.1. Saintot, A., M.-F. Brunet, F. Yakovlev, M. Sebrier, R. A. Stephenson, A. V. Ershov, F. Chalot-Prat, and T. McCann (2006), he Mesozoic-Cenozoic tectonic evolution of the Greater Caucasus, Geol. Soc. London, Mem., 32(May), 277–289, doi:10.1144/GSL.MEM.2006.032.01.16. Schmid, S. M., and E. Kissling (2000), he arc of the western Alps in the light of geophysical data on deep crustal structure, Tectonics, 19(1), 62, doi:10.1029/1999TC900057. Schmid, S. M., O. A. Pifner, N. Froitzheim, G. Schönborn, and E. Kissling (1996), Geophysical-geological transect and tectonic evolution of the Swiss-Italian Alps, Tectonics, 15(5), 1036–1064, doi:10.1029/96TC00433. Schmid, S. M., D. Bernoulli, B. Fügenschuh, L. Matenco, S. Schefer, R. Schuster, M. Tischler, and K. Ustaszewski (2008), he Alpine-Carpathian-Dinaridic orogenic system: Correlation and evolution of tectonic units, Swiss J. Geosci., 101(1), 139–183, doi:10.1007/s00015-008-1247-3. Schönborn, G. (1992), Kinematics of a transverse zone in the Southern Alps, Italy, in hrust Tectonics, edited by K. R. McClay, pp. 299–310, Springer Netherlands, Dordrecht. Sherkati, S., J. Letouzey, and D. F. De Lamotte (2006), Central Zagros fold-thrust belt (Iran): New insights from seismic data, ield observation, and sandbox modeling, Tectonics, 25(4), 1–27, doi:10.1029/2004TC001766. Sinclair, H. D., M. Gibson, M. Naylor, and R. G. Morris (2005), Asymmetric growth of the Pyrenees revealed through measurement and modeling of orogenic luxes, Am. J. Sci., 305, 369–406, doi:10.2475/ ajs.305.5.369. Teixell, A., M. L. Arboleya, M. Julivert, and M. Charroud (2003), Tectonic shortening and topography in the central High Atlas (Morocco), Tectonics, 22(5), doi:10.1029/2002TC001460. hieulot, C. (2011), FANTOM: Two- and three-dimensional numerical modelling of creeping lows for the solution of geological problems, Phys. Earth Planet. Inter., 188(1–2), 47–68, doi:10.1016/j. pepi.2011.06.011. Vergés, J., H. Millán, E. Roca, J. A. Muñoz, M. Marzo, J. Cirés, T. Den Bezemer, R. Zoetemeijer, and S. Cloetingh (1995), Eastern Pyrenees and related foreland basins: pre-, syn- and post-collisional crustal-scale cross- sections, Mar. Pet. Geol., 12(8), 903–915, doi:10.1016/0264-8172(95)98854-X. Vergés, J., M. Fernàndez, and A. Martínez (2002), he Pyrenean orogen: Pre-, syn-, and post-collisional evolution, J. Virtual Explor., 8, 55–74, doi:10.3809/jvirtex.2002.00058. Chapter 4. Numerical modelling of strain partitioning 93

Vogt, K., L. Matenco, and S. Cloetingh (2017a), Crustal mechanics control the geometry of mountain belts. Insights from numerical modelling, Earth Planet. Sci. Lett., 460, 12–21, doi:10.1016/j.epsl.2016.11.016. Vogt, K., E. Willingshofer, L. Matenco, D. Sokoutis, T. Gerya, and S. Cloetingh (2017b), he role of lateral strength contrasts in orogenesis: A 2D numerical study, Tectonophysics, doi:10.1016/j.tecto.2017.08.010. Willett, S. D., C. Beaumont, and P. Fullsack (1993), Mechanical model for the tectonics of doubly vergent compressional orogens, Geology, 21(4), 371–374, doi:10.1130/0091-7613(1993)021<0371:MMFTTO>2. 3.CO.

Chapter 5

Conclusions and perspectives 96 Chapter 5. Conclusions and perspectives

In Chapter 1 of this thesis, a number of questions related to the dynamic relationship of pro- and retro- wedges were outlined. In this chapter, the new insights gained from this work are summarised, and an attempt is made to answer these questions.

How does a doubly vergent orogen evolve? Detailed analysis of the Eastern Pyrenees has revealed that shortening was distributed roughly equally across the European and Iberian margins during early inversion of the rit. From the end of the Maastrichtian onwards, shortening was distributed asymmetrically in the Eastern Pyrenees. Around 80% of shortening was accommodated in the southern Pyrenees and the northern Pyrenees were temporarily abandoned during the Paleocene. From the hanetian onwards, the retro-wedge was reactivated, but the pro-wedge remained dominant. Because the change in shortening distribution occurred coeval with the estimated onset of main collision and subduction of the Iberian lower crust, it appears that this change is the result of an internal threshold (i.e. collision) being reached that is intrinsic to inverted rit systems. his change may thus have occcurred independent from external forcing (e.g., plate kinematics). his evolution from symmetric inversion to asymmetric collision is reproduced by our numerical models that include extensional inheritance. Quantitatively, the models also agree with the evolution as observed in the Eastern Pyrenees: early shortening is distributed 50/50, and shortening during main collision is distributed roughly 80/20 between both plates. A temporary abandonment of the retro-wedge is also observed in our models, but the amount of shortening accommodated by the lower plate during the upper plate’s quiescence is much higher in the models than reconstructed for the Eastern Pyrenees.

Do the pro- and retro-wedge interact? Based on the data from the Eastern Pyrenees, the pro- and retro-wedge do interact. Firstly, the change in shortening distribution between the pro- and retro-wedge that is observed in the Pyrenees can be regarded as interaction between both wedges. Secondly, the reactivation of the retro-wedge occurred around the same time as the onset of exhumation in the Axial Zone is recorded. According to the reconstruction in Chapter 2, shortening and subsidence of the northern Pyrenees during the Eocene was mainly driven by backthrusting of the Axial Zone onto the European plate. he pro-wedge thus drives shortening in the retro-wedge during this phase. he numerical models also reproduce this. However, an even more direct interaction between the pro- and retro-wedge can be seen in the numerical models: he pro- and retro-wedge do not grow at a constant rate, even at geological timescales. he shortening rate of the retro-wedge changes as soon as a new crustal thrust is formed in the pro-wedge. Unfortunately, interactions at these timescales are unlikely to be preserved in a natural system. However, theoreticaly, such interactions should indeed exist, because of the force balance that must exist in the system as a whole. Any event in one wedge thus has an immediate efect on the other wedge.

What factors inluence the shortening distribution, and how? In this work, extensional inheritance is the only factor that can dramatically alter the shortening distribution. he numerical models show that, of the investigated factors, only extensional inheritance has enough inluence to cause an orogen to become singly or doubly vergent. Surface processes and thin- skinned deformation can signiicantly alter the crustal structure, but the shortening distribution between the pro- and retro-wedge remains largely unafected.

Secondary questions What causes the Paleocene quiescence in the Pyrenees remains debatable. he hypothesis developed on the basis of the East Pyrenean study states that such a quiescent period is part of the transition from equal to Chapter 5. Conclusions and perspectives 97 asymmetric shortening, and is to be expected for all inverted rit systems. he viability of this mechanism was conirmed by the numerical models, which reproduced a similar temporary abandonment of the retro- wedge. A key implication of this mechanism is that shortening of the lower plate continues while the retro- wedge is inactive. his is in direct contrast to explanations that implicate external factors (plate kinematics, intraplate shortening) as the cause for the Pyrenean Paleocene quiescence, which require shortening to slow down or stop in both wedges simultaneously. Unfortunately, data on Paleocene shortening of the southern Pyrenees is sparse and oten ambiguous, and plate kinematic reconstructions already predict a Paleocene phase of slow convergence. Both mechanisms thus remain valid for the time being. A hybrid solution could also have been the case: through sheer (and literal) coincidence, the transition from symmetric to asymmetric shortening may have occurred around the same time as a slowing in convergence. hin-skinned deformation can inluence orogenic crustal sructure. he numerical models described in Chapters 3 and 4 show that thin-skinned deformation can inluence the crustal structure. When combined with syn-tectonic sedimentation and erosion, eicient decoupling of the sedimentary cover promotes formation of a crustal antiformal stack. he sedimentary décollement creates a system of two wedges that share a common top surface. he mid-crustal décollement can support a higher taper than the sedimentary décollement. his results in the cover wedge constantly collapsing to keep taper low, while the crustal wedge constantly deforms internally to increase taper. he high internal deformation of the crust is what creates the antiformal stack.

Perspectives he timing of the change in shortening distribution coeval with the onset of subduction of the lower plate, both in the Pyrenees and numerical models, suggests that this transition is intrinsic to inverted rit systems, and not driven by external factors. hat implies that a similar change should be observed in other inverted rit systems. his hypothesis should be tested further by performing a similar analysis in other natural systems. Perhaps a broader model can be developed, that can also be applied to orogens that inherited an asymmetry from oceanic subduction prior to collision. he method used in Chapter 2 to reconstruct a high resolution shortening rate history should be tested for sensitivities. On the whole it appears that the reconstruction for the Eastern Pyrenees is fairly representative for the whole orogen. It would be interesting to apply the same method to other cross sections further west to test this. If the method returns representative results along strike, it will allow more detailed results for the same amount of work as the commonly used method of stepwise restoration. Should the shortening reconstruction method prove robust enough, it could be used to prove gravitational sliding in convergent settings. To do this, the rate of horizontal displacement of the cover and basement should be compared. Cover thrust sheets move only if they are pushed from behind, or if the subhorizontal component of the normal force vector is large enough on its own (assuming an inclined surface, of course). hus, in a standard setting with no gravitational sliding, a cover thrust sheet can never outpace the basement. However, if the cover does move faster than basement, only gravity can be responsible. hat is the theory, at least. he practical reality of constraining shortening rates of basement thrust sheets will, no doubt, make this very diicult. In this work, movement of the cover thrust sheets was used to constrain basement thrusting, which precludes an analysis as described above. he dual approach of detailed study of a natural system in combination with testing factors and mechanisms with numerical models has revealed the crustal-scale dynamics of a complex orogenic system in more detail than before. he beneit of this dual approach goes further than simply testing ield hypotheses, a task models are not always suited for. Working with numerical models allows one to visualise deformation, see processes occur on a human timescale that makes them much more relatable and easier to understand. hese insights can then be applied to the interpretation of geological data. Vice versa, knowledge of natural systems and their characteristics informs the models. A similar approach of close collaboration between ield studies and modelling should thus continue, also in other tectonic settings.

Appendix A

Supplement to Chapter 2: Pyrenean case study 100 Appendix A. Supplement to Chapter 2: Pyrenean case study

Contents Figure A.1. A comparison of shortening rates through time from diferent authors and diferent parts of the Pyrenees. Table A.1. A stratigraphic table that correlates the PYRAMID stratigraphic groups used in Chapter 2 and the BRGM map units along the north Pyrenean cross section. Table A.2. Material constants for each lithology used in decompaction calculations Table A.3. Subsidence of the north Pyrenean foreland Table A.4. Subsidence of the south Pyrenean foreland Table A.5. Field measurements

Figure A.1 presents the (minimum) shortening rate in the Pyrenees through time. he data used to plot the shortening rate in the Eastern Pyrenees is already presented in Figure 2.8b, c in Chapter 2. he method used to obtain this data is described in Chapter 2. As a result of assuming a constant average rate for each structure, rather than gradually increasing and decreasing, some artefacts may be present in the data where two structures briely overlap in time. he spike around 47 Ma is probably such an artefact. Table A.1 shows a detailed correlation of stratigraphic units used in the paper and the BRGM 1:50,000 geological maps. Tables A.2 to A.4 present values used for, and results of the subsidence analysis of the northern and southern foreland basins. he methodology of the subsidence analysis is explained in Chapter 2. Table A.5 presents all ield data collected and used during this study. he measurements are divided into three categories, each shown in a separate tab: plane measurements (bedding, cleavage, etc.), fault measurements (normal, reverse, etc.) and lineation measurements (fold axis, striae, etc.).

7 a Western Pyrenees b Central Pyrenees (ECORS) c Eastern Pyrenees

6 this study

5 Teixell et al. (2016) Beaumont et al. (2000)

4 mantle mantle protocollision closure 3 closure

shortening rate (mm/y) full collision 2 ?

1 Mouthereau et al. (2014) mantle closure crustal shortening 0 Phase 1 Ph2 Phase 3 Ph4 Phase 1 Ph2 Phase 3 Ph4 Phase 1 Ph2 Phase 3 Ph4

90 80 70 60 50 40 30 20 90 80 70 60 50 40 30 20 90 80 70 60 50 40 30 20 Time (Ma) Time (Ma) Time (Ma)

Figure A.1. Shortening rates for the Western, Central and Eastern Pyrenees. West Pyrenean shortening rate from Teixell et al. [2016]. Central Pyrenean shortening rates from Beaumont et al. [2000], and Mouthereau et al. [2014]. East Pyrenean shortening rate from this work (Chapter 2). Appendix A. Supplement to Chapter 2: Pyrenean case study 101

Table A.1. Stratigraphy correlation of PYRAMID stratigraphic groups with BRGM map units in the northern Pyrenees along section S3.

Table A1 Stratigraphy correlation along S3

harmonised S3 Stratigraphic units on BRGM 1:50,000 geological maps Epoch Age PYRAMID stratigraphic groups stratigraphy 1076 Lavelanet 1058 Mirepoix 1036 Castelnaudary g2c g2c Oligocene rupelian g1-2 g2b g2b g1-2a g1-2a Priabonian e7 e7 e7 Carcassonne Group Bartonian e6 e6 e6 Molasse de Castelnaudary e6 upper e5b e5b e5b Lutetian Palassou Fm lower e3c-5 e5a Palassou Fm e5a Eocene upper e5a e3-4c e3m-s e3b-c Oyster Sandstone + Ypresian e3b lower Coustouge Group e3a-b e3 Turitella Marl + e3a-b Alveolina Limestone e3 e3-4 e3a1 Aude Valley Group e2bM e2bM Albas Fm upper Thanetian Rieubach e2bC e2bC Lecarla Fm Paleocene lower Group e2a e2a e2 Selandian e1 e1 Esperaza Fm e1 Danian C7b-e1 Aude Valley Group upper c7b-e1 C7b Alet Fm Maastrichtian lower C7aG Plantaurel c6b-7a C6bM Labarre Fm upper Group Campanian c6 C6 C6bG C5b-6 lower Petites Pyrénées Group Plagne Fm upper Upper C5a3 Santonian c5 C5 Cretaceous lower C5a2 C5a1 Coniacian Grey Flysch Group upper c3-4 C3-4 middle Turonian lower c1-3 C1-3 Cenomanian Conglomérat de upper n7c-d Freychenet middle n7b Albian lower n7a n5-7 n5-7 Black Flysch Group (flysch) (Urgonia n6bU upper n5-6 n) n6a2 n6a3 Aptian Lower n6a1 Cretaceous lower n5 upper n4b Bartonian lower n4a Hauterivian upper Mirande Group n1-4 Calcaires jaunes à Bryozoaires + Calcaires Valanginian lower n1-3 graveleux à Pfendérines + Calcaires roux en plaquettes + Calcaires à Tocholines et Berriasian Dasycladacées + Brèche-limite Tithonian Upper Kimmeridgian Jurassic Oxfordian Callovian jD Middle Bathonian Jurassic Bajocian Black Dolomite Group Jurassic Aalenian Toarcian I5-8 Lower Pliensbachian Jurassic Sinemurian I1-4 Hettangian Rhaetian Keuper t7-10 Keuper Fm Upper Norian Keuper Group Triassic Carnian Middle Ladinian Triassic Anisian Lower Olenekian Permo-Triassic r-t Triassic Induan sandstone

Table A.2. Values for decompaction calculationsa b c d Lithology ρ (kg/m3) c (1/km) φ0 (%) Shale 2500 0.51 63 Siltstone 2530 0.45 59 Sandstone 2600 0.27 49 Conglomerate 2600 0.30 50 Marl 2500 0.59 58 Limestone 2670 0.70 50 Gypsum 2310 0.90 50 ahese represent the ‘pure’ lithologies. Bulk values for each stratigraphic unit were assumed to be a weighted average based on the lithological composition of the stratigraphic unit. Values taken from Vergés et al. [1998]. bDensity cCompaction coeicient dInitial Porosity 102 Appendix A. Supplement to Chapter 2: Pyrenean case study f,j (mm/y) Subs. Rate Rate Subs. f Tectonic Tectonic Subs. (km) Subs. f (km) Total Subs. Subs. Total i φ0 (%) h c (1/km) g ρ (kg/m3) d,f (km) Sea Level T1 DR4 MRL1 d,f 0 0.20 2509 0.57 57 3.173 1.426 0.144 0.01 0.14 2573 0.35 52 0.329 0.109 0.020 0.01 0.12 2557 0.41 54 0.415 0.183 0.022 0.01 0.22 2642 0.56 50 0.019 -0.163 0.020 0.005 0.21 -0.205 -0.296 (km) Water Depth Water e a 65Lst 16Mrl Lithology 19Sst; 81Mrl19Sst; 0 0.18 2537 0.59 56 0.128 -0.045 0.040 24Cgl; 30Mrl (km) hickness hickness d Age (Ma) c e1 59.2 0.058 36Lst 55Mrl; 9Sst; 0 0.15 2571 0.60 54 3.884 1.749 0.002 c6 75 2.588 1Lst 92Mrl; 7Sst; e3 51.9 0.310 18Lst 76Mrl; 6Sst; 0.005 0.17 2519 0.53 56 0.858 0.376 0.060 e2a 57.5 0.165 100Lst 0.005 0.16 2670 0.70 50 3.971 1.762 0.008 e2a 57.5 0.211 100Lst 0.005 0.16 2670 0.70 50 0.567 0.225 0.055 c7b 66 0.576 5Lst 87Mrl; 8Sst; 0 0.16 2517 0.57 57 3.819 1.719 0.062 c7b 66 0.192 91Mrl 9Sst; 0 0.16 2569 0.59 54 0.361 0.117 0.054 Unit e2bC 56.5 0.052 100Lst 0 0.16 2670 0.70 50 3.987 1.756 -0.006 e3a-b 54 0.087 33Lst 55Mrl; 12Sst; 0.005 0.17 2509 0.56 57 0.630 0.249 0.017 e2bM 56 0.118 22Lst 47Mrl; 31Sst; 0 0.17 2569 0.51 53 4.043 1.771 0.031 c7b-e1 61.6 0.038 100Lst 0 0.15 2670 0.70 50 3.856 1.745 0.006 BRGM BRGM unconf. 83.6 - - 0.15 0.23 - - - 0.492 0.188 0.052 unconf. 56 - - 0.005 0.17 - - - 0.560 0.215 -0.007 unconf. 59.2 - - 0 0.15 - - - 0.371 0.132 0.002 basement 100.5 Subsidence of the north Pyrenean foreland the Pyrenean north of Subsidence b PG c6bG 69 0.295 46Mrl 54Sst; 0 0.18 2554 0.42 53 3.405 1.534 0.018 PG c6b-7a 69 0.223 RG RG RG CG CG PPG CSG e6 37.8 0.072 29Sst; 16Slt; 2Shl; CSG e5c 41.2 0.119 7Cgl; 64Sst; 16Slt; GFGGFG c3-4 c1-3 86.3 93.9 0.190 0.128 14Cgl; 18Sst; 4slst; 100Mrl 0.01 0.23 2500 0.59 58 0.352 0.048 0.028 AVG AVG AVG AVG AVG Strat. Strat. Group Table A.3. Table Appendix A. Supplement to Chapter 2: Pyrenean case study 103 , 2016]. Rougier et al. et Rougier , 2016; hole Syn1. Locations in Figure 2.4. Locations Syn1. hole Ford et al. et Ford glomerate; Mrl, marl; Lst, limestone; Gps, gypsum. Gps, limestone; Lst, marl; Mrl, glomerate; , 1988b]. ude Valley Group; RG, Rieubach Group; CG, Coustouge Group; CSG, Group; CG, Coustouge Group; Rieubach RG, Group; Valley ude Bilotte et al. et Bilotte Syn1 0 0.17 2557 0.40 54 0.001 -0.130 0.024 0 0.16 25170 0.48 54 0.04 0.194 2557 0.007 0.48 0.020 55 1.935 1.015 0.010 0 0.08 2611 0.67 53 1.891 0.957 0.026 , 1976; 0.01 0.12 2591 0.35 50 1.739 0.850 0.073 0.005 0.17 -0.160 -0.230 0.005 0.16 2670 0.70 50 0.381 0.113 0.065 Cavaillé , 1975a; 7Lst 3Gps 10Lst 8Cgl; 20Mrl; 1Lst 8Cgl; 20Mrl; 7Cgl; 20Mrl; 15Gps 7Cgl; 20Mrl; Cavaillé et al. et Cavaillé e2a 57.5 0.162 100Lst e3-4 51.9 0.149 40Sst; 26Slt; 5Shl; e3a-b 51.9 0.213 33Lst 55Mrl; 12Sst; 0 0.17 2569 0.59 54 0.624 0.237 0.033 c7b-e1 61.6 0.182 91Mrl 9Sst; 0 0.15 2509 0.56 57 0.172 0.010 0.034 unconf. 56 - - 0.005 0.17 - - - 0.374 0.103 -0.007 basement 56 PG unconf. 59.2 - - 0 0.15 - - - 0.167 0.002 -0.003 RG CG CG CSG e5a 45 0.180 30Sst; 26Slt; 2Shl; CSG e7 33.9 0.179 69Lst; 14Mrl; 14Shl; CSG g1-2 28.1 0.014 40Cgl; 10Shl; 40Mrl; CSG e6 38 0.550 13Cgl; 67Sst; 13Mrl; CSG e5b 41.2 0.407 83Cgl; 8Shl; 8Lst 0.01 0.14 2598 0.35 51 1.296 0.615 0.054 CSG e5a 45 0.262 18Lst 76Mrl; 6Sst; 0 0.16 2537 0.59 56 0.922 0.411 0.025 AVG Bulk compaction coeicient Bulk compaction At top of unit. of top At Using ANR PYRAMID harmonised stratigraphic framework for central and eastern north Pyrenees (Figure 2.3b) [ north Pyrenees eastern (Figure and central for framework stratigraphic ANR PYRAMID harmonised Using Bulk density Measured in boreholes Dr4, T1, MRL1 (Bureau Exploration-Production des Hydrocarbures; beph.net), and synthetic bore synthetic and beph.net), des Hydrocarbures; Exploration-Production Dr4, T1, MRL1 (Bureau in boreholes Measured Composition as percentages of ‘pure’ lithologies. Shl, shale; Slt, siltstone; Sst, sandstone; Cgl, con sandstone; Sst, siltstone; Slt, shale; Shl, lithologies. ‘pure’ of percentages as Composition From BRGM 1:50000 geological maps [ 1:50000 geological maps BRGM From Relative to present-day sea level. present-day to Relative Initial porosity Initial Tectonic subsidence rate subsidence Tectonic GFG, Grey Flysch Group; PPG, Petites Pyrénées Group; PG, Plantaurel Group; AVG, A AVG, Group; Plantaurel PG, Pyrénées Group; Petites PPG, Group; Flysch Grey GFG, Carcassonne Group. a b c f g h i j d e 104 Appendix A. Supplement to Chapter 2: Pyrenean case study

Table A.4. Subsidence in the south Pyrenean forelanda Stratigraphic Age hickness Water Depth Sea Level ρ Total Subs. Tectonic Subs. Rate Name (Ma)b (km) (km)b,c (km)b,c (kg/m3)d (km)c Subs. (km)c (mm/y)e Lower Pedraforca equivalentf Paleocene 56 0.39 0 0.165 2600 2.025 0.928 0.009 Maastrichtian 66 1.07 0 0.160 2550 1.800 0.839 0.080 Campanian 72.1 0.64 0.01 0.185 2550 0.835 0.354 0.049 Santonian 83.6 0.06 0.01 0.230 2550 -0.060 -0.220 0.025 Cenomanian 86.3 0.03 0.01 0.230 2550 -0.160 -0.280 0.001 pre-Cenom. 100.5 0.01 0.210 -0.200 -0.300

Gombréng Bellmunt 42.65 1.2034 -0.05 0.148 2560 4.283 1.842 0.078 Coubet 45.82 0.1268 0.015 0.165 2520 3.577 1.593 -0.038 Beuda 46.18 0.17 0.05 0.166 2390 3.537 1.607 -0.142 Campdevánol 46.85 0.7232 0.2 0.170 2540 3.552 1.702 0.098 Armáncies 49.36 1.0766 0.1575 0.173 2520 2.924 1.456 0.531 Corones 50.69 0.2256 0.005 0.171 2600 1.587 0.750 0.044 Sagnari 52.05 0.465 0.0375 0.170 2590 1.385 0.691 0.117 Paleocene 55.9 0.49 -0.05 0.165 2590 0.629 0.240 0.051 Basement 65 0 0.157 -0.157 -0.225

Jabalíg Solsona 33.77 0.698 -0.05 0.074 2500 3.580 1.660 0.028 Igualada-Tossa 36.81 1.41 0.09 0.105 2530 3.260 1.580 0.182 Folgueroles 40 0.155 0.125 0.131 2590 2.040 1.000 0.074 Banyoles 41.5 0.58 0.075 0.142 2550 1.860 0.890 0.269 Tavertet 43 0.417 0.02 0.150 2650 1.140 0.480 0.093 Penya 46.26 0.143 0.0125 0.167 2720 0.560 0.180 0.039 Cadí 49 0.297 0.015 0.174 2680 0.350 0.070 0.045 Basement 55.9 0 0.165 -0.170 -0.240

Puig-reigg Solsona 34.36 0.5632 -0.05 0.080 2540 3.100 1.490 0 Igualada-Tossa 36.81 1.6628 0.09 0.105 2400 2.910 1.490 0.231 Folgueroles 40 0.0975 0.125 0.131 2540 1.430 0.750 0.076 Banyoles 41.5 0.218 0.075 0.142 2520 1.260 0.640 0.141 Tavertet 43 0.127 0.02 0.150 2640 0.930 0.430 0.029 Penya 46.26 0.0967 0.0125 0.167 2670 0.760 0.330 0.019 Cadí 49 0.4605 0.015 0.174 2590 0.650 0.280 0.075 Basement 55.9 0 0.165 -0.170 -0.240 Appendix A. Supplement to Chapter 2: Pyrenean case study 105

Stratigraphic Age hickness Water Depth Sea Level ρ Total Subs. Tectonic Subs. Rate Name (Ma)b (km) (km)b,c (km)b,c (kg/m3)d (km)c Subs. (km)c (mm/y)e Santpedorg Solsona 36.15 0.15 -0.05 0.098 2530 1.592 0.823 -0.141 Igualada-Tossa 36.81 0.92 0.09 0.105 2440 1.637 0.916 0.158 Collbas 40 0.16 0.025 0.131 2580 0.796 0.413 0.072 Pontils 41.5 0.36 -0.01 0.142 2520 0.621 0.305 0.033 Orpí 52.7 0.15 0.015 0.169 2610 0.094 -0.065 0.054 Basement 55.9 0 0.165 -0.165 -0.237

Castellfollitg Solsona 34.31 0.574 -0.05 0.079 2520 1.806 0.964 0.021 Igualada- 36.81 0.683 0.09 0.105 2410 1.583 0.912 0.133 Cardona Collbas 40 0.108 0.025 0.131 2520 0.885 0.488 0.059 Pontils 41.5 0.521 -0.01 0.142 2530 0.750 0.399 0.051 Orpí 52.7 0.049 0.015 0.169 2650 -0.070 -0.174 0.056 Basement 53.9 0 0.168 -0.168 -0.241

Montserratg (17.1n) 37.2 0.37 -0.1 0.109 2600 1.688 0.505 0.271 37.47 0.05 -0.08 0.112 2600 1.440 0.432 -0.219 (17.2.3n) 37.6 0.4325 -0.04 0.113 2580 1.448 0.460 0.141 38.11 0.028 0.025 0.117 2550 1.152 0.388 0.271 (1.8n) 38.42 0.3995 -0.04 0.120 2570 1.049 0.304 0.110 40.13 0.068 -0.05 0.132 2570 0.621 0.116 0.033 (1.9n) 41.25 0.12 -0.05 0.140 2580 0.544 0.078 0.210 Mediona-La- 41.52 0.4238 -0.05 0.142 2570 0.410 0.022 0.018 Salut Basement 55.9 0 0.165 -0.165 -0.237 aBorehole locations in Figure 2.9. bAt top of unit. cRelative to present-day sea level. dDensity eTectonic subsidence rate fRough estimate based on stratigraphic section ‘La Nou’ on Spanish 1:50,000 geological map 255 ‘La Pobla de Lillet’ [Vergés et al., 1994]. A compaction coeicient of 0.6 and initial porosity of 55% were used to estimate decompaction. gConverted from Vergés et al. [1998].

next page: Table A.5. Field data 106 Appendix A. Supplement to Chapter 2: Pyrenean case study

Planes type waypoint lat long dip azimuth strike stratigraphy comment Bedding 1.3 43.419968 1.651450 51.3 202.4 112.4 c7ag best measurement hard sandstone surface and relatively flat. Bedding 1.3 43.419968 1.651450 66.9 187.9 97.9 c7ag measured in marls not so reliable Bedding 1.3 43.419968 1.651450 52.3 194.8 104.8 c7ag silt/sandstone in tracks average reliability Bedding 1.3 42.928951 1.917953 29.4 200.7 110.7 c6bm unreliable measured in small stream more sandy level of Bedding 1.6 42.928951 1.917953 64.4 204.3 114.3 c7ag very good quality measurement Bedding 1.6 42.928951 1.917953 56.7 189.5 99.5 c7ag quite good quality Bedding 1.6 42.928951 1.917953 53.3 206.0 116.0 c7ag very good Bedding 1.8 42.940868 1.898085 55.1 6.9 276.9 c6bg Bedding 1.8 42.940868 1.898085 64.9 34.5 304.5 c6bg Bedding 1.8 42.940868 1.898110 56.2 20.2 290.2 c6bg Bedding 1.8 42.940868 1.898110 59.6 21.9 291.9 c6bg Bedding 1.8 42.940392 1.897982 58.4 25.1 295.1 c6bg Bedding 1.8 42.940392 1.897982 60.0 16.6 286.6 c6bg Bedding 1.9 42.920753 1.897063 24.2 195.9 105.9 e2 base of thanetian limestones contact with black marls/shales Bedding 1.10 42.958481 1.862358 52.1 6.6 276.6 e3 base of turitella malrs Bedding 1.11 42.958294 1.860599 35.4 29.5 299.5 e3 ilerdian basal limestones Bedding 2.5 42.947819 1.905628 51.7 6.9 276.9 e2 thanetian limestones Bedding 2.5 42.947826 1.905752 60.6 16.5 286.5 e2 thanetian limestones less reliable Bedding 2.5 42.947998 1.905791 55.8 12.5 282.5 e2 thanetian limestones Bedding 2.5 42.947998 1.905791 59.5 11.9 281.9 e2 thanetian limestones less reliable Bedding 2.5 42.949249 1.904099 21.3 32.0 302.0 e3 ilerdian basal limestones not very reliable but best we can Bedding 2.6 42.947330 1.906368 57.2 21.9 291.9 e2 beautiful surface thanetian limestones Bedding 2.6 42.947330 1.906368 56.7 25.6 295.6 e2 same surface different spot Bedding 2.6 42.947918 1.906671 57.2 17.0 287.0 e2 probable bedding plane slightly steeper than s of here but Bedding 2.7 42.951885 1.996932 28.0 5.0 275.0 e3 Bedding 2.7 42.951885 1.996932 29.3 3.2 273.2 e3 Bedding 2.7 42.951885 1.996932 36.7 4.5 274.5 e3 Bedding 2.7 42.951885 1.996932 24.7 6.5 276.5 e3 Bedding 2.7 42.951885 1.996932 24.3 6.8 276.8 e3 Bedding 2.8 42.986580 2.004328 6.2 34.6 304.6 e3c‐5 Bedding 2.8 42.986580 2.004328 2.6 6.9 276.9 e3c‐5 Bedding 2.8 42.986580 2.004328 21.8 38.3 308.3 e3c‐5 Bedding 2.8 42.986580 2.004328 10.5 7.0 277.0 e3c‐5 Bedding 2.8 42.986572 2.004330 6.8 7.6 277.6 e3c‐5 Bedding 2.8 42.986572 2.004330 12.2 297.0 207.0 e3c‐5 Bedding 2.12 42.909618 2.005900 39.5 197.2 107.2 e2 probably thanetian limestone bed base of small gully dug Bedding 2.14 42.909748 2.006697 28.5 185.8 95.8 e2 best plane visible medium reliabiliy Bedding 2.14 42.909687 2.006691 22.2 168.1 78.1 e2 best plane beautiful Bedding 2.15 42.909534 2.009521 55.0 253.7 163.7 e2 Bedding 3.1 42.889126 1.792310 36.4 201.0 111.0 n7c‐d Bedding 3.2 42.882740 1.809537 77.2 3.4 273.4 n7c‐d no clear bedding but probably subvertical and the strike is Bedding 3.3 42.892540 1.806664 37.4 219.0 129.0 c3‐4 right on border not clear wich unit Bedding 3.3 42.892563 1.806816 46.3 205.9 115.9 c3‐4 same material north side of road Bedding 3.3 42.892563 1.806816 41.2 213.2 123.2 c3‐4 idem Bedding 3.4 42.892467 1.808852 53.1 201.6 111.6 c5a1 high quality measurement overturned Bedding 3.7 42.883301 1.843118 72.6 166.3 76.3 n7c‐d or c1‐3 sandstone bed interbedded in black marls or shales exposed in path. dip estimated strike correct Bedding 3.7 42.883350 1.843082 76.2 169.0 79.0 n7c‐d or c1‐3 another sandstone bed dip estimated strike correct Bedding 3.8 42.883591 1.842013 47.0 166.9 76.9 c1‐3 outcrop of more marly part in path dip not precise strike correct Bedding 3.10 42.909325 1.864726 30.4 201.8 111.8 c6 limestone beds turning in anticline core. probable southern limb Bedding 3.11 42.908852 1.864850 47.2 213.4 123.4 c6 bedding measured behind travertine waterfall (not directly behind water more next to but strike reaches behind waterfall) Bedding 4.1 42.869225 1.894742 62.7 173.5 83.5 n5‐6 n7a dip varies strike correct Bedding 4.1 42.869225 1.894742 75.4 165.5 75.5 n5‐6 n7a dip varies strike correct Bedding 4.1 42.867401 1.897466 82.7 356.8 266.8 n5‐6 n7a Bedding 4.1 42.867401 1.897466 77.1 356.3 266.3 n5‐6 n7a Plane Type 1 4.1 42.867302 1.897581 79.5 67.9 337.9 n5‐6 n7a bedding not clear many planes. plane 1 Plane Type 2 4.1 42.867302 1.897581 42.0 351.5 261.5 n5‐6 n7a bedding not clear many planes plane 2 Plane Type 3 4.1 42.867302 1.897581 31.3 185.5 95.5 n5‐6 n7a bedding not clear many planes plane 3 Plane Type 1 4.1 42.866760 1.899530 70.5 73.6 343.6 n5‐6 n7a Plane Type 2 4.1 42.866760 1.899530 84.4 357.2 267.2 n5‐6 n7a probable bedding similar strike as other measurements and what's visible down in riverbed from distance although hat might also be joint Plane Type 3 4.1 42.866760 1.899530 58.5 284.0 194.0 n5‐6 n7a Bedding 4.4 42.866379 1.900299 87.0 352.4 262.4 n5‐6 n7a Bedding 4.4 42.866379 1.900299 82.5 356.3 266.3 n5‐6 n7a Bedding 4.4 42.866379 1.900299 83.7 353.3 263.3 n5‐6 n7a Bedding 4.4 42.866432 1.900370 85.1 358.7 268.7 n5‐6 n7a Cleavage 4.4 42.866226 1.900417 49.3 150.7 60.7 n5‐6 n7a possible cleavage measured in more strong possibly limey part. no really clear plane Appendix A. Supplement to Chapter 2: Pyrenean case study 107

Planes type waypoint lat long dip azimuth strike stratigraphy comment Bedding 4.5 42.864925 1.903605 49.7 167.0 77.0 n5‐6 n7a Bedding 4.6 42.865078 1.910591 46.9 330.6 240.6 n5‐6 n7a dip uncertain strike correct Bedding 4.6 42.865070 1.910575 49.4 336.9 246.9 n5‐6 n7a dip slightly more certain strike correct Bedding 4.6 42.865051 1.910786 50.3 329.1 239.1 n5‐6 n7a good surface but irregular. Bedding 4.7 42.865253 1.911923 86.4 167.5 77.5 n5‐6 n7a the steepest bed in the outcrop slightly overturned. other beds aren't but are still subvertical Bedding 4.8 42.864990 1.913026 79.6 161.0 71.0 n5‐6 n7a Bedding 4.8 42.864788 1.913069 79.9 155.8 65.8 n5‐6 n7a Bedding 4.8 42.864861 1.913131 76.5 154.3 64.3 n5‐6 n7a Bedding 4.8 42.864861 1.913131 78.1 156.2 66.2 n5‐6 n7a Bedding 4.8 42.864624 1.913384 82.6 164.0 74.0 n5‐6 n7a Bedding 4.9 42.864182 1.926726 70.3 167.5 77.5 n5‐6 n7a Bedding 4.9 42.864182 1.926726 74.7 164.8 74.8 n5‐6 n7a Bedding 4.9 42.864182 1.926726 67.3 161.6 71.6 n5‐6 n7a Bedding 4.9 42.864182 1.926726 74.3 165.7 75.7 n5‐6 n7a Bedding 4.10 42.864925 1.935388 63.6 164.4 74.4 n5‐6 n7a Bedding 4.11 42.814030 1.967165 64.0 356.4 266.4 jD unclear could be this. matches with map Bedding 4.13 42.817894 1.984360 50.8 162.6 72.6 n1‐3 weathered surface Bedding 4.13 42.817894 1.984360 38.6 181.4 91.4 n1‐3 weathered surface Bedding 4.14 42.823914 1.995807 59.1 147.3 57.3 U1 marty most recurring planar feature possible bedding plane. that would mean overturned northern flank. Bedding 4.18 42.886559 1.953603 31.1 275.2 185.2 n4a no good surfaces but brdding is approximately correct Bedding 5.1 43.044670 1.933478 12.5 189.6 99.6 e5b‐1 Bedding 5.1 43.044670 1.933478 20.9 200.7 110.7 e5b‐1 Bedding 5.1 43.044670 1.933478 18.5 193.7 103.7 e5b‐1 Bedding 5.2 43.031769 1.948481 23.4 194.5 104.5 e5b‐1 most reliable bed in a sandstone outcrop with a lot of troughs.

Bedding 5.3 43.059387 1.985749 14.2 28.9 298.9 e5c2 footwall syncline? measured in marly bed Bedding 5.4 43.065891 1.980308 28.6 35.4 305.4 e5c2 best approximation of bedding taken in sandstone. Bedding 6.4 42.843102 1.877044 83.0 309.5 219.5 n4b estimated possible bedding irregular. Bedding 6.5 42.843792 1.877202 58.1 357.7 267.7 n4b Bedding 6.6 42.845997 1.875002 67.5 1.8 271.8 n4b measured next to waterfall Bedding 6.7 42.845997 1.875002 66.8 355.1 265.1 n5‐6 n7a not very reliable Bedding 6.7 42.846085 1.875109 79.7 9.2 279.2 n5‐6 n7a probably slightly wrong strike Bedding 6.7 42.845943 1.874590 70.0 1.2 271.2 n5‐6 n7a reasonably ok Bedding 6.7 42.846455 1.874623 77.2 353.3 263.3 n5‐6 n7a good Bedding 6.8 42.849926 1.875373 82.1 178.5 88.5 n5‐6 n7a Bedding 6.8 42.849926 1.875373 82.6 179.1 89.1 n5‐6 n7a Bedding 6.8 42.849892 1.875331 78.3 180.6 90.6 n5‐6 n7a Bedding 6.8 42.849892 1.875331 76.5 180.7 90.7 n5‐6 n7a Bedding 6.8 42.849892 1.875331 86.1 180.4 90.4 n5‐6 n7a srike best approached Bedding 6.9 42.859928 1.879832 18.2 38.8 308.8 n5‐6 n7a not a very good outcrop but dip is consistent enough everywhere to believe Bedding 6.10 42.860176 1.880414 53.3 358.8 268.8 n5‐6 n7a bad quality Bedding 6.10 42.860153 1.880492 89.0 6.9 276.9 n5‐6 n7a could be cleavage instead Cleavage 6.10 42.860153 1.880492 37.8 19.8 289.8 n5‐6 n7a could be bedding instead not high quality measurement

Bedding 6.11 42.861431 1.882061 63.1 193.1 103.1 n5‐6 n7a strike approximate Bedding 6.11 42.861431 1.882061 75.5 183.9 93.9 n5‐6 n7a may be cleavage Cleavage 6.11 42.861465 1.882128 80.6 4.0 274.0 n5‐6 n7a may be bedding. Cleavage 6.12 42.821552 1.859688 37.0 327.2 237.2 marble breccia?bedding undulates a lot Cleavage 6.12 42.821552 1.859688 27.9 343.0 253.0 marble breccia?bedding undulates a lot Cleavage 6.12 42.821602 1.859777 43.9 310.6 220.6 marble breccia?bedding undulates a lot Bedding 7.1 42.916245 1.982443 31.9 200.1 110.1 c7aG Bedding 7.1 42.916470 1.982203 29.2 211.6 121.6 c7aG Bedding 7.2 42.917164 1.983906 16.1 6.9 276.9 c7aG vey high quality Bedding 7.2 42.917130 1.983954 11.9 354.6 264.6 c7aG Bedding 7.2 42.917168 1.984024 16.8 345.0 255.0 c7aG Bedding 7.2 42.917183 1.984108 16.2 23.2 293.2 c7aG not very reliable Bedding 7.2 42.917091 1.984259 17.4 3.5 273.5 c7aG Bedding 7.2 42.917198 1.984292 18.7 353.1 263.1 c7aG Bedding 7.3 42.920551 1.989191 69.7 32.4 302.4 c7aG Bedding 7.4 42.932060 1.998419 13.7 344.3 254.3 c7b Bedding 7.7 42.938812 1.997894 13.6 12.5 282.5 e2a very irregular surface Bedding 7.7 42.938812 1.997894 41.4 32.2 302.2 e2a better but steeper than general outcrop Bedding 7.7 42.938812 1.997787 13.8 327.6 237.6 e2a Bedding 7.7 42.938812 1.997787 13.9 356.5 266.5 e2a Bedding 9.1 42.899036 1.850392 24.8 174.4 84.4 c5b‐6 bedding very approximate Bedding 9.2 42.893642 1.854856 18.3 167.2 77.2 c5b‐6 might be loose Bedding 9.2 42.893593 1.854873 50.6 158.9 68.9 c5b‐6 very good fixed in streambed gutter Bedding 9.2 42.893444 1.854515 41.8 177.9 87.9 c5b‐6 Bedding 9.2 42.893444 1.854515 43.1 194.5 104.5 c5b‐6 very good plane Bedding 9.2 42.893276 1.854544 35.4 181.7 91.7 c5b‐6 ver good plane 108 Appendix A. Supplement to Chapter 2: Pyrenean case study

Planes type waypoint lat long dip azimuth strike stratigraphy comment Bedding 9.2 42.893276 1.854544 30.4 201.4 111.4 c5b‐6 undulating Bedding 9.3 42.891582 1.851803 46.6 160.7 70.7 c5a2 irregular but ok Bedding 10.1 42.951637 1.922155 21.2 350.6 260.6 e3a1 embedded in road Bedding 10.1 42.951637 1.922155 25.4 332.0 242.0 e3a1 idem best Bedding 10.1 42.951637 1.922155 19.6 323.7 233.7 e3a1 idem Bedding 10.1 42.951763 1.922121 16.2 4.4 274.4 e3a1 idem Bedding 10.1 42.951763 1.922121 17.2 7.1 277.1 e3a1 idem Bedding 10.1 42.951763 1.922121 13.4 17.0 287.0 e3a1 idem Bedding 10.2 42.949974 1.922069 22.0 12.5 282.5 e3a1 approximation Bedding 10.2 42.949974 1.922069 17.5 354.9 264.9 e3a1 approximation Bedding 10.2 42.949974 1.922069 22.8 0.3 270.3 e3a1 approximation in marls Bedding 10.2 42.949974 1.922069 13.7 30.8 300.8 e3a1 approximation Bedding 10.2 42.949974 1.922069 14.5 356.3 266.3 e3a1 approximation Bedding 10.3 42.959412 1.900889 11.4 326.2 236.2 e3a1 approximate Bedding 10.4 42.962368 1.903509 12.0 358.4 268.4 e3b approximated strike correct and dip estimated. dip difficultnto see in outcrop Bedding 10.4 42.962296 1.903571 7.2 317.4 227.4 e3b much better Bedding 10.4 42.962296 1.903571 10.1 300.9 210.9 e3b much better Bedding 10.4 42.962296 1.903571 11.2 345.1 255.1 e3b dip estimated Bedding 10.5 42.962135 1.889214 19.4 8.3 278.3 e3b Bedding 10.5 42.962135 1.889214 18.3 17.8 287.8 e3b Bedding 10.5 42.962135 1.889214 13.8 341.4 251.4 e3b approximated Bedding 10.5 42.962135 1.889214 19.0 22.2 292.2 e3b Bedding 10.5 42.962135 1.889214 15.9 348.0 258.0 e3b Bedding 10.5 42.962063 1.889216 18.4 7.2 277.2 e3b Bedding 10.6 42.971687 1.911127 10.3 251.7 161.7 e3b Bedding 10.6 42.971687 1.911127 3.9 331.5 241.5 e3b not as good as previous Bedding 10.6 42.971561 1.911532 9.8 17.3 287.3 e3b very good underside of bed Bedding 10.6 42.971561 1.911532 5.8 355.1 265.1 e3b Bedding 10.6 42.971561 1.911532 10.9 20.4 290.4 e3b Bedding 10.6 42.971561 1.911532 13.2 29.0 299.0 e3b Plane Type 1 11.1 42.961346 1.775601 62.7 35.3 305.3 c7b‐e1 recurring structural plane in danian limestone outcrop Plane Type 1 11.1 42.961346 1.775601 56.2 55.2 325.2 c7b‐e1 idem planes very planar on outcrop scale but not on fieldbook‐scale Plane Type 1 11.1 42.961346 1.775601 73.8 270.7 180.7 c7b‐e1 possible conjugate? sense: normal Plane Type 1 11.1 42.961369 1.775589 56.6 75.3 345.3 c7b‐e1 Plane Type 1 11.1 42.961372 1.775583 85.0 296.2 206.2 c7b‐e1 conjugate? Bedding 11.1 42.961681 1.776178 14.4 339.8 249.8 c7b‐e1 best bedding Bedding 11.1 42.962032 1.776134 21.4 14.6 284.6 c7b‐e1 also relatively good Bedding 11.3 42.964359 1.773709 19.7 24.9 294.9 e2a good surface Bedding 11.3 42.964359 1.773709 16.5 7.5 277.5 e2a idem Bedding 11.3 42.964359 1.773709 17.7 6.4 276.4 e2a idem Bedding 11.3 42.964359 1.773709 14.6 14.4 284.4 e2a not as good Bedding 11.4 42.965450 1.772800 21.6 349.4 259.4 e2a fairly good Bedding 11.4 42.965450 1.772800 21.2 3.6 273.6 e2a medium Bedding 11.4 42.965450 1.772800 27.6 12.4 282.4 e2a idem Bedding 11.4 42.965450 1.772800 21.9 21.5 291.5 e2a better Bedding 11.5 42.966923 1.773761 19.4 350.1 260.1 e2a very good underside of bed but small surface making it difficult to keep ipad in same orientation Bedding 11.5 42.966923 1.773761 20.5 350.7 260.7 e2a very good topside of thin bed Bedding 11.5 42.966923 1.773761 14.5 326.8 236.8 e2a relatively good underside approx 10 m n along road from previous measurement Bedding 11.5 42.967197 1.773628 20.1 19.1 289.1 e2a inside very massive bed missing piece of small layer. approx 10 m north along road from previous Bedding 11.5 42.967197 1.773628 31.9 346.3 256.3 e2a topside of bed bedding contacts much more chaotic here. approx 20 m n along road from previous Bedding 11.5 42.967197 1.773628 12.1 357.8 267.8 e2a underside of bed slightly chaotic. same location as previous

Bedding 11.5 42.967422 1.773374 21.0 15.6 285.6 e2a relatively good underside approx 10 m along road from previous Bedding 11.5 42.967506 1.773338 18.4 67.9 337.9 e2a good underside of bed with thin loose layers. approx 14 m north along road Bedding 11.5 42.967506 1.773269 16.6 4.4 274.4 e2a another underside very good quality approx 5 m north along road from previous. Bedding 11.5 42.967506 1.773269 26.7 357.3 267.3 e2a underside slightly undulating. approx 10 m north from previous. limestone more evenly bedded (30‐50 cm) and darker beige in colour. Bedding 11.5 42.967506 1.773269 24.4 355.3 265.3 e2a same location underside not so good Bedding 11.5 42.967506 1.773269 21.2 358.8 268.8 e2a underside same location again slightly better than previous

Bedding 11.5 42.967655 1.773161 19.6 351.4 261.4 e2a underside approx 5 m north from previous. wet. Appendix A. Supplement to Chapter 2: Pyrenean case study 109

Planes type waypoint lat long dip azimuth strike stratigraphy comment Bedding 11.5 42.967762 1.773041 17.7 306.2 216.2 e2a top of bed not in cliff but more open part between middle and top of outcrop stratigraphy. dip direction could be off due to erosion. approx 20 m north from previous

Bedding 11.5 42.967937 1.772846 21.2 2.9 272.9 e2a underside of very thin bed really good measurement on really good plane. first measurement in top of outcrop (stratigraphically speaking). approx 20 m north along road from previous Bedding 11.5 42.967937 1.772846 14.7 4.3 274.3 e2a underside dip direction may be off. same location as previous

Bedding 11.5 42.967937 1.772846 15.6 6.9 276.9 e2a underside seems like a very good plane. same location as previous Bedding 11.5 42.967937 1.772846 9.1 321.6 231.6 e2a idem Bedding 11.5 42.967934 1.772731 19.4 353.3 263.3 e2a underside seems very good plane. approx5 m north from previous Bedding 11.7 42.968327 1.772397 19.9 11.1 281.1 e2a underside very similar to top of outcrop 11.5. relatively good surface but undulates. Bedding 12.1 42.907330 2.022579 16.0 214.9 124.9 e3 good underside of bed Bedding 12.6 42.883930 2.038083 27.3 159.6 69.6 l1‐4 very good underside of smooth bed Bedding 12.6 42.883930 2.038083 32.2 168.3 78.3 l1‐4 more chaotic Bedding 12.7 42.883385 2.038876 20.9 205.8 115.8 l1‐4 good underside of bed Bedding 12.8 42.882923 2.039047 17.7 174.0 84.0 n1‐3 very nice Bedding 12.8 42.882923 2.039047 13.6 196.4 106.4 n1‐3 idem Bedding 12.8 42.882923 2.039047 15.3 201.5 111.5 n1‐3 idem Bedding 12.8 42.882923 2.039047 17.2 168.1 78.1 n1‐3 idem Bedding 12.12 42.872719 2.023239 14.7 357.9 267.9 n1‐3 could also be third direction in joint set Bedding 12.12 42.872719 2.023239 16.1 350.3 260.3 n1‐3 idem Joint 12.12 42.872719 2.023239 63.7 106.6 16.6 n1‐3 number 1 of conjugate Joint 12.12 42.872719 2.023239 56.7 236.9 146.9 n1‐3 number 2 option 1 Joint 12.12 42.872719 2.023239 58.0 293.4 203.4 n1‐3 number 2 alternative Bedding 12.13 42.851608 2.103537 81.0 8.6 278.6 n1‐3 Bedding 12.13 42.851608 2.103537 86.7 1.7 271.7 n1‐3 Bedding 12.13 42.851608 2.103537 87.3 1.1 271.1 n1‐3 Bedding 12.13 42.851608 2.103537 86.2 0.6 270.6 n1‐3 Bedding 12.13 42.851608 2.103537 87.7 4.0 274.0 n1‐3 Bedding 12.13 42.851608 2.103537 83.8 4.6 274.6 n1‐3 Bedding 12.13 42.851608 2.103537 89.0 4.3 274.3 n1‐3 Bedding 12.13 42.851665 2.103532 84.2 9.0 279.0 n1‐3 Bedding 12.13 42.851665 2.103532 89.0 5.8 275.8 n1‐3 Bedding 12.13 42.851665 2.103532 89.0 5.9 275.9 n1‐3 Cleavage 12.13 42.851749 2.103414 75.8 178.3 88.3 n1‐3 approximation of possible cleavage Bedding 12.13 42.851742 2.103505 85.3 4.9 274.9 n1‐3 Bedding 12.13 42.851742 2.103505 81.7 2.2 272.2 n1‐3 Bedding 12.13 42.851742 2.103505 88.2 184.5 94.5 n1‐3 Bedding 12.13 42.851742 2.103505 86.7 15.6 285.6 n1‐3 Bedding 12.13 42.851822 2.103425 88.8 8.1 278.1 n1‐3 Bedding 12.13 42.851822 2.103425 88.7 187.2 97.2 n1‐3 Bedding 12.13 42.851868 2.103452 89.0 2.0 272.0 n1‐3 Bedding 12.13 42.851894 2.103465 84.0 5.7 275.7 n1‐3 Bedding 12.13 42.851925 2.103464 86.1 3.9 273.9 n1‐3 Bedding 12.13 42.852184 2.103313 88.3 197.0 107.0 n1‐3 Bedding 12.13 42.852325 2.103278 88.1 8.1 278.1 n1‐3 Bedding 12.13 42.852486 2.103317 89.0 190.2 100.2 n1‐3 Bedding 12.13 42.852516 2.103295 88.1 358.8 268.8 n1‐3 Bedding 12.13 42.852592 2.103310 86.0 186.1 96.1 n1‐3 Bedding 12.13 42.852646 2.103310 86.1 178.5 88.5 n1‐3 Bedding 12.13 42.852753 2.103316 89.0 183.9 93.9 n1‐3 Bedding 12.13 42.852753 2.103316 89.0 176.1 86.1 n1‐3 found next to cleavage measurement Cleavage 12.13 42.852753 2.103316 63.6 186.6 96.6 n1‐3 definitely cleavage found in more marly part Bedding 12.13 42.852901 2.103330 89.0 191.0 101.0 n1‐3 Bedding 12.13 42.852901 2.103330 87.6 184.9 94.9 n1‐3 Bedding 12.13 42.853016 2.103364 86.4 189.8 99.8 n1‐3 Bedding 12.14 42.853630 2.103269 49.3 86.0 356.0 n1‐3 Bedding 12.14 42.853630 2.103182 46.8 92.2 2.2 n1‐3 Bedding 12.14 42.853661 2.103207 35.9 80.9 350.9 n1‐3 Bedding 12.14 42.853664 2.103227 48.9 83.2 353.2 n1‐3 Plane Type 1 13.1 42.807365 1.922065 43.6 169.2 79.2 granite mineral banding in granite outcrop. don't know if true outcrop or separate block Plane Type 1 13.6 42.819290 1.922106 85.9 153.9 63.9 jD plane of stromatolites. approximation but very reliable because stromatolites visible on intersecting planes giving 3d orientation Bedding 13.6 42.819290 1.922106 78.9 166.0 76.0 jD presumed bedding parallel with stromatoliyes. measured on tiny part of bedding plane 110 Appendix A. Supplement to Chapter 2: Pyrenean case study

Planes type waypoint lat long dip azimuth strike stratigraphy comment Bedding 13.7 42.823303 1.922760 43.1 185.6 95.6 n1‐3 erosional surface uneven Bedding 13.7 42.823303 1.922760 46.0 192.1 102.1 n1‐3 dip correct strike approximated Bedding 15.1 42.882248 1.799874 61.3 324.8 234.8 h2 Bedding 15.1 42.882256 1.799878 53.5 341.0 251.0 h2 Bedding 15.1 42.882198 1.799864 47.5 347.3 257.3 h2 Bedding 15.1 42.882198 1.799864 45.1 338.0 248.0 h2 Bedding 15.1 42.882206 1.799856 64.2 343.8 253.8 h2 Bedding 15.1 42.882206 1.799880 57.3 346.4 256.4 h2 Bedding 15.1 42.882217 1.799908 60.4 2.3 272.3 h2 Bedding 15.1 42.882191 1.799853 61.6 11.0 281.0 h2 Bedding 15.1 42.882141 1.799814 48.3 348.4 258.4 h2 Bedding 15.1 42.882164 1.799761 49.3 15.8 285.8 h2 Bedding 15.2 42.881630 1.799273 39.3 303.3 213.3 h2 Bedding 15.2 42.881645 1.799297 36.5 295.9 205.9 h2 Bedding 15.2 42.881641 1.799286 33.3 286.6 196.6 h2 Bedding 15.2 42.881641 1.799301 37.4 298.3 208.3 h2 Bedding 15.2 42.881649 1.799328 35.7 295.4 205.4 h2 Bedding 15.2 42.881634 1.799340 34.9 300.7 210.7 h2 Bedding 15.2 42.881634 1.799345 31.4 313.3 223.3 h2 Bedding 15.2 42.881630 1.799339 32.1 307.2 217.2 h2 Bedding 15.2 42.881630 1.799339 29.7 308.9 218.9 h2 measured on next bed Bedding 15.2 42.881676 1.799336 36.2 302.9 212.9 h2 measured on next bed Bedding 15.2 42.881634 1.799334 50.1 316.6 226.6 h2 curving bed 1 5 m south of good bed Bedding 15.2 42.881622 1.799303 30.3 294.9 204.9 h2 idem other side of curved bed Bedding 15.4 42.879990 1.800484 28.4 150.5 60.5 h2 Plane Type 1 15.5 42.879326 1.800185 51.1 264.2 174.2 h2 joint maybe? two directions of lines are visible on this plane forming a diamond pattern. those are probably bedding and cleavage making this third plane a joint. old measurement first logged with wrong location made new measurement with correct location and changed location of this to correct

Joint 15.5 42.879326 1.800292 50.8 264.3 174.3 h2 probably joint. two line directions visible on this plane forming a diamond crosshatch pattern. those are probably bedding and cleavage making this a joint. Bedding 15.6 42.877178 1.798476 84.3 203.0 113.0 d4‐7 Bedding 15.6 42.877159 1.798466 86.5 203.2 113.2 d4‐7 Bedding 15.6 42.877155 1.798415 81.8 202.0 112.0 d4‐7 Bedding 15.6 42.877106 1.798371 81.3 198.0 108.0 d4‐7 Bedding 15.6 42.877075 1.798359 85.7 209.1 119.1 d4‐7 Bedding 15.6 42.877022 1.798311 88.0 210.7 120.7 d4‐7 Bedding 15.6 42.876991 1.798189 87.1 213.4 123.4 d4‐7 Bedding 15.6 42.876961 1.798190 83.0 221.2 131.2 d4‐7 Bedding 15.6 42.876858 1.798092 70.2 191.4 101.4 d4‐7 Bedding 15.6 42.876873 1.798081 72.7 191.2 101.2 d4‐7 Bedding 15.6 42.876850 1.798038 89.0 188.7 98.7 d4‐7 Bedding 15.8 42.871540 1.770496 60.2 139.2 49.2 d1‐3 undulating surface Bedding 15.8 42.871552 1.770517 59.8 130.8 40.8 d1‐3 good Bedding 15.8 42.871517 1.770474 67.3 135.7 45.7 d1‐3 good Bedding 15.8 42.871418 1.770467 59.0 160.0 70.0 d1‐3 ok Bedding 15.8 42.871418 1.770467 71.9 147.4 57.4 d1‐3 very good plane in similar direction but bedding? i don't know for sure. it is recurring though Bedding 15.11 42.849194 1.753420 16.7 341.7 251.7 k‐o on steepest part of folded shale Cleavage 15.11 42.849213 1.753395 46.2 359.9 269.9 k‐o nice cleavage recurring throughout always same orientation

Bedding 15.11 42.849220 1.753257 0.0 0.0 0.0 k‐o in horizontal part of shales Bedding 16.1 42.855999 2.101587 83.9 358.1 268.1 n4b Bedding 16.1 42.855995 2.101596 79.2 2.1 272.1 n4b Bedding 16.1 42.855988 2.101544 78.8 357.5 267.5 n4b Bedding 16.1 42.856018 2.101575 86.1 4.5 274.5 n4b Bedding 16.1 42.856083 2.101620 84.1 8.3 278.3 n4b Bedding 16.1 42.856094 2.101661 83.0 1.3 271.3 n4b Bedding 16.1 42.856106 2.101647 87.3 2.4 272.4 n4b Joint 16.1 42.856083 2.101754 75.9 275.0 185.0 n4b recurring fracture orientation in one bed Bedding 16.1 42.856098 2.101633 80.6 355.8 265.8 n4b Joint 16.1 42.856121 2.101707 70.6 274.5 184.5 n4b Bedding 16.1 42.856251 2.101837 89.0 350.4 260.4 n4b very irregular bedding plane has striations in several directions. Bedding 16.1 42.856289 2.101866 75.3 354.5 264.5 n4b very irregular Joint 16.1 42.856289 2.101868 45.3 304.4 214.4 n4b recurring joint Joint 16.1 42.856308 2.101943 67.5 300.3 210.3 n4b Joint 16.1 42.856342 2.101966 79.1 276.3 186.3 n4b Bedding 16.1 42.856316 2.101960 72.1 6.0 276.0 n4b very irregular Joint 16.1 42.856342 2.101923 73.2 283.7 193.7 n4b irregular Appendix A. Supplement to Chapter 2: Pyrenean case study 111

Planes type waypoint lat long dip azimuth strike stratigraphy comment Joint 16.1 42.856354 2.101910 56.4 269.2 179.2 n4b Bedding 16.1 42.856316 2.101911 79.4 349.6 259.6 n4b irregular Joint 16.1 42.856297 2.101939 54.4 85.0 355.0 n4b very minor Joint 16.2 42.856419 2.102146 37.6 232.9 142.9 n4b Joint 16.2 42.856441 2.102111 38.5 243.0 153.0 n4b Bedding 16.2 42.856438 2.102199 36.1 46.8 316.8 n4b possible bedding. could also be just another joint. Joint 16.2 42.856567 2.102309 40.6 93.5 3.5 n4b i have no clue what this plane us supposed to be so i've marked it as just a joint Joint 16.2 42.856541 2.102453 56.0 104.5 14.5 n4b idem Joint 16.2 42.856560 2.102402 54.9 121.0 31.0 n4b idem Joint 16.5 42.856941 2.103208 72.2 0.1 270.1 n4b very irregular but still very flat overall Joint 16.5 42.856918 2.103245 48.0 133.1 43.1 n4b supersmooth and flat Joint 16.5 42.856915 2.103261 17.5 212.1 122.1 n4b idem Joint 16.5 42.856911 2.103250 17.6 215.8 125.8 n4b idem Joint 16.5 42.856922 2.103292 45.2 231.6 141.6 n4b smooth and flat Joint 16.5 42.856937 2.103326 59.5 81.2 351.2 n4b idem Joint 16.5 42.856918 2.103369 59.3 228.8 138.8 n4b one of 2 intersecting joints Joint 16.5 42.856941 2.103363 61.1 82.8 352.8 n4b two of 2 intersecting joints both supersmooth and flat Bedding 16.6 42.857906 2.103939 69.9 197.3 107.3 n5 measured in more shaly part Bedding 16.6 42.857948 2.103960 82.7 205.9 115.9 n5 Cleavage 16.6 42.857937 2.103965 42.9 177.6 87.6 n5 Bedding 16.6 42.858028 2.103951 88.4 208.1 118.1 n5 Bedding 16.6 42.858101 2.103951 89.0 21.1 291.1 n5 Bedding 16.6 42.858128 2.104034 86.4 207.4 117.4 n5 Bedding 16.7 42.858727 2.104521 53.3 38.5 308.5 n5 hard to measure here Joint 16.8 42.859131 2.105999 76.4 119.7 29.7 n5 Joint 16.8 42.859138 2.106031 79.5 120.0 30.0 n5 Joint 16.8 42.859135 2.106027 72.7 114.6 24.6 n5 Joint 16.8 42.859135 2.106027 75.1 118.3 28.3 n5 Bedding 16.9 42.859509 2.106434 88.9 18.7 288.7 n5 poor surface for measurement Bedding 16.9 42.859524 2.106473 87.1 20.6 290.6 n5 good Bedding 16.9 42.859539 2.106514 76.7 25.1 295.1 n5 good surface slightly less steep than others Bedding 16.10 42.859566 2.107135 85.8 31.5 301.5 n5 bedding or fault plane. has slickensides but in many directions

Bedding 16.13 42.864090 2.114105 80.9 198.6 108.6 n6a2 surface best approaches true bedding but impossible to measure die to cleavage and way of weathering. Bedding 16.13 42.864109 2.114095 89.0 13.4 283.4 n6a2 idem Cleavage 16.13 42.864124 2.114113 74.5 197.3 107.3 n6a2 best point to measure cleavage Bedding 16.14 42.868607 2.154400 35.3 152.3 62.3 n7a Bedding 16.14 42.868607 2.154400 29.1 151.2 61.2 n7a Cleavage 16.14 42.868622 2.154436 85.1 192.8 102.8 n7a Bedding 17.1 42.823887 1.995786 53.9 156.3 66.3 U1 marty Bedding 17.1 42.823879 1.995787 55.6 145.3 55.3 U1 marty Bedding 17.1 42.823895 1.995762 59.1 149.2 59.2 U1 marty Bedding 17.1 42.823887 1.995776 56.5 152.7 62.7 U1 marty Bedding 17.1 42.823864 1.995811 59.2 162.4 72.4 U1 marty plane with striations very vague 2 directions Joint 17.1 42.823971 1.995772 23.8 220.7 130.7 U1 marty joint with lineation Bedding 17.3 42.816418 1.986550 52.5 19.6 289.6 m6 marty Bedding 17.4 42.816738 1.986023 28.0 196.0 106.0 m6 marty mary's measurement Bedding 17.5 42.816849 1.984895 32.0 235.0 145.0 m6 marty mary Bedding 17.6 42.816654 1.987630 89.0 138.1 48.1 m6 marty probable bedding in cleavage heavy outcrop Cleavage 17.6 42.816547 1.987679 52.2 189.0 99.0 m6 marty cleavage in cleavage heavy outcrop Cleavage 17.6 42.816624 1.987724 67.3 188.8 98.8 m6 marty cleavage in cleavage heavy outcrop cleavage is refracted this is more parallel to bedding Bedding 17.9 42.811920 1.988350 35.7 27.3 297.3 meso meta marty Bedding 17.9 42.811928 1.988315 34.8 18.0 288.0 meso meta marty Bedding 17.9 42.811935 1.988313 34.2 10.1 280.1 meso meta marty Bedding 17.9 42.811935 1.988313 39.4 7.0 277.0 meso meta marty Bedding 17.9 42.811943 1.988311 32.4 358.3 268.3 meso meta marshallow turning into fault Joint 17.10 42.813526 1.986720 68.5 20.4 290.4 meso meta marcommon foliation in marble Bedding 18.1 42.834232 1.869058 51.3 351.9 261.9 n1‐3 Bedding 18.4 42.834450 1.870809 28.0 356.0 266.0 n1‐3 folded bed steepens to north southern measurement taken approx 1m apart Bedding 18.4 42.834450 1.870809 42.0 0.0 270.0 n1‐3 folded beds steepening north northern measurement taken approx 1m apart Bedding 18.4 42.834656 1.870842 42.0 350.0 260.0 n1‐3 approx 10 m northeast of previous two Bedding 18.5 42.843117 1.877119 43.9 19.6 289.6 n4b Cleavage 18.6 42.846199 1.875110 83.8 12.8 282.8 n5‐6 n7a oriented sample Bedding 18.6 42.846268 1.875306 81.6 330.9 240.9 n5‐6 n7a Cleavage 18.8 42.855446 1.877487 19.7 192.1 102.1 n5‐6 n7a Bedding 18.8 42.855671 1.877934 49.2 184.2 94.2 n5‐6 n7a oriented sample to look for cleavage frau 7 Bedding 18.9 42.855675 1.877794 50.0 196.0 106.0 n5‐6 n7a marys measurement same place as mine accidentally registered under 18.8 112 Appendix A. Supplement to Chapter 2: Pyrenean case study

Planes type waypoint lat long dip azimuth strike stratigraphy comment Bedding 18.10 42.856155 1.877836 35.9 213.2 123.2 n5‐6 n7a Bedding 18.10 42.856159 1.877890 33.9 183.0 93.0 n5‐6 n7a approx 3d Bedding 18.11 42.864189 1.886069 29.6 145.3 55.3 n5‐6 n7a could be loose block Bedding 18.12 42.865780 1.888044 71.0 358.7 268.7 n5‐6 n7a Cleavage 18.12 42.865791 1.888031 13.3 213.7 123.7 n5‐6 n7a Bedding 18.13 42.865807 1.888121 66.5 351.4 261.4 n5‐6 n7a Cleavage 18.13 42.867519 1.889999 35.9 152.8 62.8 n5‐6 n7a Bedding 18.14 42.892696 1.927176 68.7 321.0 231.0 n5 Plane Type 1 18.17 42.892105 1.932804 80.0 161.3 71.3 n5‐6 n7a Plane Type 1 18.17 42.892071 1.932886 78.2 167.0 77.0 n5‐6 n7a Plane Type 1 18.17 42.891884 1.933066 63.7 170.6 80.6 n5‐6 n7a Plane Type 2 18.17 42.891884 1.933066 60.0 180.0 90.0 n5‐6 n7a Cleavage 18.17 42.891884 1.933067 85.4 352.4 262.4 n5‐6 n7a Bedding 18.18 42.891178 1.934045 29.5 357.7 267.7 n5‐6 n7a question mark Bedding 18.18 42.891216 1.934088 25.8 12.9 282.9 n5‐6 n7a sedimentological features visible on eroded surface confirm this plane is bedding Plane Type 1 19.1 42.766727 1.850349 89.0 345.5 255.5 meso meta marmetamorphic foliation in marble Bedding 20.9 42.784637 1.975004 63.2 196.8 106.8 d4‐7 overturned devonian close to hinge Bedding 20.10 42.785149 1.976014 11.6 140.6 50.6 d4‐7 top flank 1 Bedding 20.10 42.785160 1.976074 8.0 176.2 86.2 d4‐7 top flank 2 Bedding 20.10 42.785145 1.976202 79.3 188.7 98.7 d4‐7 bottom flank 1 Cleavage 20.10 42.785114 1.976191 45.7 215.8 125.8 d4‐7 axial plane cleqvage Bedding 20.13 42.776936 2.140635 62.6 32.4 302.4 turonian Bedding 20.13 42.777752 2.139368 40.5 355.7 265.7 turonian Bedding 20.13 42.777370 2.138630 47.5 3.0 273.0 turonian Bedding 20.13 42.778290 2.138091 56.5 343.4 253.4 turonian Plane Type 1 20.17 42.806110 2.077482 54.7 33.1 303.1 orthogneiss foliation Bedding NPZ2015‐2 42.816669 1.987665 82.2 167.6 77.6 alb marty Cleavage NPZ2015‐2 42.816666 1.987659 68.6 205.2 115.2 alb marty Bedding NPZ2015‐25 42.819550 1.900269 18.6 59.9 329.9 Meso meta marty Bedding NPZ2015‐25 42.819508 1.900245 58.2 335.9 245.9 Meso meta marty Bedding NPZ2015‐25 42.819485 1.900390 22.2 72.5 342.5 Meso meta marty Bedding NPZ2015‐25 42.819321 1.900504 69.3 351.3 261.3 Meso meta marty Bedding NPZ2015‐25 42.819328 1.900469 43.7 352.8 262.8 Meso meta marty Bedding NPZ2015‐25 42.819324 1.900471 47.9 347.2 257.2 Meso meta marty Bedding NPZ2015‐25 42.819355 1.900455 42.7 22.1 292.1 Meso meta marty Bedding NPZ2015‐25 42.819359 1.900443 37.0 29.4 299.4 Meso meta marty Bedding NPZ2015‐25 42.819355 1.900433 39.6 11.7 281.7 Meso meta marty Bedding NPZ2015‐25 42.819355 1.900452 50.2 6.5 276.5 Meso meta marty Bedding NPZ2015‐25 42.819347 1.900459 37.5 15.1 285.1 Meso meta marty Bedding NPZ2015‐25 42.819355 1.900442 37.6 22.6 292.6 Meso meta marty Bedding NPZ2015‐25 42.819458 1.900400 48.4 10.1 280.1 Meso meta marty Bedding NPZ2015‐25 42.819473 1.900360 48.3 8.3 278.3 Meso meta marty Bedding NPZ2015‐26 42.817749 1.865870 59.6 187.7 97.7 Meso meta marty Bedding NPZ2015‐47 42.801258 1.912681 57.0 25.7 295.7 Meso meta marty Bedding NPZ2015‐49 42.801380 1.908615 81.7 314.7 224.7 Meso meta marty Bedding NPZ2015‐49 42.801388 1.908599 89.0 135.8 45.8 Meso meta marty Bedding NPZ2015‐50 42.801662 1.907800 53.8 324.6 234.6 Meso meta marty Bedding NPZ2015‐50 42.801670 1.907813 46.7 316.9 226.9 Meso meta marty Bedding NPZ2015‐50 42.801662 1.907847 65.5 342.5 252.5 Meso meta marty Bedding NPZ2015‐50 42.801666 1.907924 57.9 333.1 243.1 Meso meta marty Bedding NPZ2015‐50 42.801662 1.907949 50.3 331.5 241.5 Meso meta marty Bedding NPZ2015‐50 42.801670 1.907967 50.2 326.9 236.9 Meso meta marty Bedding NPZ2015‐50 42.801662 1.907977 37.1 336.0 246.0 Meso meta marty Bedding NPZ2015‐50 42.801662 1.907973 25.6 311.1 221.1 Meso meta marty Bedding NPZ2015‐51 42.802006 1.907022 76.4 335.9 245.9 Meso meta marty Bedding NPZ2015‐51 42.801643 1.907050 66.2 324.1 234.1 Meso meta marty Bedding NPZ2015‐51 42.801674 1.907046 53.5 328.7 238.7 Meso meta marty Bedding NPZ2015‐53 42.802723 1.906337 54.5 356.6 266.6 Meso meta marty Bedding NPZ2015‐53 42.802696 1.906368 54.0 335.7 245.7 Meso meta marty Bedding NPZ2015‐56 42.805229 1.900069 59.1 150.0 60.0 Meso meta marty Bedding NPZ2015‐56 42.805225 1.900125 54.0 154.9 64.9 Meso meta marty Bedding NPZ2015‐56 42.805229 1.900196 69.9 174.7 84.7 Meso meta marty Bedding NPZ2015‐56 42.805233 1.900232 73.2 161.0 71.0 Meso meta marty Bedding NPZ2015‐56 42.805222 1.900224 66.9 178.8 88.8 Meso meta marty Bedding NPZ2015‐56 42.805244 1.900225 66.9 169.9 79.9 Meso meta marty Bedding NPZ2015‐58 42.807655 1.896838 69.5 305.1 215.1 Meso meta marty Joint NPZ2015‐58 42.807652 1.896958 46.9 7.0 277.0 Meso meta marty Bedding NPZ2015‐58 42.807686 1.896940 65.9 269.0 179.0 Meso meta marty Bedding NPZ2015‐58 42.807686 1.896940 70.8 270.9 180.9 Meso meta marty Bedding NPZ2015‐58 42.807686 1.896940 74.4 281.3 191.3 Meso meta marty Bedding NPZ2015‐58 42.807583 1.897089 38.5 241.6 151.6 Meso meta marty Bedding NPZ2015‐58 42.807552 1.897065 56.1 254.1 164.1 Meso meta marty Appendix A. Supplement to Chapter 2: Pyrenean case study 113

Planes type waypoint lat long dip azimuth strike stratigraphy comment Bedding NPZ2015‐58 42.807507 1.897061 33.8 248.4 158.4 Meso meta marty Bedding NPZ2015‐58 42.807491 1.897081 45.1 223.5 133.5 Meso meta marty Bedding NPZ2015‐58 42.807484 1.897088 42.7 202.6 112.6 Meso meta marty Bedding NPZ2015‐58 42.807491 1.897121 47.8 177.7 87.7 Meso meta marty Bedding NPZ2015‐58 42.807465 1.897106 52.8 170.1 80.1 Meso meta marty Bedding NPZ2015‐58 42.807461 1.897120 44.1 187.7 97.7 Meso meta marty Bedding NPZ2015‐58 42.807457 1.897129 42.8 202.1 112.1 Meso meta marty Bedding NPZ2015‐58 42.807457 1.897129 48.6 235.6 145.6 Meso meta marty Bedding NPZ2015‐58 42.807453 1.897130 39.1 230.5 140.5 Meso meta marty Bedding NPZ2015‐58 42.807442 1.897112 50.4 228.4 138.4 Meso meta marty Bedding NPZ2015‐59 42.808178 1.896383 48.1 84.7 354.7 Meso meta mar? Plane Type 1 NPZ2015‐59 42.808167 1.896356 83.3 124.0 34.0 Meso meta marlight dark banding in white marble Bedding NPZ2015‐61 42.799168 2.035429 43.2 8.3 278.3 Meso meta marty Bedding NPZ2015‐62 42.811462 1.890566 50.9 286.3 196.3 Meso meta marty Bedding NPZ2015‐62 42.811531 1.890529 46.6 335.7 245.7 Meso meta marty Bedding NPZ2015‐62 42.811577 1.890523 55.8 300.9 210.9 Meso meta marty Bedding NPZ2015‐62 42.811596 1.890534 17.1 107.9 17.9 Meso meta marty Bedding NPZ2015‐62 42.811527 1.890475 58.6 315.9 225.9 Meso meta marty Bedding NPZ2015‐62 42.811554 1.890491 29.8 174.7 84.7 Meso meta marty Bedding NPZ2015‐62 42.811531 1.890472 15.7 171.5 81.5 Meso meta marty Bedding NPZ2015‐64 42.812172 1.886002 49.6 57.7 327.7 Meso meta marty Bedding NPZ2015‐64 42.812206 1.886002 48.4 65.7 335.7 Meso meta marty Bedding NPZ2015‐64 42.812241 1.885979 42.8 62.9 332.9 Meso meta marty Bedding NPZ2015‐64 42.812256 1.885955 40.3 45.2 315.2 Meso meta marty Bedding NPZ2015‐64 42.812256 1.885955 30.0 65.0 335.0 Meso meta marty Bedding NPZ2015‐64 42.812260 1.885939 39.2 67.6 337.6 Meso meta marty Bedding NPZ2015‐64 42.812260 1.885961 42.1 57.9 327.9 Meso meta marty Bedding NPZ2015‐64 42.812260 1.885909 30.6 73.8 343.8 Meso meta marty Bedding NPZ2015‐64 42.812256 1.885869 36.0 41.2 311.2 Meso meta marty Bedding NPZ2015‐64 42.812241 1.885848 46.8 53.4 323.4 Meso meta marty Bedding NPZ2015‐65 42.812496 1.884062 80.3 248.6 158.6 Meso meta marty Bedding NPZ2015‐65 42.812489 1.884078 89.0 235.6 145.6 Meso meta marty Bedding NPZ2015‐67 42.812634 1.882487 19.7 357.4 267.4 Meso meta marty Bedding NPZ2015‐67 42.812588 1.882504 48.6 95.6 5.6 Meso meta marty Bedding NPZ2015‐68 42.812508 1.881925 70.8 303.2 213.2 Meso meta marty Bedding NPZ2015‐68 42.812485 1.882324 87.8 301.4 211.4 Meso meta marty Bedding NPZ2015‐72 42.814522 1.875264 49.8 15.6 285.6 Meso meta marty Bedding NPZ2015‐72 42.814564 1.875293 42.6 37.9 307.9 Meso meta marty Bedding NPZ2015‐72 42.814495 1.875392 55.8 19.0 289.0 Meso meta marty Bedding NPZ2015‐72 42.814510 1.875398 29.0 79.5 349.5 Meso meta marty Bedding NPZ2015‐72 42.814507 1.875490 28.3 55.6 325.6 Meso meta marty Bedding NPZ2015‐73 42.814289 1.874563 61.9 241.6 151.6 Meso meta marty Bedding NPZ2015‐74 42.815281 1.872690 50.1 66.1 336.1 Meso meta marty Bedding NPZ2015‐74 42.815266 1.872699 47.4 63.1 333.1 Meso meta marty Bedding NPZ2015‐74 42.815174 1.872813 55.8 130.3 40.3 Meso meta marty Bedding NPZ2015‐74 42.815140 1.872781 42.5 103.9 13.9 Meso meta marty Bedding NPZ2015‐75 42.817909 1.862480 64.0 312.6 222.6 Meso meta marty Bedding NPZ2015‐77 42.820393 1.859922 57.5 40.0 310.0 Meso meta marty Bedding NPZ2015‐77 42.820393 1.859922 46.5 46.6 316.6 Meso meta marty Bedding NPZ2015‐79 42.795708 2.067406 22.3 199.6 109.6 black‐grey dolomite Bedding NPZ2015‐79 42.795696 2.067387 30.2 176.0 86.0 black‐grey dolomite Bedding NPZ2015‐80 42.795254 2.066427 27.8 213.3 123.3 black‐grey dolomite Bedding NPZ2015‐80 42.795250 2.066455 30.6 202.4 112.4 black‐grey dolomite Bedding NPZ2015‐81 42.794796 2.065963 68.3 208.1 118.1 black‐grey dolomite Bedding NPZ2015‐82 42.794315 2.065840 37.0 176.4 86.4 black‐grey dolomite Bedding NPZ2015‐83 42.793926 2.065784 73.5 209.7 119.7 black‐grey dolomite Bedding NPZ2015‐84 42.793056 2.065214 66.8 22.1 292.1 Meso meta marty Bedding NPZ2015‐84 42.793049 2.065260 71.6 333.9 243.9 Meso meta marty Plane Type 2 NPZ2015‐86 42.792049 2.063689 73.7 230.1 140.1 schists foliatipn of schists Plane Type 2 NPZ2015‐87 42.791954 2.062254 72.7 22.6 292.6 schists foliation in schist Plane Type 2 NPZ2015‐88 42.787937 2.061447 57.5 62.8 332.8 schists foliation in strongly foliated granitic rock Plane Type 2 NPZ2015‐89 42.784092 2.060651 44.9 145.2 55.2 schists estimate Bedding NPZ2015‐107 42.850761 1.836826 82.1 45.2 315.2 beige limestone Bedding NPZ2015‐107 42.850758 1.836795 65.7 67.8 337.8 beige limestone Bedding NPZ2015‐107 42.850819 1.837069 77.6 280.8 190.8 beige limestone Bedding NPZ2015‐107 42.850830 1.837054 81.3 337.2 247.2 beige limestone Bedding NPZ2015‐107 42.850853 1.837039 82.8 346.9 256.9 beige limestone Bedding NPZ2015‐109 42.850983 1.836652 74.5 87.5 357.5 beige limestone Plane Type 1 NPZ2015‐110 42.852531 1.837412 77.8 296.0 206.0 Meso meta marfoliation trend in outcrop Bedding NPZ2015‐112 42.853241 1.834209 76.3 51.6 321.6 zim brown red crystalline limestone Bedding NPZ2015‐112 42.853252 1.834055 67.8 55.5 325.5 zim brown red crystalline limestone Bedding NPZ2015‐112 42.853313 1.834045 79.5 71.4 341.4 zim brown red crystalline limestone Bedding NPZ2015‐112 42.853390 1.834089 84.5 51.8 321.8 zim brown red crystalline limestone 114 Appendix A. Supplement to Chapter 2: Pyrenean case study

Planes type waypoint lat long dip azimuth strike stratigraphy comment Bedding NPZ2015‐112 42.853176 1.834325 68.8 43.9 313.9 zim brown red crystalline limestone Bedding NPZ2015‐112 42.853657 1.833858 65.4 75.9 345.9 zim brown red crystalline limestone Bedding NPZ2015‐116 42.851521 1.834615 76.1 182.6 92.6 zim brown red crystalline limestone Bedding NPZ2015‐116 42.851627 1.834349 63.8 211.1 121.1 zim brown red crystalline limestone Bedding NPZ2015‐116 42.851643 1.834376 60.4 215.3 125.3 zim brown red crystalline limestone Bedding NPZ2015‐117 42.852005 1.835385 84.4 357.1 267.1 zim brown red crystalline limestone Bedding NPZ2015‐117 42.852001 1.835246 78.8 228.3 138.3 zim brown red crystalline limestone Bedding NPZ2015‐117 42.851921 1.835178 40.4 75.1 345.1 zim brown red crystalline limestone Bedding NPZ2015‐117 42.851906 1.835194 54.8 81.4 351.4 zim brown red crystalline limestone Bedding NPZ2015‐117 42.851856 1.835246 57.4 67.7 337.7 zim brown red crystalline limestone Bedding NPZ2015‐117 42.851852 1.835101 25.8 54.1 324.1 zim brown red crystalline limestone Appendix A. Supplement to Chapter 2: Pyrenean case study 115

Faults type waypoint lat long dip azimuth strike stratigraphy comment Sinistral Strike S1.6 42.928951 1.917953 54.9 282.7 192.7 c7aG slickenside 1 Sinistral Strike S1.6 42.928951 1.917953 71.2 283.3 193.3 c7aG slickenside 2 Unknown 1.8 42.940353 1.897991 65.0 108.0 18.0 c6bG slickenside 3 Unknown 2.6 42.947918 1.906671 73.4 21.5 291.5 e2 plane with striations with slickensides still visible under coating of mud. sense: Dextral Strike S 4.1 42.867420 1.897573 64.3 245.7 155.7 n5‐6 n7a dextral

Dextral Strike S 4.4 42.866379 1.900385 68.0 54.0 324.0 n5‐6 n7a very small and irregular but with slickensides so measured it extremely flat recurring multiple times on other side stream Normal Fault 6.4 42.843102 1.877044 43.7 17.9 287.9 n4b in western cliff. has striations. Reverse Fault 11.4 42.965450 1.772800 32.4 300.3 210.3 e2a with slickensides reverse Unknown 12.7 42.883137 2.038856 36.1 308.5 218.5 l1‐4 no signs of movement direction probable bedding but moves around a lot through outcrop Unknown 12.9 42.881886 2.038496 71.2 77.3 347.3 n1‐3 and is hard to recognise Unknown 12.9 42.881886 2.038496 70.7 83.8 353.8 n1‐3 idem Unknown 12.9 42.881886 2.038496 66.6 84.2 354.2 n1‐3 idem Unknown 12.9 42.881886 2.038496 77.3 76.0 346.0 n1‐3 idem Unknown 12.9 42.881886 2.038496 69.1 79.1 349.1 n1‐3 idem Unknown 16.1 42.856220 2.101801 55.4 306.2 216.2 n4b not completely flat has striations 1 good surface but small so could only fit 2 corners of ipad on Unknown 16.1 42.856270 2.101821 41.6 239.0 149.0 n4b fault. has striations 2 Dextral Strike S 16.1 42.856293 2.101766 48.2 208.2 118.2 n4b bigger fault plane. with striations 4 Dextral Strike S 16.2 42.856438 2.102143 47.0 285.0 195.0 n4b has striations 5 dextral Dextral Strike S 16.2 42.856457 2.102132 38.9 245.0 155.0 n4b has striations 6 dextral fault along which the bedding and cleavage have been Normal Fault 16.13 42.864033 2.114037 22.0 25.8 295.8 n6a2 deformed in bands Reverse Fault 17.8 42.812855 1.979467 50.8 228.8 138.8 M1‐4 marty fault contact with metamorphic. bright white marble. Reverse Fault 17.8 42.812824 1.979453 34.0 200.0 110.0 M1‐4 marty Mary's measurement Reverse Fault 17.8 42.812874 1.979467 36.9 221.2 131.2 M1‐4 marty measurement three ipad smaller fault and actually more inclined than fault but a Reverse Fault 17.8 42.812977 1.979425 55.6 112.4 22.4 M1‐4 marty repeating surface less inclined than second fault so dip azimuth different but it Reverse Fault 17.8 42.812901 1.979399 33.7 55.2 325.2 M1‐4 marty is a big plane best approximation of actual fault because no really Reverse Fault 17.8 42.812851 1.979373 30.4 91.9 1.9 M1‐4 marty representative planes could be found

Normal Fault 17.9 42.811928 1.988314 26.8 224.6 134.6 meso meta marsmall fault cuts through beds reducing dip pf beds below

another fault that steepens dip of strata above and below it Normal Fault 17.9 42.811848 1.988303 46.2 229.3 139.3 meso meta marbut has slickensides indicating normal movement attempt 2 of fault plane with slickensides actual plane with Normal Fault 17.9 42.811852 1.988248 30.9 223.9 133.9 meso meta marslickensides this time. (rough plane) Sinistral Strike S17.10 42.813530 1.986720 72.9 257.5 167.5 meso meta marwith slickensides sinistral Normal Fault 17.11 42.813549 1.987239 85.5 190.6 100.6 meso meta marhas slickensides direction ambiguous maybe normal another fault plane same block also with slickensides. Dextral Strike S 17.11 42.813572 1.987222 63.7 138.7 48.7 meso meta mardirection ambiguous maybe dextral Unknown 17.12 42.813488 1.986387 45.6 342.3 252.3 meso meta marbig flat plane in quarry Unknown 17.12 42.813473 1.986372 43.5 346.1 256.1 meso meta marsame plane different place for accuracy Unknown 17.12 42.813480 1.986382 45.7 338.1 248.1 meso meta maridem Reverse Fault 18.15 42.894493 1.929424 60.3 176.9 86.9 n5 fault 1 with striae 1 reverse Unknown 18.15 42.894436 1.929473 35.3 128.2 38.2 n5 fault 2 with striae 2 both reverse and normal? Sinistral Strike S18.15 42.894394 1.929448 14.4 24.4 294.4 n5 fault 3a Sinistral Strike S18.15 42.894432 1.929408 29.2 70.3 340.3 n5 fault 3b with striae 3 sinistral Unknown 20.14 42.776520 2.141117 50.1 353.7 263.7 black dolomite breccia Normal Fault 20.14 42.776524 2.141097 38.3 22.7 292.7 black dolomite normal Unknown NPZ2015‐112 42.853455 1.834091 43.0 297.9 207.9 zim brown red crystalline limestone Unknown NPZ2015‐112 42.853687 1.833946 80.9 119.7 29.7 zim brown red crystalline limestone Unknown NPZ2015‐112 42.853752 1.834033 67.1 115.9 25.9 zim brown red crystalline limestone Unknown NPZ2015‐118 42.852467 1.836233 89.0 61.4 331.4 zim brown red crystalline limestone Unknown NPZ2015‐118 42.852406 1.836199 78.2 55.8 325.8 zim brown red crystalline limestone Unknown NPZ2015‐118 42.852379 1.836162 61.4 80.2 350.2 zim brown red crystalline limestone 116 Appendix A. Supplement to Chapter 2: Pyrenean case study

Lineations type waypoint lat long plunge azimuth stratigraphy comment Slickenside 1.6 42.928951 1.917953 2.0 7.1 c7ag slickenside 1 sense unclear maybe sinistral Slickenside 1.6 42.928951 1.917953 30.4 201.9 c7ag slickenside 2 sinistral normal Slickenside 1.8 42.940353 1.897991 46.0 50.3 c6bg slickenside 3 Slickenside 2.6 42.947918 1.906671 73.8 54.6 e2 striations in thanetian limestones Slickenside 3.2 42.882740 1.809537 68.2 127.4 n7c‐d northern block up contrary to regional sense with slickensides still visible under coating of mud. sense: Slickenside 4.1 42.867420 1.897573 21.4 317.5 n5‐6 n7a dextral Fold Axis 4.4 42.866379 1.900385 1.7 265.2 n5‐6 n7a possible cleavage bedding intersection Slickenside 4.4 42.866379 1.900385 23.2 332.8 n5‐6 n7a dextral measured on fault plane sense: looks like normal but feels the same as reverse. unclear but guessing normal. striations Slickenside 6.4 42.843102 1.877044 41.9 6.9 n4b have been partly dissolved. no growth of mineral just striation in sandstone. sense: Slickenside 7.1 42.916245 1.982443 28.7 263.9 c7aG dextral Slickenside 11.4 42.965450 1.772800 10.8 237.3 e2a top to ene Line Type 1 15.5 42.879311 1.800301 2.0 175.5 h2 intersection of probable bedding with joint Line Type 2 15.5 42.879337 1.800309 21.8 187.7 h2 intersection of probable cleavage with joint. Fold Axis 15.6 42.877052 1.798257 47.4 121.3 d4‐7 on bedding plane of limestone bedding‐cleavage intersection not easily measured best Fold Axis 15.11 42.849213 1.753293 7.9 76.8 k‐o location Slickenside 16.1 42.856251 2.101811 26.2 19.5 n4b striations 1. direction impossible to tell here. Slickenside 16.1 42.856205 2.101872 3.5 329.8 n4b striations 2. direction unknown most common direction of striations on irregular bedding Slickenside 16.1 42.856277 2.101780 76.7 45.1 n4b plane. direction unknown Slickenside 16.1 42.856285 2.101823 1.1 125.3 n4b striations 4 dextral. (top to wnw) Slickenside 16.2 42.856430 2.102158 2.2 206.1 n4b striations 5 Slickenside 16.2 42.856392 2.102188 2.7 332.6 n4b striation 6 dextral i think Slickenside 17.1 42.823875 1.995799 51.5 181.0 U1 marty on bedding plane 5 might indicate movement Slickenside 17.1 42.823853 1.995754 17.3 68.7 U1 marty direction 2 of lineation on same plane lineation on joint measured on slightly loose hanging piece. Slickenside 17.1 42.823978 1.995820 17.2 184.2 U1 marty actual plunge a few degrees steeper bedding cleavage intersection in cleavage heavy outcrop if Fold Axis 17.6 42.816528 1.987627 60.6 249.7 m6 marty bedding is corrrect Slickenside 17.9 42.811829 1.988332 29.4 169.9 meso meta marslickensides indicate normal movement Slickenside 17.9 42.811825 1.988193 30.4 181.1 meso meta marattempt 2 of slickensides same place Slickenside 17.10 42.813526 1.986720 31.7 337.9 meso meta maron fault sinistral

Slickenside 17.11 42.813526 1.987253 56.5 110.8 meso meta maron previous fault plane direction ambiguous maybe normal on second fault plane on same block direction ambiguous Slickenside 17.11 42.813576 1.987176 36.5 63.6 meso meta marmaybe dextral Slickenside 18.5 42.843109 1.876790 41.5 358.5 n4b normal Fold Axis 18.6 42.846214 1.875133 75.4 304.7 n5‐6 n7a bedding cleavage intersection Slickenside 18.15 42.894432 1.929458 48.3 224.3 n5 striae 1 reverse Slickenside 18.15 42.894382 1.929474 30.1 175.1 n5 striae 2 reverse and normal? Slickenside 18.15 42.894470 1.929421 12.4 18.1 n5 striae 3 sinistral Fold Axis 19.16 42.813480 1.852060 3.7 340.4 meso meta marty Fold Axis 20.10 42.785137 1.976164 22.6 142.6 d4‐7 different partnofnsame hinge Fold Axis 20.10 42.785091 1.976126 2.8 108.7 d4‐7 approximation of hinge newr other measurements Lineation 20.14 42.776527 2.141088 45.9 21.3 black dolomite breccia Lineation 20.14 42.776524 2.141097 30.7 353.0 black dolomite normal Lineation 21.6 42.767666 2.299687 73.6 119.3 black marble lineation in marble Appendix B

Supplement to Chapter 3: Numerical modelling of orogen crustal structure 118 Appendix B. Supplement to Chapter 3: Numerical modelling of orogen crustal structure

Supplementary methods Mathematical description FANTOM is an arbitrary Lagrangian-Eulerian inite-element code that solves the Stokes equation, assuming incompressibility and negligible inertial forces [hieulot, 2011]:

(B.1)

is the stress tensor, the spatial dimensions, is density and the gravitational acceleration. In the viscous regime, materials follow a power-law rheology, meaning strain rate depends on stress to a power .

(B.2)

is the strain rate tensor, is a pre-exponential material constant, is the deviatoric stress tensor, the exponent, is the activation energy, is pressure, is the activation volume, is the gas constant and is temperature. Material constants , , , and are determined from laboratory data [Karato and Wu, 1993; Gleason and Tullis, 1995]. Frictional (brittle) behaviour is approximated by choosing an efective viscosity that places the state of stress at yield.

(B.3)

is the second invariant of the strain rate tensor, deined as . Yield stress is determined using a Drucker-Prager yield criterion, which is equivalent to the Coulomb failure criterion in 2D. (B.4) is the second invariant of the deviatoric stress, cohesion, and the angle of internal friction. Strain sotening of frictional materials is simulated by reducing and linearly with increasing strain in the range 0.5 < < 1.5, where and is the second invariant of the deviatoric strain rate tensor. he initial value of = 15° for the upper crust has been chosen to represent the strength of the crust with hydrostatic pore luid pressure. he fully weakened value = 2° has been experimentally determined to best approximate behaviour of natural examples [Lavier et al., 2000; Huismans and Beaumont, 2007]. he temperature is determined by the heat transport equation

(B.5)

is density, the heat capacity, is time, the spatial dimensions, the thermal conductivity, and is radiogenic heat production. Density variations due to temperature are accounted for: (B.6) is the reference density, the thermal expansion coeicient and the temperature at which the reference density was measured. Model set up he model domain is 1200 km wide and 600 km deep (Figure B.1). We use a vertically variable resolution that is distributed as follows: In the top 25 km, cells are 500x200 m. he 100 km below that has a resolution of 500x800 m, and the bottom 475 km has a resolution of 500x9800 m. All cells are stretched vertically to accommodate topography. he model comprises a laterally homogeneous, 35 km thick crust, composed of 3 km of pre-orogenic sediment, 1 km of a weak décollement layer, 21 km of upper crust and 10 km of lower crust. Lithospheric mantle lies below the crust to a depth of 120 km, and below that lies sublithospheric Appendix B. Supplement to Chapter 3: Numerical modelling of orogen crustal structure 119

a 6 x 6 km weak seed temperature (°C) 3 km centered around 28 km depth 0 550 1330 0.5 cm/y 0 1 km Crust 35 Lithospheric mantle 21 km 120 Sublithospheric mantle

10 km Depth (km)

600 0 600 1200 Horizontal scale (km) Materials b Strain softening c Strength 0 Predeformation sediment 15° Wet Quartz, ƒ=1

Décollement layer ) ε

( Upper crust

Upper crust eff

φ Wet Lower crust Lower crust Quartz, ƒ=100 2° Depth (km) Lithospheric mantle 0.5ε 1.5 Sublithospheric mantle Dry Olivine, ƒ=1 20 Lithospheric mantle

4 Cohesion (MPa) 120 0.5ε 1.5 0Stress (MPa) 400

Figure B.1. Model set up. (a) Initial geometry of the models, including velocity boundary conditions and initial thermal proile. (b) Implementation of strain sotening in frictional materials. (c) Initial strength proile, showing the strong coupling of the lower crust and lithospheric mantle. Ater Erdős et al. [2014]. mantle. At the center of the model, a weak seed of 6x6 km, centered around 28 km depth has been fully strain weakened. he side boundaries are thermally insulated, while the bottom and top boundaries are kept at 1330° C and 0° C, respectively. he initial temperature at the Moho is 550° C. he top boundary of the model is a true free surface, while the bottom allows free slip. On the let and right boundaries, we introduce material at 5 cm/y in the upper 132 km (10 cm/y in total). Below 172 km, material is extracted at ~0.15 cm/y to balance the volume of introduced material. Between 132 and 172 km, the velocity is interpolated linearly to create a smooth transition. During the extensional phase, where applicable, the velocity boundary conditions are inverted. Model parameters he material properties used in the models are listed in Table B.1. Shortening distribution measurement method he distribution of shortening between the pro- and retro-wedge is quantiied by tracking the horizontal position of a column of points that originates at the centre of the model. his column will be displaced towards the side that accommodates the most deformation. he amount of displacement is a measure for the ratio of shortening accommodated by both sides. his method does not distinguish between the pro- and retro-wedge as originally deined by Willett et al. [1993], where the boundary between the pro- and retro-wedge is a topographic drainage divide. Instead we track the tectonic boundary between pro- and retro-verging structures, which starts aligned with the plate boundary, or point S in Willett et al. [1993]. 120 Appendix B. Supplement to Chapter 3: Numerical modelling of orogen crustal structure -6 -5 19 - - Salt viscous 0.8 ∙ 10 3.1 ∙ 10 3 -28 se a separate scaling se a separate -5 -6 stic . Shale 0.8 ∙ 10 3.1 ∙ 10 base 8.574 ∙ 10 222.815 ∙ 10 A -n 3 -28 -6 -5 0 0 - Sediment 8.574 ∙ 10 222.815 ∙ 10 3 -14 - -6 b sphere 15 ∙ 10 Astheno 429.83 ∙ 10 1.393 ∙ 10 3 -15 -6 Plastic rheologyPlastic Viscous rheology 0 0 0.8 ∙ 10 0 0 3.1 ∙ 10 mantle dry olivine olivine wet ∙ 1 qtz wet ∙ 1 qtz wet ∙ 1 qtz wet 540.41 ∙ 10 2.4168 ∙ 10 Density and thermal parameters thermal and Density a 3 -36 -6 -5 0.8 ∙ 10 3.1 ∙ 10 8.574 ∙ 10 222.815 ∙ 10 3 -28 -6 -5 Upper crustUpper crust Lower Lithospheric visco-plastic visco-plastic visco-plastic visco-plastic visco-plastic visco-pla ) 0 0 25 ∙ 10 ) 803.57 803.57 681.82 681.82 803.57 803.57 803.57 ) 222.815 ∙ 10 ) 0.8 ∙ 10 -1 ) 8.574 ∙ 10 -1 ) 2.25 2.25 2.25 2.25 2.25 2.25 2.25 ) 2800 2800 3360 3300 2300 2300 2300 -1 -3 -1 -1 -3 Mechanical and thermal material properties thermal material and Mechanical ) 3.1 ∙ 10 K ∙s -1 -1 -n mol 3 n 4 4 3.5 3.0 4 4 4 ϕ 15° → 2° 15° → 2° 15° → 2° 15° → 2° 15° → 2° 2° → 1° - (°C) 273.15 273.15 273.15 273.15 273.15 273.15 273.15 0 Type α (K T η (Pa∙s) power-law power-law power-law power-law power-law power-law 10 c (MPa) 20 → 4 20 → 4 20 → 4 20 → 4 20 → 4 2 - Material (J kg Rheology ∙ 1 qtz wet ∙ 100 qtz wet k (W m ρ (kg m A (Pa p H (µW m Q (kJ mol c V (cm Arrows point to fully strain weakened values. fully weakened to strain point Arrows For calculating the efective viscosity in the viscous regime for scaled rheologies, FANTOM does not u does not scaled FANTOM for rheologies, regime in the viscosity viscous calculating the efective For Table B.1. Table a b factor f, but instead incorporates it into the pre-exponential factor A. he conversion is A=f is conversion A. he factor the pre-exponential into it incorporates instead but f, factor Appendix B. Supplement to Chapter 3: Numerical modelling of orogen crustal structure 121 his ensures that all shortening accommodated by the upper and lower plates will be correctly accounted for, even though our tracked boundary does not necessarily represent the suture zone. Our method allows easy comparison to quantiied shortening in natural systems, where shortening on either side of a tectonic boundary can be more reliably reconstructed back in time than the location of the drainage divide. To further facilitate comparison of our results to natural systems, we eliminated the efect of early distributed deformation by only counting shortening accommodated since deformation began to localise in the irst crustal shear zone. When shear zones are already present at the onset of convergence (Models 1 and DR1), shortening was counted from the onset of convergence.

Table B.2. Models Model Extensional Décollement Sedimentation Erosion rateb (s-1) inheritance (km) rheologyc base levela (km) Model 1 50 salt 0 4.35 ∙ 10-15 Model DR1 50 shale 0 4.35 ∙ 10-15 Model DR2 0 shale 0 4.35 ∙ 10-15 aSedimentation is started at 10 My. bErosion is altitude-dependant, set so that a 4 km high topography erodes by 1 km in 2 My. cSalt has efective viscosity = 1019 Pa∙s, shale has efective angle of internal friction = 2°

Supplementary model descriptions We compared our results to two models that have been previously published [Erdős et al., 2014, 2015]. he main characteristics of all three models are listed in Table B.2. For completeness, we describe the full evolution of the two supplementary models below. DR1: rit inheritance + shale + surface processes he pre-orogenic rit phase of model DR1 is practically identical to that of model 1 described in the main paper (phase 0). A narrow, symmetric rit is bounded by two conjugate frictional-plastic shear zones that root in the weak middle crust. Two similar shear zones in the lithospheric mantle accommodate extension in the mantle. he only diference is a lack of gravitational sliding of the sedimentary cover in model DR1. Phase 1 is the symmetric inversion of the rit. Both inherited shear zones are preferentially reactivated simultaneously during the irst 25 km of convergence (Figure B.2a). A keystone block is uplited, and the cover does not slide into the foredeeps, but remains on top. Phase 2 begins ater 25 km of convergence, when a single new shear zone is created through the crust (Figure B.2b). his new shear zone creates an asymmetry that allows subduction of the lower crust and lithospheric mantle, while the other, inherited shear zones are temporarily abandoned, as shown in the strain rate plot (Figure B.2b). At 10 My (50 km of convergence), sedimentation is started, illing the pro- foredeep. he single pro-wedge thrust sheet steepens as it overthrusts this new basin, reactivating the retro-wedge very slowly ater 55 km of convergence. A new thrust sheet is accreted to the pro-wedge ater 80 km of convergence. his uplits the older, upper thrust sheet and increases shortening rates in the retro-wedge (Figure B.2c). he pro-foreland basin is uplited on top of the thick-skinned thrust sheet and a new thin-skinned thrust is formed in the footwall (Figure B.2c). More convergence results in an outward growing thick-skinned pro-wedge, while the sedimentary cover is mostly uplited and eroded along with the thick-skinned thrust sheets (Figure B.2d). Ater 225 km of convergence, the orogen comprises a wide thick-skinned pro-wedge and well-developed retro-wedge. hin-skinned thrusts always link back to the nearest thick-skinned shear zone. 122 Appendix B. Supplement to Chapter 3: Numerical modelling of orogen crustal structure

Model DR1: inheritance + shale + surface processes

a t = 7 My, ∆xc = 23 km pre-orogenic sediments 0 350° C shale upper crust strain rate 10-13 lower crust 40 550° C km lithospheric mantle s-1

80 0

b t = 9 My, ∆xc = 38 km 0 350° C

40

km 550° C

80

c t = 15 My, ∆xc = 95 km syn-orogenic sediments 0 350° C

40

km 550° C

80

d t = 28 My, ∆xc = 225 km 0 350° C

40 km 550° C

80 200 250 300 350 400 450 500 550 600 km

Figure B.2. Evolution of model DR1, with extensional inheritance and a shale décollement level. Δxc is total convergence. Plot windows were chosen so the upper plate appears stationary. he white line in the keystone block marks the column of points tracked for measuring the shortening distribution. Points on this line that are above topography (red) are ignored. Strain rate plots (let) show second invariant of the strain rate tensor. Grey dashed lines in strain rate plots indicate fully strain weakened areas. (a) Symmetric inversion. (b) Onset of collision and asymmetry, retro-wedge abandoned. (c) Retro-wedge reactivated and narrow thin-skinned pro-foreland fold-and-thrust belt. (d) Final coniguration. Well-developed retro- wedge, wide thick-skinned pro-wedge directly overlain by thin-skinned pro-foreland fold-and-thrust belt. DR2: no inheritance + shale + surface processes Initial convergence is accommodated by distributed deformation until a single shear zone forms ater 6 My. his shear zone reaches down into the lithospheric mantle, creating an asymmetry that determines the plate polarity (Figure B.3a). No strain-weakened retro-shear zone exists at this point. In the footwall of this thrust, a small thin-skinned wedge develops (Figure B.3b). 33 km of convergence is accommodated along the single shear zone before a thrust sheet is accreted into the pro-wedge. Both shear zones remain active, although the frontal shear zone is dominant. Meanwhile, the pro-foredeep is illed with sediment and the sedimentary cover of the uplited upper plate begins to erode. As new thrust sheets are stacked in the pro- wedge, the upper, older thrust sheets migrate very slowly onto the upper plate along a new retro-shear zone (Figure B.3c). his process continues until 215 km of convergence, where the oldest and uppermost thrust sheets have been deeply eroded (Figure B.3d). he resulting orogen is highly asymmetrical, with a wide pro-wedge and only a single retro-wedge shear zone that did not accommodate much shortening. Appendix B. Supplement to Chapter 3: Numerical modelling of orogen crustal structure 123

Model DR2: no inheritance + shale + surface processes

a t = 7 My, ∆xc = 13 km pre-orogenic sediments 0 350° C shale upper crust lower crust 40 550° C km lithospheric mantle strain rate 0 s-1 10-13 80

b t = 9 My, ∆xc = 28 km 0 350° C

40

km 550° C

80

c t = 15 My, ∆xc = 85 km syn-orogenic sediments 0 350° C

40

km 550° C

80

d t = 28 My, ∆xc = 215 km 0 350° C

40 km 550° C

80 200 250 300 350 400 450 500 550 600 km

Figure B.3. Evolution of model DR2, with no extensional inheritance and a shale décollement level. Δxc is total convergence. Plot windows were chosen so the upper plate appears stationary. he white line in the keystone block marks the column of points tracked for measuring the shortening distribution. Points on this line that are above topography (red) are ignored. Strain rate plots (let) show second invariant of the strain rate tensor. Grey dashed lines in strain rate plots indicate fully strain weakened areas. (a) initial asymmetry, no conjugate shear zone. (b) Continued deformation only in pro-wedge. (c) Stack of thick- skinned pro-wedge thrust sheets builds and begins to activate a retro-shear zone. (d) Final coniguration, highly asymmetrical orogen with very little deformation of the single retro-shear zone, a wide thick- skinned pro-wedge overlain by a thin-skinned pro-foreland fold-and-thrust belt. References Erdős, Z., Huismans, R.S., and van der Beek, P., 2015, First-order control of syntectonic sedimentation on crustal- scale structure of mountain belts: Journal of Geophysical Research: Solid Earth, v. 120, p. 1–16, doi: 10.1002/2014JB011785. Erdős, Z., Huismans, R.S., van der Beek, P., and hieulot, C., 2014, Extensional inheritance and surface processes as controlling factors of mountain belt structure: Journal of Geophysical Research: Solid Earth, v. 119, p. 9042–9061, doi: 10.1002/2014JB011408. Gleason, G.C., and Tullis, J., 1995, A low law for dislocation creep of quartz aggregates determined with the molten salt cell: Tectonophysics, doi: 10.1016/0040-1951(95)00011-B. Huismans, R.S., and Beaumont, C., 2007, Roles of lithospheric strain sotening and heterogeneity in determining the geometry of rits and continental margins: Geological Society, London, Special Publications, doi: 10.1144/SP282.6. Karato, S. -i., and Wu, P., 1993, Rheology of the Upper Mantle: A Synthesis: Science, v. 260, p. 771–778, doi: 10.1126/science.260.5109.771. 124 Appendix B. Supplement to Chapter 3: Numerical modelling of orogen crustal structure

Lavier, L.L., Buck, W.R., and Poliakov, A.N.B., 2000, Factors controlling normal fault ofset in an ideal brittle layer: Journal of Geophysical Research, v. 105, p. 23431, doi: 10.1029/2000JB900108. hieulot, C., 2011, FANTOM: Two- and three-dimensional numerical modelling of creeping lows for the solution of geological problems: Physics of the Earth and Planetary Interiors, v. 188, p. 47–68, doi: 10.1016/j. pepi.2011.06.011. Willett, S.D., Beaumont, C., and Fullsack, P., 1993, Mechanical model for the tectonics of doubly vergent compressional orogens: Geology, v. 21, p. 371–374, doi: 10.1130/0091-7613(1993)021<0371:MMFTTO> 2.3.CO. Appendix C

Supplement to Chapter 4: Numerical modelling of strain partitioning 126 Appendix C. Supplement to Chapter 4: Numerical modelling of strain partitioning

Supplementary models hree supplementary models are described here. hese models are similar to M6, M7 and M8 described in Chapter 4, but feature an even weaker salt décollement. he efective viscosity of these supplementary models is 1018 Pas. Here we only describe the evolution during the irst 125 km of convergence, as opposed to 300 km. he salt décollement in these models is so weak that deformation reaches the edge of the model domain, which invalidates the model result ater this point.

Table C.1. Supplementary models Décollement Model Extension Rheologya Distribution Sedimentationb Erosionc S1 Yes Weak salt Full width No No S2 Yes Weak salt Full width Yes No S3 Yes Weak salt Full width Yes Yes a 18 Weak salt ηef = 10 Pa∙s. bbase level = 0 m, start at 10 My. c -15 -1 Ef = 4.35∙10 s .

a S1: extension + weak salt pre-orogenic t = 19 My; x = 125 km ∆ E shale sediments 0 350° C upper crust lower crust 40 km 550° C lithospheric mantle

80 b S2: extension + weak salt + sedimentation

t = 19 My; ∆xE = 125 km syn-orogenic sediments 0 350° C

40

km 550° C

80 c S3: extension + weak salt + sedimentation + erosion

t = 19 My; ∆xE = 125 km 0 350° C

40

km 550° C

80 0 100 200 300 400 500 600 km

Figure C.1. Model results with a weaker salt décollement layer, without and with surface processes (S1,

S2 and S3). Convergence (ΔxC) was measured from restored crustal thickness onwards. he white line indicates the tracked points that measure the distribution of shortening. Points above topography are ignored and coloured red. Appendix C. Supplement to Chapter 4: Numerical modelling of strain partitioning 127

Model S1 he overall development of model S1 is very similar to model M6. During the irst 5 My, the weak salt décollement allows the sedimentary cover to slide into the rit. he rit is symmetrically inverted along both inherited shear zones, past full inversion at 6.5 My. As soon as the central block is uplited, the cover slides down towards both sides. At 15 km beyond full inversion, a new shear zone is created that establishes an asymmetry and the retro-wedge is abandoned. A large thin-skinned pro-foreland fold-and- thrust belt forms as the sedimentary cover of the lower plate is stacked against the orogen. A new pro- wedge thrust sheet is accreted around 77.5 km ater full inversion. his uplits the keystone block and older pro-wedge thrust sheet. he rigid top of the keystone block is separated from its ductile root as two large blocks slide downward onto the upper plate. he retro-foreland fold-and-thrust belt is transported over the weak salt décollement in front of these basement blocks. Meanwhile, in the pro-foreland, a wide thin- skinned fold-and-thrust belt has formed, and crustal thrust sheets in the pro-wedge almost completely overlap each other. Ater 125 km of convergence, the crustal structure is extremely similar to model M6. A narrow thick-skinned orogen lanked by a wide thin-skinned pro-foreland fold-and-thrust belt and slightly narrower retro-foreland fold-and-thrust belt (Figure C.1a). he thin-skinned foreland fold-and- thrust belts are a bit wider. Model S2 Model S2 includes sedimentation, similar to model M7. he evolution prior to the onset of synorogenic sedimentation is the same as S1. Ater 10 My, sedimentation starts and the thin-skinned deformation front in the pro-foreland rapidly migrated outward. his creates wide, relatively stable zones of wedge- top basins separated by narrow zones of intense deformation. Ater 75 km of convergence, a new pro- wedge thrust sheet is accreted. his uplits the keystone block and older pro-wedge thrust sheet while the sedimentary cover of the new pro-wedge thrust sheet slides of the thrust sheet and into the foredeep under the inluence of gravity. Ater 125 km of convergence, the geometry remains similar to model M7, except the pro-foreland fold-and-thrust belt is wider (Figure C.1b). Model S3 Erosion is included in model S3. he irst 10 My are almost identical to S1 and S2: Symmetric inversion of the rit, full inversion at 6.5 My. he cover slides of the keystone block and the asymmetry is established at 15 km of shortening beyond full inversion. Similar to model s2, the pro-foreland establishes wide wedge- top basins separated by narrow zones of intense deformation. he cover of each newly accreted crustal thrust sheet slides of into the foreldeep. Erosion of the central orogen causes deformation to focus there. Ater 125 km of convergence, the crustal structure of the central orogen is more chaotic than models M8 and S2 (Figure C.1c).

Appendix D

Raman data 130 Appendix D. Raman data

Raman spectra of carbonaceous material Raman spectra of carbonaceous material (RSCM) analysis is a paleothermometry technique for rocks that have experienced temperatures from ~200° C to ~600° C [Beyssac et al., 2002; Lahid et al., 2010]. It is used to measure the peak temperature a given sample has experienced and gives no age. Raman spectra of carbonaceous material have several distinct peaks at determined Raman shits. he precise location and amplitude of these peaks is a function of the temperature experienced by the sample. Because preparation of the sample into a thin section involves lots of abrasive processes, the temperature experienced by the surface of the thin-section is an unreliable thermometer. herefore, it is necessary to ind carbonaceous material that is not at the surface of the sample, but underneath a transparent mineral (usually calcite). Because the analysis and data processing involve highly subjective steps, a large number of measurements per sample is needed to give a statistically meaningful result. 10 spectra per sample is a good minimum value [Lahid et al., 2010].

Unsuccessful analysis 93 rock samples were collected in the ield in the North Pyrenean Zone, in particular around the Metamorphic Internal Zone and the contact with the non- or less metamorphic North Pyrenean Zone. he goal was to establish the peak temperatures at various locations and produce a map that might help unravel the structure of the Metamorphic Internal Zone and its contacts. 41 Samples were processed into thin sections, and 13 samples were analysed using the RSCM technique. Most samples were white or pinkish, highly recrystallized marbles. he organic content proved insuicient to ind enough good spots to analyse, oten only resulting in a single, two, or three spectra per sample. Several samples returned very large diferences between measurements. he analysis was therefore not statistically reliable enough to use. he sample names, locations and unsuccessful Raman results are added here. Following this failed analysis, the samples were handed to Corentin Gouache, a master student at the time, for reanalysis. His results (among other things) can be found in his MSc thesis [Gouache, 2017].

References Beyssac, O., B. Gofé, C. Chopin, and J. N. Rouzaud (2002), Raman spectra of carbonaceous material in metasediments: A new geothermometer, J. Metamorph. Geol., 20(9), 859–871, doi:10.1046/j.1525- 1314.2002.00408.x. Gouache, C. (2017), Etudes tectono-métamorphiques des brèches de la zone interne métamorphique des Pyrénées ariégeoises., Université de Lorraine. Lahid, A., O. Beyssac, E. Deville, F. Negro, C. Chopin, and B. Gofé (2010), Evolution of the Raman spectrum of carbonaceous material in low-grade metasediments of the Glarus Alps (Switzerland), Terra Nov., 22(5), 354–360, doi:10.1111/j.1365-3121.2010.00956.x. Appendix D. Raman data 131 1 0.44 443 4 0.58 0.05 385 12 4 0.40 0.23 462 51 5 0.71 0.01 323 4 4 0.58 0.05 385 12 1 0.44 443 3 0.52 0.04 411 10 5 0.71 0.01 323 4 42.811962 1.988238 42.816399 1.986591 42.834396 1.870940 42.846058 1.874768 42.816399 1.986591 42.811962 1.988238 42.834396 1.870940 42.846058 1.874768 position (WGS84) position (WGS84) Barremian? 42.799599 2.068770 5 0.32 0.05 500 9 Barremian? 42.799599 2.068770 5 0.32 0.05 500 9 ‐ ‐ limestone 42.799179 2.035431 2 0.27 0.38 525 119 limestone 42.799179 2.035431 1 0.53 403

Berriasian Berriasian

dolomite flysch 42.896000 1.932800 4 0.61 0.08 371 18 dolomite flysch 42.896000 1.932800 4 0.61 0.08 371 18 limestone 42.816900 1.984900 4 0.68 0.07 339 15 limestone 42.816900 1.984900 4 0.68 0.07 339 15 limestone 42.815617 1.921148 3 0.15 0.17 574 43 limestone 42.815617 1.921148 2 0.06 0.08 614 27

sacharoid 42.813431 1.986271 4 0.33 0.22 496 50 sacharoid 42.813431 1.986271 3 0.43 0.06 448 16

fit

crystalline crystalline shale 42.855700 1.877900 5 0.69 0.02 332 3 shale shale shale 42.855700 1.877900 5 0.69 0.02 332 3

marty 42.804417 1.998489 3 0.41 0.09 457 24 marty 42.804417 1.998489 3 0.41 0.09 457 24

grey grey of

limestone siltstone? limestone siltstone? heavy heavy

‐ ‐ black black black black marble, marble,

meta meta limestone, marble? 42.814163 1.986788 4 0.53 0.05 405 11 marble? 42.814163 1.986788 4 0.53 0.05 405 11 limestone,

coarse coarse

grey grey/black grey grey grey/black grey

crystalline crystalline

quality

ignored

of

spectra

Frau dark Frau albian Frau albian Frau dark Frau albian Frau albian regardless

la la la la la la fitting

Sault meso Sault meso

de de de de de de

de de

badly

included,

and

spectra

all outliers

1 Belesta black/dark 1 Belesta black/dark

with with

5 Gorges 3 Gorges 6 Gorges 3 Gorges 5 Gorges 6 Gorges 16A Niort 5 Comus grey 22 Mazuby grey, 4A Roquefeuil white 10A Rodome black 1 Roquefeuil black 16A Niort 1 Roquefeuil black 10A Rodome black 22 Mazuby grey, 5 Comus grey 4A Roquefeuil white 6 Roquefeuil dark 3a Roquefeuil dark 2 Roquefeuil cleavage 6 Roquefeuil dark 3a Roquefeuil dark 2 Roquefeuil cleavage

‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ P15 P15 P15 P15 P15 Roq P15 Roq Roq Frau Frau Belesta Frau results sample name location lithology lat long n R2 stdev T (°C) SE (°C) P15 Roq P15 P15 P15 P15 Roq P15 Roq Frau Frau Belesta Frau results sample name location lithology lat long n R2 stdev T (°C) SE (°C) Raman results 132 Appendix D. Raman data

D1 (~1350) D2 (~1620) G (~1590) Beyssac et al 2002 count? sample spectrum area area area R2 T1 1Belesta 1 16HR_3009 30073.5 4766.73 8365.17 0.69605883 331.253822 1 16HR_3010 49846.9 11438.4 12450.1 0.676024 340.16932 0 16HR_3011 ‐ ‐ 1 16HR_3012 70525.5 38676.2 26469.7 0.51982584 409.677499 1 16HR_3013 45495.3 21112.8 18374.5 0.53534841 402.769956 mean 0.60681427 370.96765 stdev 0.07973226 35.4808552 SE 0.03986613 17.7404276

0 Frau 6 16HR_3014 ‐ ‐ 1 16HR_3015 23758 7599.66 3722.14 0.67725586 339.621144 0 16HR_3016 ‐ ‐ 1 16HR_3017 17729.9 4667.89 3029.41 0.69728086 330.710015 1 16HR_3018 73270.3 15447.4 18498 0.68339152 336.890774 1 16HR_3019 51030 12398.3 10516.9 0.69010565 333.902988 1 16HR_3020 50096.9 5108.23 14189.8 0.72191009 319.750011 mean 0.69398879 332.174986 stdev 0.01730469 7.70058535 SE 0.00773889 3.44380646

1Frau 5 16HR_3021 9750.03 1933 2208.06 0.70189092 328.658539 0 16HR_3022 ‐ ‐ 1 16HR_3023 18752.3 2151.27 6338.8 0.68835054 334.684011 1 16HR_3024 16599.4 289.45 5720.43 0.73418525 314.287562 1 16HR_3025 11032.5 1714.09 2645.46 0.71676612 322.039075 1 16HR_3026 17297.4 1875.8 4438.76 0.73256943 315.006605 mean 0.71475245 322.935158 stdev 0.01975863 8.79258892 SE 0.00883633 3.93216531

1Frau 3 16HR_3027 4158.81 449.9 3202.15 0.53243945 404.064445 1 16HR_3028 4011.64 399.084 2948.57 0.54511207 398.425128 0 16HR_3029 613.335 94.5259 9085.28 ‐ ‐ 0 16HR_3030 ‐ ‐ 1 16HR_3031 2762.75 473.921 2632.12 0.47075283 431.514989 mean 0.51610145 411.334854 stdev 0.03978092 17.7025098 SE 0.02296753 10.2205488

1Roq 2 (XZ) 16HR_3034 10662.8 1225.94 2645.7 0.73362304 314.537749 1 16HR_3035 25907.2 1924.38 7660.58 0.72994149 316.176039 1 16HR_3036 12676.6 4248.3 2434.17 0.65481451 349.607542 1 Roq 2 (YZ) 16HR_3037 12046.8 3510.74 4613.03 0.59724638 375.225359 mean 0.67890635 338.886672 stdev 0.06544022 29.1208982 SE 0.03272011 14.5604491

1Roq 3a 16HR_3058 8817.79 549.671 5226.4 0.60421228 372.125536 1 16HR_3059 7460.81 1835.03 3704.45 0.57389566 385.616432 1 16HR_3060 4786.98 1427.61 3344.68 0.50076836 418.158078 1 16HR_3061 7378.43 1081.19 3352.93 0.62462635 363.041274 mean 0.57587566 384.73533 stdev 0.05423597 24.1350049 SE 0.02711798 12.0675025

1P15‐1 16HR_3062 14551.3 5453.1 7221.36 0.53446809 403.161699 1 16HR_3063 20402.8 5035.03 10144.1 0.57340341 385.835483 1 16HR_3064 18049.6 4028.55 10870.1 0.54781665 397.22159 1 16HR_3065 29867.2 12038.1 22739.8 0.462018 435.401989 mean 0.52942654 405.40519 stdev 0.04775454 21.2507721 SE 0.02387727 10.6253861

1Roq 6 16HR_3066 10316.4 4023.03 8854.81 0.44478284 443.071635 extremely poor data and very large fit error 0 16HR_3067 ‐ ‐ no satisfactory fit 0 16HR_3068 ‐ ‐ 0 16HR_3069 ‐ ‐ no satisfactory fit mean 0.44478284 443.071635 stdev #DIV/0! #DIV/0! SE #DIV/0! #DIV/0!

1P15‐4A 16HR_3070 4915.35 4197.25 3605.59 0.38648188 469.015563 poor data and extremely large fit error 0 16HR_3071 0 1574.44 2771.33 ‐ ‐ very poor data an extremely large fit error 0 16HR_3072 ‐ ‐ 1 16HR_3073 10158.6 4992.77 9502.95 0.41204138 457.641587 very poor data and very large fit error 0 16HR_3074 ‐ ‐ 1 16HR_3075 10520.8 3678.83 6683.67 0.50379011 416.813401 very poor data and very large fit error mean 0.43410446 447.823517 stdev 0.06168784 27.451089 SE 0.03561549 15.8488936

0 P15‐10A 16HR_3078 ‐ ‐ 1 16HR_3079 1530.34 469.661 2322.95 0.35400355 483.46842 poor data. Surprisingly good fit but temps are all wrong 1 16HR_3080 5524.7 725.655 9236.14 0.35674308 482.249327 relatively reasonable data. Very large fit error Appendix D. Raman data 133

1 16HR_3081 2355.04 423.096 5048.3 0.30090836 507.095781 1 16HR_3082 2178.77 503.138 6298.41 0.24261613 533.035822 best data from set? 1 16HR_3083 1957.35 395.433 3519.49 0.33332067 492.6723 so‐so data, ok fit mean 0.31751836 499.70433 stdev 0.04744918 21.1148837 SE 0.02121992 9.44286305

1P15‐16A 16HR_3084 3951.59 256.435 6791.36 0.35925554 481.131285 1 16HR_3085 1749.68 501.611 1107.31 0.520955 409.175023 only slightly better than 3084 0 16HR_3086 ‐ ‐ 1 16HR_3087 4954.89 963.05 7776.86 0.36180813 479.995381 worse than 3085 mean 0.41400623 456.76723 stdev 0.09262915 41.219973 SE 0.05347947 23.7983625

0 P15‐22 16HR_3088 0 4017.6 1364.08 ‐ ‐ poor data, poor fit 1 16HR_3089 16672.5 6911.5 7580.66 0.53498097 402.933469 poor data, poor fit 0 16HR_3090 ‐ ‐ mean 0.53498097 402.933469 stdev #DIV/0! #DIV/0! SE #DIV/0! #DIV/0!

1P15‐5 16HR_3091 1343.8 605.018 9303.05 0.11942906 587.854069 almost no signal, very poor fit 1 16HR_3092 0 3484.85 3730.56 0 641 weak signal, poor fit 0 16HR_3093 1609.21 2171.83 1041.27 ‐ ‐ veryu weak signal, very poor fit. All three spectra are different mean 0.05971453 614.427035 stdev 0.0844491 37.5798482 SE 0.05971453 26.5729655

Stellingen 135 Stellingen

It is a Dutch tradition to place a list of propositions (‘stellingen’) at the end of the thesis. Originally, these were statements related to the topic of research that the PhD candidate had to successfully defend publicly. Over time, the emphasis shited more and more towards judging the written work of a PhD candidate, and the propositions lost their original purpose. Nowadays, the list of stellingen should still include some related to the topic, but may also contain unrelated topics that the PhD candidate happens to be interested in or feels strongly about. he last stelling is usually an attempt at humor.

• Using a balanced and fully restored cross section in combination with fault timings to calculate shortening gives better results with equal or less work than stepwise restoration.

• he rate of displacement of a suprasalt cover unit in relation to its basement can be used to recognise gravitational sliding in convergent settings.

• A lux steady state (accretionary inlux = erosional outlux) of orogens or orogenic wedges is iction. • here should be a half-open access science publishing platform analogous to Spotify or Youtube. • he ratio of references cited in a study that are also cited by other references cited in the same study to ‘fresh’ references can be used as a measure of quality.

• Isolation and motivation are inversely correlated. • Successful procrastination is a ine art that requires lots of practice.

Du système de rit à l'orogène à double vergence : Un modèle évolutif basé sur l’étude de cas des Pyrénées Orientales et une étude des facteurs de contrôle à partir des modèles numériques. Les orogènes à double vergence sont classiquement déinis comme deux prismes critiques opposés (pro et retro) qui évoluent ensemble. Les études récentes montrent que les rétro-prismes et leurs bassins d’avant-pays associés se comportent diféremment des pro-prismes. Cependant, ni les facteurs qui mènent un orogène à devenir doublement vergent, ni la relation entre le pro- et rétro-prisme ne sont bien compris. Le but de cette étude est d'améliorer notre connaissance 1) de la relation entre le pro- et le rétro-prisme pendant l'orogénèse, 2) des facteurs contrôlant l'évolution d'un orogène à double vergence, et 3) d’un lien dynamique possible entre le pro- et le rétro-prisme. Répondre à ces questions nécessite une connaissance améliorée de l'évolution d'un orogène à double vergence. Nous nous sommes concentrés sur les Pyrénées Orientales, en raison de la grande quantité de données disponibles. Nous avons efectué une étude de terrain tectono-stratigraphique détaillée à l’est du Massif de Saint Barthelemy et dans l’avant-pays autour de Lavelanet (plaque Européenne). Notre interprétation d’une coupe restaurée intègre une coniguration crustale pré-orogenique en tant qu'une marge hyper-amincie. Nous relions l'évolution détaillée du rétro-prisme à celle du pro-prisme (plaque Ibérique), ain de mieux contraindre la dynamique à l'échelle crustale. Nous subdivisons l'évolution des Pyrénées Orientales en quatre phases. La première phase (Crétacé Supérieur) est caractérisée par la fermeture d'un domaine de manteau exhumé entre les plaques et l'inversion synchrone d'un système de rit riche en sel et thermiquement déséquilibré. Le raccourcissement était distribué de façon égale entre les deux marges pendant cette première phase d’inversion. Une phase de quiescence (Paléocène), limitée au rétro-prisme, enregistre la transition entre l'inversion et la phase de collision. La phase de collision principale (Éocène) enregistre le taux de raccourcissement le plus élevé, et était principalement accommodé dans le pro-prisme. Pendant la phase inale (Oligocène) le rétro-prisme était largement inactif et le raccourcissement du pro-prisme a ralenti. Cela démontre que la relation entre le pro- et rétro-prisme change avec le temps. Nous avons utilisé des modèles numériques 2D thermomécaniques à l'échelle lithosphérique pour simuler l'évolution d'un orogène à double vergence s'initie après avec un rit. Nos résultats montrent un modèle évolutif similaire à celui observé dans les Pyrénées Orientales avec une phase d'inversion du rit approximativement symétrique suivie d'une phase de collision asymétrique. L'héritage du rit est essentiel pour permettre le développement d’un orogène à double vergence. Des autres facteurs, comme les processus de surface et la déformation de la couverture, ont un efet signiicatif sur la structure crustale et la répartition du raccourcissement entre les deux prismes. Un niveau de décollement (sel) à la base de la couverture favorise la formation d'un empilement antiformal d’écailles crustales, similaire à la géométrie observée dans la Zone Axiale des Pyrénées, en formant un prisme à faible pente qui force la déformation crustale à se concentrer dans l'arrière-pays. Enin, nous montrons que l'évolution des pro- et rétro-prismes est inextricablement liée : des événements ou des conditions d'un côté de l'orogène ont un efet direct sur l'autre côté de l'orogène. Ceci est clairement démontré dans nos modèles par des variations constantes des taux de raccourcissement du pro- et rétro-prisme en réponse à l'accrétion dans le pro-prisme. Le Haut Atlas (Maroc) et Pyrénées peuvent être respectivement considérés comme des exemples d'inversion de rit symétrique et de phases de collision asymétrique ultérieures. From rit system to doubly vergent orogen: An evolutionary model based on a case study of the Eastern Pyrenees and controlling factors from numerical models. he doubly vergent nature of some natural orogens is classically understood as two opposing thrust wedges (pro and retro) that comply with critical taper theory. he evidence that retro-wedges and their associated basins behave diferently from their pro- wedge counterparts has been steadily increasing over the past few decades. However, what causes an orogen to become doubly vergent is currently not well understood. Nor is the relationship between the pro- and retro-wedge during the evolution of a doubly vergent orogen. It is the aim of this work to improve our understanding of: 1) how the pro- and retro-wedges relate to each other during the orogenic process, 2) what factors control the evolution of a doubly vergent orogen and 3) a possible link between the pro- and retro-wedge. Answering these questions requires an improved knowledge of the evolution of a doubly vergent orogen. We focussed on the Eastern Pyrenees as a type example of a doubly vergent orogen, due to the large amount of available data. We performed a detailed tectonostratigraphic study of the retro-foreland of the Eastern Pyrenees (European plate), updating the interpretation based on recent insights into its hyperextended rit origins. We link the evolution of the retro-foreland to that of the pro-foreland (Iberian plate) in order to derive insight into the crustal scale dynamics. Based on cross section restoration, reconstructed shortening rates and subsidence analysis, we subdivide the East Pyrenean evolution into four phases. he irst (Late Cretaceous) phase is characterised by closure of an exhumed mantle domain between the European and Iberian rited margins, and simultaneous inversion of a salt-rich, thermally unequilibrated rit system. Shortening was distributed roughly equally between both margins during this early inversion phase. Following inversion, a quiescent phase (Paleocene) was apparently restricted to the retro- foreland. his phase may record the period of transition between inversion and full collision in the Eastern Pyrenees. he main collision phase (Eocene) records the highest shortening rates, which was predominantly accommodated in the pro-wedge. Retro-wedge shortening rates were lower than during the rit inversion phase. During the inal phase (Oligocene) the retro- wedge was apparently inactive and shortening of the pro-wedge slowed. his demonstrates that the relationship between the pro- and retro-wedges changes through time. We used lithosphere-scale thermo-mechanical numerical models to simulate the evolution of a doubly vergent orogen. Our results show a similar evolutionary pattern as observed in the Pyrenees: A roughly symmetrical rit inversion phase is followed by an asymmetric collision phase. Rit inheritance was found to be essential for enabling double vergence. Other factors, such as surface processes and thin-skinned deformation, were found to have a signiicant efect on the crustal structure and strain partitioning between both wedges. A salt décollement layer in the sedimentary cover promotes the formation of a crustal antiformal stack such as observed in the Pyrenees and Alps by forming a wide and low-taper thin-skinned fold-and-thrust belt that forces crustal deformation to focus in the hinterland. Finally, we show that the evolution of the pro- and retro-wedges is inextricably linked: events or conditions on one side of the doubly vergent orogen have an immediate efect on the other side of the orogen. his is clearly demonstrated in our models by constant variations in shortening rates of the pro- and retro-wedge in response to accretion of new pro-wedge thrust sheets. he High Atlas (Morocco) and Pyrenees can be seen as examples of symmetric rit inversion and later asymmetric collision phases, respectively. PhD hesis, Grool, A.R., 2018, Université de Lorraine, Nancy, France & University of Bergen, Norway