PONTIFICIA UNIVERSIDAD CATOLICA DE CHILE SCHOOL OF ENGINEERING INTRA-ARC TECTONICS IN THE SOUTHERN ANDES: CONSTRAINTS FROM COMBINED FIELD GEOLOGY AND CRUSTAL SEISMICITY GERD GUSTAV SIELFELD CORVALÁN Thesis submitted to the Office of Graduate Studies in partial fulfilment of the requirements for the Degree of Doctor in Engineering Sciences Advisor: JOSÉ CEMBRANO. DIETRICH LANGE Santiago de Chile, December, 2018 © MMXVIII, GERD GUSTAV SIELFELD CORVALAN PONTIFICIA UNIVERSIDAD CATOLICA DE CHILE SCHOOL OF ENGINEERING INTRA-ARC TECTONICS IN THE SOUTHERN ANDES: CONSTRAINTS FROM COMBINED FIELD GEOLOGY AND CRUSTAL SEISMICITY GERD GUSTAV SIELFELD CORVALÁN President of the Committee GUSTAVO LAGOS Members of the Committee: JOSÉ CEMBRANO DIETRICH LANGE GONZALO YAÑEZ PABLO SALAZAR TRhAeFsiAs EsLu bRmIDitDteEd LtLo the Office of Graduate Studies in partial fulfilment of the requirements for the Degree of Doctor in THOMAS MITCHELL Engineering Sciences Santiago de Chile, December, 2018 © MMXVIII, GERD GUSTAV SIELFELD CORVALAN This work is dedicated to Verónica Corvalán and Kurt Sielfeld For melting me with nature iii Acknowledgements Deep thanks with tectonic forces to all the people that kept always an eye on the development of this thesis, and were physically and/or mentally connected with me during these five years. I want to thank all my family, Christel, Kurt, Rolf, nephews, María Isabel Corvalán (R.I.P.), Gloria Nicolás (R.I.P.), friends, colleagues and people from life, that at some point freely took the time to talk and/or discuss the topic here presented. From the most universal perspective, I am deeply and spiritually thankful to José, my mentor, who showed me tools and the way how-to-use them, to face the living-experience of growing in science, life and love for nature. The human team surrounding José is unique and I am proud to be part of it. Thanks to the Geoscience-PUC team & collaborators, for its originality and heterogeneity. I want to sincerely thank Gonzalo Yáñez, Gloria Arancibia, Mariel Castillo, Felipe Aron, Tom Mitchell, Luis Lara, W. Ashley Griffith, Jorge Crempien, Carlos Marquardt, Constantino Mpodozis and Andrés Veloso for their conspicuous generosity. I am also very proud and thankful for having the chance to meet and learn from Dietrich, my co-advisor, who guided me from the very beginning on the discovery and understanding of the art of seismicity. I took a tremendous step in my career thanks to the patience and generosity of Dietrich. I also what to thank Heidrum Kopp and the GEOMAR – Dynamics of ocean floor team for the kind reception and cordiality I experienced during my three internships with them. Special thanks to Jasmin Mögeltönder, Florian Petersen, Margit Weiprich, Antje Schwimoska, Bahne Thomsen, Heiner + Bärbel Hoffmann, Anna Loges, Marina Viñas, Laura Gómez de la Peña, Marlon Pardo and Jairo Orduy for the support, friendship, ‘Fahrräder’, ‘Bierchen’ and conversations during my days in Kiel. Special thanks to Diego Morata for supporting one fundamental step of this thesis: the deployment of the SAIAS seismic network. Seismometers and EarthDataLoggers for the seismic network were provided by the Geophysical Instrument Pool Potsdam (GIPP, GFZ Potsdam, loan: LOFZ). I thank Christian Haberland for his professional and technical support. I am thankful to all Chilean landowners, companies and institutions for allowing us to install the seismic station on their property, including CONAF, Oscar Iturrieta, Comunidad Namuncai, Mario and Ursula from El Coihüe, A. Trepper and Ch. Gibson, Mario Neihual (Trafún rural School), María Mardones and iv husband, the Pilahuepillan School, Aldo Matus, Cecilia from El Rincon hotsprings, among many others. I have to recognize the tremendous help provided by the field crews during the seismic deployment and services: thanks to P. Guzman, A. Stanton-Yonge, N. Pérez-Estay, P. Iturrieta, P. Azúa, D. Tardani, M. Lizama, J. Riquelme, J. Puratic, F. Rodriguez, P. Pérez-Flores, R. Rivera, M.P. Lira, K. Muñoz, V. Vicenzio, E. Lira, A. Órdenes, A. Salazar; without your help this work could never have been concluded. I am also thankful to the OVDAS crew for providing their seismicity catalogue, which allows the improvement of the detection quality and magnitude calibration of the SAIAS catalogue. I also want to thank all landowners around the Callaqui and Tatara-San Pedro-Pellado volcanoes, specially to the Energy Development Company, ENEL and the people of Los Álamos village for their generosity (Marcelo Gonzalez, Juano, Boris, Maggie, tía María, Roberto Torres and Roberto Torres Jr. among others). Frank thanks to my research-team mates and friends Pamela Pérez-Flores, Ashley Stanton-Yonge, Rodrigo Gomila, Javiera Ruz, Nicolas Pérez-Estay, Pablo Iturrieta, Pablo Sánchez- Alfaro, Danielle Tardani, Gert Heuser, Pedro Guzman, Tomás Roquer, Isabel Santibañez, Andrés Bosch, Tiaren Garcia, Felipe del Valle, and Elias Lira. I am also thankful to my friends, just because of their unconditional friendship during these years: Juan I. Rubio, Julian Rosenblatt & fam., Johannes Dietsche & fam., Alfred & fam., Andrea Brogi & fam., Sebastian Salas & fam., Felipe Durán, J. Fco. Jullian & fam., Ludovico, Monica, Iside e Dorotea Ginanneschi, Carlos Lama, Tomas Schlack, Melanie Duclos, Sebastian Martini & fam., Benjamín Malandre & fam., Andrés Lama & fam., Antonella Bernucchi, Jorge Knabe, Danilo González & fam., Fco. J. Ossa & fam., Rodrigo Galleguillos & fam., Delia Rodriguez, Roberto Cristy & fam., Constanza Jara, Valentina Catalán, Juan y Margarita Escobedo, Liliana Espinoza, Marcelo Vega, Lucia Misi, J.M. Bustos & fam., Vanna, Rayén Gutierrez, Fernanda Corvalán, Leonardo and Camilo Paredes-Mariño, Rosa Vega, Francisco and Mari, Agnés Bes, German Rey, Marcia de Mello and Thomas Eggers. Finally, special thanks to Joa, who has been carefully treating with love my emotional diseases. v Abstract In active volcanic chains, there is a strong interplay between deformation and volcanism. The aim of this thesis is to better understand how long-term tectonics and current crustal brittle deformation are coupled and how the resulting architecture and local stress fields can drive volcanism and hydrothermal activity. To achieve this, I studied both long-term and short-term deformation at a classic subduction setting, by using the Southern Andes active magmatic arc as natural laboratory. Here, the intra-arc architecture is dominated by margin-parallel (e.g. Liquiñe-Ofqui fault system, LOFS) and margin-transverse faults (Andean Transverse Faults, ATF). These long-lived crustal faults are in spatial and temporal association with the loci of geofluids circulation and emplacement. The causal relationship among these geological features have been largely studied; however, many scientific questions regarding the role that these fault systems play in the development of volcanism and geothermal activity remain unanswered: How do these faults interact and how this interaction promotes permeability enhancement forwarding volcanism and geothermal activity? Does this causal relationship imply a thermo-mechanical feedback loop at the volcanic-chain-, and/or volcanic- complex-scales? The brittle-deformation nature of faults and fractures was examined using (1) crustal seismicity techniques for a geologically instantaneous/regional-scale approach, and (2) structural analyses of exhumed faults and hydraulic fractures (e.g. dikes and volcanic fissures) for a long- term/basement- and volcano-scale approach. The instantaneous/regional approach (1) consisted of monitoring the intra-arc crustal seismicity for a 16-month time-window (i.e. a snap-shot in geological time) with 34 seismic stations covering a ~200 km N-S long section of the Southern Volcanic Zone (38˚S-40˚S). The study region includes segments of the LOFS and ATF, as well as large stratovolcanoes such as Llaima, Lonquimay and Villarrica. A total of 356 crustal events with magnitudes between Mw 0.6 and Mw 3.6 were identified. Hypocenters occur at all depths down to 40 km, demarking an arc- orthogonal convex shape of the crustal seismogenic layer bottom as seen in a W-E section. The latter is characterized by up to 40 km of thickness in the forearc and approximately 12 km of thickness and relative high seismicity at the intra-arc zone. Within here, focal mechanisms indicate strike-slip to oblique-slip faulting consistent with an ENE-WSW oriented far-field maximum principal stress axis and local stress-field compartmentalization. In turn, the long-term/basement- and volcano-scale geological approach (2) was conducted by exhaustive field mapping at the Callaqui stratovolcano (CSV) and the Tarara-San Pedro-Pellado volcanic complex (TSPPC). Dike-geometry and minor eruptive vents vi morphometry analyses at the CSV yields to a consistent model of dike propagation and minor eruptions following an ENE-oriented transtensional wing-crack like array. Such geometry implies a strike-slip component accompanying dominant local extension during the Late-Pleistocene to Holocene volcano-building process. At the TSPPC, in turn, structural analysis of fault-slip data and dike geometry allows the identification of an ENE-oriented transtensional interacting damage zone, limited to the west by margin-parallel seismic faulting at El Melado Fault and to the east, by strike-slip faulting at La Plata and Los Condores ATF (hitherto described). The interacting damage zone accommodates oblique-slip with a large extensional component where the locus of TSPPC volcanism and geothermal activity take place. In summary, the results of this thesis indicate that active faulting within the intra- arc region is consistent with long-lived fault geometries and kinematics enhancing geofluid migrations along local transtensional domains. The maximum rate of crustal seismicity occurs within the intra- arc, where elevated geothermal gradients are expected and the seismogenic layer is thinner (<12 km) compared to forearc and retro-arc regions. The observation suggests that elevated geothermal gradients lead to a thinner brittle portion of crust, which is then mechanically easier to (re-)break up. Furthermore, seismicity reveals detailed information on how ongoing deformation accommodates within crustal faults characterized by diverse orientations and kinematics. A complex bulk strain compartmentalization pattern is documented, dominated by stick-slip mechanism on critically loaded faults. The slip interaction among faults enhances rock damage and permeability development on optimally oriented hybrid faults and extensional fractures. Seismological and field data contained in this thesis improves the understanding of the tectonic role of margin parallel faults and ATF in magma migration processes within the upper crust. These results are crucial for further understanding the mechanisms of fault interactions and crustal fluid flow. This in turn may help improving geothermal exploration strategies and seismic hazard assessment. vii Resumen En cadenas volcánicas activas existe una fuerte interacción entre deformación y volcanismo. El objetivo de esta tesis es precisar los mecanismos mediante los cuales la tectónica de largo plazo y la deformación frágil actual en la corteza se encuentran acopladas. También se busca comprender como la arquitectura y campos de estrés resultantes pueden direccionar la actividad volcánica e hidrotermal. Para lograr esto, estudié la deformación de largo y corto plazo en un contexto de subducción, utilizando el arco magmático de los Andes del Sur como laboratorio natural. En la región del intra-arco andino, la arquitectura cortical es dominada por fallas paralelas al margen (e.g. Sistema de Fallas Liquiñe-Ofqui, SFLO) y fallas transversales al margen (FTA, Fallas transversales a los Andes). Estas fallas corticales de larga vida se encuentran en asociación espacial y temporal con el loci de la circulación y emplazamiento de geofluidos. La relación causal entre estos elementos geológicos ha sido largamente estudiada, no obstante aún quedan interrogantes científicas sin responder con respecto al rol que estos sistemas de fallas cumplen en el desarrollo del volcanismo y actividad geotermal. ¿Cómo interactúan estas fallas y cómo esa interacción contribuye al desarrollo de permeabilidad precediendo el volcanismo y actividad geotermal? ¿Implica acaso esta relación causal un ciclo termo-mecánico retroalimentado tanto a escala de la cadena volcánica como a la escala de un complejo volcánico? La naturaleza de la deformación frágil en fallas y fracturas fue examinada usando (1) técnicas de sismicidad cortical para una perspectiva geológicamente instantánea a escala regional, y (2) análisis estructural de fallas y fracturas hidráulicas (e.g. diques y fisuras volcánicas) exhumadas para un enfoque de largo plazo a escalas del basamento y volcán. La perspectiva instantánea a escala regional consistió en monitorear la sismicidad cortical en el intra-arco por un lapso de 16 meses (i.e. un breve instante en tiempo geológico) mediante una red de 34 estaciones sismológicas cubriendo una franja de 200 km de largo en la Zona Volcánica Sur (38˚S-40˚S). La región de estudio incluye segmentos del SFLO y FTA, y también estratovolcanes como Llaima, Lonquimay y Villarrica. Un total de 356 eventos sísmicos corticales fueron identificados con magnitudes entre Mw 0.6 y Mw 3.6. Los hipocentros ocurren en los primeros 40 km de profundidad, demarcando la capa sismogénica cortical con una geometría convexa vista en sección. Esta última se caracteriza por tener aproximadamente 12 km de espesor y sismicidad relativamente alta en la zona de intra-arco. En esta zona, los mecanismos focales indican fallamiento de rumbo y de slip oblicuo consistentes con un eje de estrés máximo orientado ENE-WSW para el campo lejano y compartimentalización local del campo viii de estrés. En cambio, el enfoque geológico de largo plazo a escalas de basamento y volcán (2) fue conducido mediante levantamientos de campo en el estratovolcán Callaqui (SVC) y el complejo volcánico Tatara-San Pedro Pellado (CTSPP). Los resultados de análisis geométrico en diques y morfométrico en centros eruptivos menores del SVC resultan en un modelo consistente de propagación de diques y erupciones menores siguiendo una estructura transtensional tipo fractura- alada de orientación ENE. Tal geometría implica extensión local dominante con un componente de rumbo durante el proceso de construcción del volcán entre el Pleistoceno tardío y el Holoceno. En el CTSPP, el análisis estructural de datos de deslizamiento en fallas y de la geometría de diques permite la identificación de una zona de daño de interacción transtensional orientada ENE. Esta se encuentra limitada al oeste por fallamiento sísmico paralelo al margen en la Falla El Melado y al este, por fallamiento de rumbo en las fallas La Plata y Los Cóndores (aquí definidas). La zona de daño de interacción acomoda slip oblicuo con un gran componente de extensión donde tiene lugar el locus del volcanismo y actividad geotermal del CTSPP. En resumen, los resultados de esta tesis indican que el fallamiento activo en la región de intra-arco es consistente con la geometría y cinemática de fallas de larga vida favoreciendo la migración de geofluidos a lo largo de dominios transtensivos locales. La mayor tasa de sismicidad cortical ocurre en el intra-arco, donde se esperan gradientes geotermales elevados y donde la capa sismogénica es más delgada (<12 km) en comparación a las regiones de antearco y trans-arco. La observación sugiere que gradientes geotermales elevados dan lugar a porciones más delgadas de corteza frágil, la cual, en consecuencia, es mecánicamente más fácil de (re)fracturar. Más aún, la sismicidad revela información detallada del modo en que la deformación continua se acomoda en fallas corticales de orientación y cinemática diversa. También, se documenta como la deformación es compartimentalizada en fallas a lo largo del intra-arco de los Andes del sur, regidas por mecanismos de stick-slip. En conclusión, la interacción de deslizamientos entre fallas favorece el desarrollo de zonas de daño y aumento de permeabilidad en fallas híbridas y fracturas extensionales óptimamente orientadas. Los datos sismológicos y de campo contenidos en esta tesis contribuyen al entendimiento del rol tectónico de las fallas paralelas al margen y FTA en procesos de migración de magmas en la corteza alta. Estos resultados son cruciales para el perfeccionamiento de estrategias de exploración geotermal y evaluación del peligro sísmico. ix Table of contents 1 THESIS SUMMARY ............................................................................................................................. 16 1.1 INTRODUCTION ............................................................................................................................... 16 1.2 THEORETICAL BACKGROUND, RESEARCH PROBLEM AND HYPOTHESES ..................................................... 21 1.3 RESEARCH QUESTIONS, SPECIFIC OBJECTIVES AND METHODOLOGIES ....................................................... 24 SPECIFIC OBJECTIVE AND METHODOLOGY 1: DECIPHERING THE INSTANTANEOUS TECTONIC STATE OF TECTONICALLY CONTROLLED ARCS. .............................................................................................................................................. 26 SPECIFIC OBJECTIVE AND METHODOLOGY 2: THE TECTONIC SIGNIFICANCE OF INTRA-VOLCANIC DIKE AND MINOR ERUPTIVE VENTS: STATE OF STRESS AT THE TIME OF VOLCANISM .............................................................................................. 28 SPECIFIC OBJECTIVE AND METHODOLOGY 3: ESTIMATION OF PALEOSTRESS CONDITIONS OF VOLCANIC BASEMENT ARCHITECTURE ..................................................................................................................................................... 29 1.4 CONCLUDING REMARKS .................................................................................................................... 30 1.5 ONGOING AND FUTURE PROJECTS ...................................................................................................... 32 1.6 PUBLICATIONS AND CONFERENCE PARTICIPATIONS .............................................................................. 33 PUBLICATIONS RESULTING FROM THIS RESEARCH .................................................................................................... 33 ABSTRACT AND CONFERENCE PARTICIPATIONS ASSOCIATED WITH THIS RESEARCH ....................................................... 33 PUBLICATIONS AND ABSTRACTS RESULTING FROM SIDE PROJECTS ............................................................................. 35 2 INTRA-ARC CRUSTAL SEISMICITY: SEISMO-TECTONIC IMPLICATIONS ..................................... 37 2.1 INTRODUCTION ............................................................................................................................... 37 2.2 TECTONIC SETTING OF SOUTHERN ANDES ........................................................................................... 40 LIQUIÑE-OFQUI FAULT SYSTEM (LOFS) AND ANDEAN-TRANSVERSE FAULTS (ATF) .................................................... 41 x NATURE AND HISTORICAL RECORD OF INTRA-ARC SEISMICITY .................................................................................... 45 2.3 DATA PROCESSING AND METHODS .................................................................................................... 47 CONSTRUCTION AND VALIDATION OF A MINIMUM ONE-DIMENSIONAL VELOCITY MODEL ............................................. 49 HYPOCENTRE DETERMINATION AND MOMENT MAGNITUDES ................................................................................... 51 FOCAL MECHANISMS ............................................................................................................................................ 52 2.4 RESULTS ......................................................................................................................................... 53 NATURE, MAGNITUDES AND TEMPORAL DISTRIBUTION OF INTRA-ARC SEISMICITY ..................................................... 53 SPATIAL DISTRIBUTION AND KINEMATICS OF INTRA-ARC SEISMICITY ......................................................................... 54 2.5 DISCUSSION .................................................................................................................................... 65 THE FRICTIONAL-PLASTIC TRANSITION IN THE INTRA-ARC REGION .............................................................................. 66 STRAIN PARTITIONING AND COMPARTMENTALIZATION ............................................................................................. 72 FLUID-RELATED SEISMICITY AND THE VILLARRICA 2015 ERUPTION ............................................................................ 78 2.6 SUMMARY AND CONCLUSIONS .......................................................................................................... 80 2.7 ACKNOWLEDGMENTS ....................................................................................................................... 82 3 TRANSTENSION DRIVING VOLCANO-EDIFICE ANATOMY: INSIGHTS FROM ANDEAN TRANSVERSE-TO-THE-OROGEN TECTONIC DOMAINS ...................................................................... 83 3.1 INTRODUCTION AND PROBLEM STATEMENT ......................................................................................... 83 3.2 TECTONIC AND GEOLOGICAL SETTING ................................................................................................. 85 REGIONAL GEOLOGY AND BASEMENT NATURE ......................................................................................................... 90 THE CALLAQUI STRATOVOLCANO .......................................................................................................................... 92 3.3 METHODOLOGY ............................................................................................................................... 96 MEASURING DIKES ............................................................................................................................................... 96 xi MORPHOMETRY OF MINOR ERUPTIVE VENTS ........................................................................................................... 97 3.4 DATA ANALYSES AND RESULTS ......................................................................................................... 101 DIKE GEOMETRY AND SPATIAL DISTRIBUTION ........................................................................................................ 101 MORPHOMETRY OF MINOR ERUPTIVE EVENTS (MEV) ............................................................................................. 103 3.5 DISCUSSION .................................................................................................................................. 109 3.6 CONCLUSIONS ............................................................................................................................... 118 3.7 ACKNOWLEDGEMENTS .................................................................................................................... 119 4 OBLIQUE-SLIP TECTONICS AND GEOFLUID FLOW: A CASE STUDY FROM SOUTH-CENTRAL ANDES. ................................................................................................................................................ 121 4.1 INTRODUCTION AND PROBLEM STATEMENT ....................................................................................... 121 4.2 TECTONIC SETTING ......................................................................................................................... 125 REGIONAL GEOLOGY .......................................................................................................................................... 129 TATARA-SAN PEDRO-PELLADO VOLCANIC COMPLEX .............................................................................................. 132 GEOTHERMAL FEATURES ..................................................................................................................................... 134 4.3 FIELD MAPPING AND DATA ANALYSIS TECHNICS ................................................................................. 135 THE MULTIPLE INVERSE METHOD (MIM) FOR HETEROGENEOUS FAULT-SLIP ANALYSIS ............................................. 137 DILATATIONAL FRACTURE INVERSION: MIXED BINGHAM DISTRIBUTION FOR 3D ORIENTATION DATA ........................... 140 4.4 RESULTS ....................................................................................................................................... 142 MIDDLE TO LATE-MIOCENE EAST-VERGING FOLDING ............................................................................................. 142 FRICTIONAL DEFORMATION AND PALEOSTRESS FIELDS FROM FAULT-SLIP DATA ANALYSIS ........................................... 144 PALEOSTRESS FIELDS FROM DIKE-ORIENTATION DATA INVERSION ............................................................................ 147 4.5 DISCUSSION .................................................................................................................................. 150 xii 4.6 CONCLUSIONS ............................................................................................................................... 156 4.7 ACKNOWLEDGMENTS ..................................................................................................................... 158 5 CONCLUSIONS ............................................................................................................................. 159 5.1 CONCLUSIONS CHAPTER 2. ............................................................................................................. 159 5.2 CONCLUSIONS CHAPTER 3. ............................................................................................................. 160 5.3 CONCLUSIONS CHAPTER 4 .............................................................................................................. 161 6 REFERENCES ................................................................................................................................. 164 7 SUPPLEMENTARY MATERIAL ...................................................................................................... 187 7.1 SUPPLEMENTARY MATERIAL CHAPTER 2 ............................................................................................ 187 INTRODUCTION .................................................................................................................................................. 187 STATION INSTALLATION AND SET-UP. ................................................................................................................... 189 NON-LINEAR LOCATION TECHNIC AND PROCEDURE (NONLINLOC; LOMAX ET AL., 2000) ......................................... 191 DATA SET S1. .................................................................................................................................................... 198 7.2 SUPPLEMENTARY MATERIAL CHAPTER 4 ............................................................................................ 200 DATA BASE ........................................................................................................................................................ 201 xiii Tables Index Table 3-3-1 Summary of 40Ar/39Ar incremental heating experiments ............................................................ 94 Table 3-2 Weight factor for field and remote detected dikes ............................................................................ 97 Table 3-3 Input data for the morphometric analysis ........................................................................................ 109 Table 7-1 Geographical coordinates of stations used for the SAIAS catalogue construction.. .................... 196 Table 7-2 Focal mechanism in the Aki and Richards convention. ................................................................... 197 Table 7-3 List and location of structural stations where FSD was collected .................................................. 200 Figures Index Figure 1-1 Seismotectonic map of southern South America. ............................................................................ 20 Figure 1-2 Synoptic model for the brittle-plastic transition. .............................................................................. 25 Figure 2-1Seismotectonic map of the southern Andes. ..................................................................................... 45 Figure 2-2 Raw waveform example of a typical crustal event . .......................................................................... 49 Figure 2-3 a) Computed 1-D Vp and Vs velocity models. ................................................................................... 50 Figure 2-4 Magnitude-frequency histogram. ....................................................................................................... 53 Figure 2-5 Temporal properties of the seismicity catalogue .............................................................................. 55 Figure 2-6 Southern Andes crustal seismicity ...................................................................................................... 62 Figure 2-7 Crustal seismicity in the northern domain. ....................................................................................... 62 Figure 2-8 Seismicity distribution of SAIAS catalog. ............................................................................................ 63 Figure 2-9 Close-up map view with seismicity of the central domain ............................................................... 64 Figure 2-10 N74˚W oriented swath cross-sections ............................................................................................. 68 Figure 2-11Conceptual model of crustal seismogenic layers along a magmatic arc. ..................................... 78 Figure 3-1 Traditional models for differential stress field .................................................................................. 87 Figure 3-2 Tectonic setting of central-southern Andes. ...................................................................................... 89 xiv Figure 3-3 Regional geological map of Andean intra-arc zone. ......................................................................... 95 Figure 3-4 Callaqui Stratovolcano map (updated from Moreno et al., 1984). ................................................ 100 Figure 3-5 Field pictures of mapped dikes. ........................................................................................................ 108 Figure 3-6 Quickbird image (0.6 m/pixel) ........................................................................................................... 108 Figure 3-7 Field pictures of some minor eruptive vents ................................................................................... 112 Figure 3-8 Map with the distribution of minor eruptive vents. ........................................................................ 113 Figure 3-9 Kinematic model for the overall right-lateral transtensional setting ........................................... 114 Figure 4-1 Seismotectonic map of the south-central Andes.. .......................................................................... 127 Figure 4-2 Volcano-tectonic setting of Transitional Southern Volcanic Zone.. .............................................. 128 Figure 4-3 Geological map of the Tatara-San Pedro-Pellado volcanic complex.. .......................................... 131 Figure 4-4 Structural map with main structural features.. ............................................................................... 143 Figure 4-5 Field pictures and bedding analysis for Oligo-Miocene strata ...................................................... 144 Figure 4-6 Field picture compilation of fault, hybrid fractures and veins ....................................................... 145 Figure 4-7 Glaciated Plmv rocks hosting the Pellado fumarole ....................................................................... 149 Figure 4-8 Pleistocene faults and dikes. ............................................................................................................. 151 Figure 4-9 Mafic dikes with ENE- predominant orientation.. ........................................................................... 154 Figure 4-10 Conceptual model for the Tatara Interaction Damage Zone (TDZ). ........................................... 155 Figure 7-1 Geological map.. .................................................................................................................................. 188 Figure 7-2 Field pictures showing the seismic stations. ................................................................................... 190 Figure 7-3 Instruments used in the seismic deployment . ............................................................................... 191 Figure 7-4 Number of events per day (gray bars) and daily cumulative Mw ................................................. 198 Figure 7-5 Compilation of stress-state paired stereograms and stress tensor solutions ............................ 204 Figure 7-6 Field pictures of folds. ........................................................................................................................ 205 xv 16 1 Thesis summary 1.1 Introduction This work is a contribution to the understanding of the nature of tectonic processes driving deformation within the brittle crust (known in literature as schizosphere, e.g. Scholz, 1988) in spatial and temporal association with volcanic and hydrothermal activity. The natural laboratory to conduct this research is placed within the intra-arc region of south-central and southern Andes (35˚-38˚S and 38˚-~46˚S, respectively; Fig.1-1), where field geology documents a long-lived interaction between tectonics and geofluids circulation/emplacement dynamics (geofluids are defined as any sort of non-solid material from geological origin such as melts, magmas and hydrothermal solutions). The body of this thesis consists of three correlated studies (i.e. self- contained chapters 2, 3 & 4) encompassing the scientific problem, by using diverse methodological approaches including: local seismicity, geological mapping and structural analysis. First, a regional overview of the short-term tectonic state of intra-arc crust has been investigated using microseismicity as seismological approach; then, a long-term approach was carried out by detailed structural mapping on specific volcanic edifices and their basement in order to constrain the geometry and kinematics operating before, during and after the time of volcanism. This leads to two published articles (chapters 1 & 3) and one submitted manuscripts (chapter 2, under review). The core objective of this research is to constrain and understand the geometry and kinematics among diverse crustal faults is spatial association with geofluid circulation and emplacement. Throughout stress field analyses, this study attempts to understand the causal 17 relationship between brittle structures and geofluid conduits in the crust, in order to define the fundamental parameters for permeability enhancement. This principal objective was first tackled from a regional perspective to describe the nature of ongoing brittle deformation in an active volcanic zone. The pulse of crustal intra-arc seismicity has been investigated for a defined region of Southern Andes with well-documented long-lived deformation, active volcanism and hydrothermal activity. Thus, a local seismicity experiment was conducted including 33 stations covering an area of ca. 200x100 km2 between the Callaqui (~38˚S) and Mocho-Choshuenco (~40˚) stratovolcanoes. The experiment setting and findings are presented in chapter 2, corresponding to the published article in the ISI journal Tectonics: “Intra-arc crustal seismicity: Seismo-tectonic implications for Southern Andes Volcanic Zone, Chile”. This article is co-authored by Dr. Dietrich Lange and Dr. José Cembrano. Two main findings arise from this research: a) At an oblique convergence setting, the margin-parallel component of the convergence vector is accommodated not only as margin-parallel slip but also as transpressional and transtentional oblique faulting along Andean transverse fault systems. Strain is thus compartmentalized on discrete tectonic domains that heterogeneously absorb the overall slippage of the regional transpressional regime arising from Nazca-South American plates oblique convergence. b) The schizosphere thickness below the southern Andes, as obtained from earthquake depths is markedly thinner along the magmatic arc, where high geothermal gradients are expected. In turn, the schizosphere thickness is thicker towards the forearc and back- arc/foreland regions, where geothermal gradients are lower. Among the physical parameters controlling the brittle-plastic crustal transition, geothermal gradient might be the first-order 18 parameter for regional scale observations, as depicted by the convex shape of the Seismogenic Layer Bottom, as seen in sectional view (SLB; described along chapter 2). In order to complement the above mentioned observations, the main objective of this thesis was additionally undertaken from field structural-geology approaches. I attempted to understand and determine the tectonic evolution of stress and strain regimes affecting Pleisto- Holocene volcanic complexes, prior, during and after their construction. Thus, tectonically controlled stratovolcanoes and their basements have been mapped in detail, collecting a large amount of geometric and kinematic data. On one hand, the intra-volcanic tectonics has been investigated from the structural analysis of a regional dike swarm as well as from morphometric analysis of minor eruptive vents. This study was carried out at the Callaqui Volcano (38˚S), characterized by tens of ENE-aligned eruptive vents built upon a parallel dike swarms. This work (chapter 3) was published in 2016 on Quaternary International, special issue SAQint3-INQUA Project: Interactions between climatic forcing, tectonics and volcanism during the Late Quaternary: a multidisciplinary approach applied on key regions of South America. The title of the published article is: Transtension driving volcano-edifice anatomy: Insights from Andean transverse-to-the-orogen tectonic domains, and is co-authored by Dr. José Cembrano and Dr. Luis Lara. Lastly, the basement architecture constituting the plumbing system for Quaternary volcanism, has been carefully mapped and analysed for the Tatara-San Pedro-Pellado volcanic complex (TSPPC) in south-central Andes (36˚S; chapter 4). This work includes an evolutionary description from the late Miocene deformation phases until present. However, based on the amount of collected data, the most robust results arise from the description of the nature, geometry and kinematics of faults and fractures affecting the TSPPC basement, this is, enhancing 19 permeability for Pleisto-Holocene volcanism and geothermal activity. This work has been recently shared with co-authors and is intended to be submitted during October to Journal of Volcanological and Geothermal Research under the title “Oblique-slip tectonics and geofluid flow: a case study from South-Central Andes” This work is co-authored by Dr. Andrea Brogi, Dr. José Cembrano, Javiera Ruz, MSc. Ashley Stanton-Yonge, MSc. Pablo Iturrieta and Dr. Pamela Pérez- Flores. The three core chapters abovementioned (2, 3 & 4) are self-contained and explain for three selected case studies new evidence of tectonic factors controlling arc volcanism and/or geothermal activity. The chapters, altogether, constitute a general review of fundamental constrains for tectonic activity coupled with volcanism in southern and south-central intra-arc Andes. These findings are of interest for ongoing research in natural hazard assessment and geothermal exploration. 20 Figure 1-1 Seismotectonic map of southern South America. Main tectonic plate boundaries are indicated with grey lines. Convergence vectors are from Angermann (1999). Segments of the Andes belt (sensu Groeber, 1921), as well as segmentation of Southern Andean Volcanic Zone (SVZ; Stern, 2004) are shown. Main crustal faults are indicated white lines (e.g. Betka et al., 2016; Cembrano et al., 2002). Moment tensors of selected crustal seismic events are indicated in white-black stereograms (gCMT; http://www.globalcmt.org/). Refer to legend for further map information. 21 1.2 Theoretical background, research problem and hypotheses Crustal deformation in active continental margins highly concentrates in the orogenic front and is commonly in spatial association with arc magmatism. Strain partitioning models for oblique convergence settings suggest that most of crustal deformation is driven by tectonic loading transferred from seismic coupling at the plate interphase (e.g. Saint Blanquat et al., 1998; Fitch, 1972b; Jarrard, 1986; Scheuber & Gonzalez, 1999a). Moreover, from geological and seismological evidence, it has been documented that crustal strain rates are higher within the magmatic arc when compared to other crustal regions (e.g. Barrientos et al., 2004; Reutter et al., 1994; Scheuber & Gonzalez, 1999b; Sielfeld et al., submitted; Weller et al., 2012). Intra-arc upper crustal seismicity sources are either pure tectonic or related to sub- volcanic and volcanic processes (e.g. Zobin, 2003). Tectonic events are characterized by P and S body-wave onsets; the largest reported magnitudes are of ~7.5 (Mw) associated with the activation of long segments of arc-parallel strike-slip faults like in Central America or in Sumatra (White and Harlow, 1993; Reid, 1913). Intra-arc volcanic seismicity, in turn, involves a wide spectrum of low-frequency (long-period) quakes associated with subsurface to surface magmatic processes, i.e. shear and tensile failures within volcanic edifices, fluid pressurization, diking, volume changes, degassing, among others (Zobin, 2003; Chouet, 2003; Manga and Brodsky, 2006). Characteristic time-span for this seismic activity range from T = 1 s long-period (LP), followed by T = 10 s very-long period (VLP) up to 100 ≤ T ≤ 200 s ultra-long period events (ULP), following the standard Caltech’s classification (Chouet, 1996). These types of seismic events are of great interest to understand short-term volcanic behaviour variations and eventually, for eruption forecasting (i.e. thousands of LP, VLP and tremors are daily reported from world-wide 22 monitored active volcanoes). Because volcanic seismicity is a result of thermo-mechanical- magmatic interaction within the crust, it is characterized by earthquakes of small to moderate magnitudes (e.g. Zobin, 2003; Chouet, 2003). For instance, the largest recorded volcanic events reach up to ML ~5.7 (e.g. in 1981 Mt. St. Helens eruptive cycle, and 1991 Mt. Pinatubo eruption; Mori et al., 1991). Additionally, strong to mega tectonic earthquakes cause small static stress changes in the crust, which may perturb and potentially trigger eruptions at primed volcanoes (e.g. Mt Fuji’s 1707 eruption, linked to Mw 8.7 Hoei earthquake related to the Nankai megathrust; Chelsey et al., 2012; Cordón Caulle 1960 eruption related to Mw 9.5 Valdivia earthquake; Lara et al., 2004). The latter involves eruptive centres acquiring conditions pre-emptive of an eruption, such small static stress changes (e.g. viscoelastic relaxation) may enhance magma overpressure and stimulate magmatic advection, large-scale overturn of magma within reservoirs, bubble growth and/or mix-up among magmatic reservoirs (Chouet, 1996; Chouet, 2003; Manga and Brodsky, 2006). What do we know about in the Southern Volcanic Zone? First-order physical controls of the crustal seismogenic behaviour at a given continental margin may be related to across-strike different thickness of the schizosphere (i.e. depth variability of the brittle-plastic transition) arising mostly from a varying geothermal gradient. The upper limit of the brittle-plastic transition can be constrained to the ca. 320˚C isotherm by integrating experimental observations (Stesky, 1975), thermal models (e.g. Springer, 1999), local seismicity (e.g. Salazar et al., 2017; Scholz, 2002; Sielfeld et al., submitted) and direct measurements on geothermal wells (e.g. Suzuki et al., 2014). A schematic representation of main mechanical properties of southern Andean intra-arc crust is presented in the synoptic model of Fig. 1-2. 23 Mechanically, a brittle material is easier to break when thinner, for instance a piece of glass. Similarly, the brittle crust is likely easier to fail when thinner, but harder to break when thicker. As the temperature is one critical parameter governing the depth of the brittle/plastic transition, a comparatively thinner schizosphere is expected along active magmatic arcs. Thus, major intra-arc faults might be active and accommodate substantial deformation in bulk transpressional settings by seismic slip. Intra-arc exhumed faults in Southern Volcanic Zone have been largely documented including the Liquiñe-Ofqui Fault System (LOFS), Andean Transverse Faults (ATF) and margin-parallel compressional faults (e.g. Cembrano et al., 1996; Perez-Flores et al., 2016; Piquer et al., 2016; Sagripanti et al., 2015). In spite of particular evidence of ductile strain along the LOFS (Cembrano et al., 2002), Late Cenozoic exhumed faults are typically defined from fault-slip data analyses of striated fault planes (i.e. frictional deformation), which are spatially and kinematically coherent with major morphostructural lineaments (e.g. Cembrano et al., 1996). Perhaps the most uncertain features of intra-arc faults in southern and south-central Andes are the precise age of deformation and the magnitudes of fault displacements, although the intensity and distribution of rock deformation is fairly well constrained (e.g. Cembrano et al., 2005; Lavenu & Cembrano, 1999; Perez-Flores et al., 2016; Piquer, et al., 2016; Rosenau et al., 2006). For a given idealized homogeneous media, heat, strain rate and fluid pressure are the three principal parameters controlling brittle deformation to occur (Scholz, 2002; Sibson, 1989). Thus, I hypothesize that for regional scales in continental margins, the geothermal gradient is expected to play a first-order control on the variability of brittle/seismogenic crustal thickness. Therefore, regions characterized by a relative higher rate of volcanic output, such as the Transitional Southern Volcanic Zone (TSVZ) in the Southern Andes volcanic arc, should produce relative more amount of little crustal-quakes and stick-slip faulting than a slim section of the 24 same magmatic arc, such as the Northern and Southern-Southern Volcanic Zones (NSVZ & SSVZ), where eventually low rate but higher magnitude events would be expected. Such hypothesis implies that seismic hazard would be smaller in regions with higher rate but low magnitude seismicity. In order to constrain geofluid pathways and effects in the volcanic morphology, my second hypothesis is that stratovolcanoes emplaced on top of inherited basement faults parallel or close-to parallel to the shortening axis in regional transpressional settings may achieve an elongated anatomy driven by parallel dike propagation. Feeder dike predominant propagation and eruption direction may become organized along either (a) trends parallel or close-to-parallel to the maximum horizontal shortening at distal zones of the main volcano edifice (Andersonian dikes), and/or (b) dilate and propagate through favourable oriented long-lived basement faults beneath the volcano-edifice (non-Andersonian dikes). Predominant orientation may prevail in time enhanced by elongated volcano construction and normal-to-the-volcano elongation de- confinement. Finally, it is here hypothesized that thermally enhanced long-term deformation within the magmatic arcs, might result in complex structural arrays promoting the development of local transtensional domains where the loci for volcanism and geothermal activity might take place. Such transtentional damage zones would improve fault and fracture connectivity favouring effectiveness of the permeability structure. 1.3 Research questions, specific objectives and methodologies The main goal of this thesis is to better constrain and understand the mechanical interaction between crustal faulting and geofluid circulation, by using the Southern Andean 25 Volcanic Zone as natural laboratory. Specific key questions arise, such as (1) if there is any seismic activity along margin-parallel and Andean Transverse faults, which is the nature of the seismic events and how do they distribute in space and time? (2) Which is the nature and significance of tectonically controlled dikes and eruptive centres emplaced within upper crustal levels and the surface, respectively? (3) Which is the tectonic significance of complex arrays of interacting crustal faults within the magmatic arc and how does this physical condition relate to volcanism and geothermal activity? This work combines field and crustal seismicity data in order to accomplish three specific objectives: 1) instantaneous brittle behaviour, i.e. natural seismicity; 2) state of stress and deformation at the time of volcanism; and 3) deformation heritage and fault-fracture architecture enhancing fluid flow for volcanism and hydrothermal activity. Specific objectives and associated methodologies are briefly described one-by-one in the following section, whereas detailed explanation of each methodology can be found in the methodology sections of chapters 2, 3 and 4. 26 Figure 1-2 Synoptic model for the brittle-plastic transition at a crustal shear zone. Adapted from Scholz 1988. This study is focused on the mechanical properties of the upper 12 km of the crust (schizosphere). Specific objective and methodology 1: deciphering the instantaneous tectonic state of tectonically controlled arcs. In the frame of this thesis, one fundamental aspect to understand is the activity of frictional faulting undergoing in the schizosphere. Specifically, underneath an active volcanic front where margin-parallel and margin-oblique shearing has been widely documented. Active faulting can be recognized by the distribution of natural seismicity assuming that fault-slip in brittle environments is manifested as elastic waveforms generated by released energy from stick- slip process (e.g. Byerlee & Brace, 1968; Shearer, 2009). The SAIAS (Southern Andes Intra-Arc Seismicity) network was deployed within a 200x100km2 area where major crustal faults and strato-volcanoes spatially co-exist. These include major and splay faults of the northern termination of the LOFS interacting with fault segments of the Lanalhue and the Biobio-Aluminé ATF (LATF and BBATF in Fig. 1-1). In this region twelve Quaternary stratovolcanoes are present: Callaqui-Copahue, Lonquimay-Tolhuaca, Llaima- 27 Sierra Nevada, Sollipulli, Villarrica-Quetrupillán-Lanin and Mocho-Choshuenco. Tectonic seismic events were depicted by impulsive P- and S-wave onsets on ordinary seismograms. Additional information provided by seismic waveforms are wave amplitudes, which are used for local and moment magnitude calculations and P-wave onset polarities, which in turn, serve for estimate the source radiation pattern from which the mechanism of rupture and kinematics can be obtained (focal mechanism). In summary, event detection and their spatial and temporal distribution across and along the Southern Andean volcanic chain yields to a snapshot for the location of active faulting process within the crust. In turn, focal mechanism calculation identifies the stress field driving the rock rupture process (strain). Therefore, the understanding of the active faulting within intra-arc crust shed lights on the tectonic state that is spatially associated with the volcanic activity. 28 Specific objective and methodology 2: the tectonic significance of intra- volcanic dike and minor eruptive vents: state of stress at the time of volcanism This specific objective points to the understanding of how dikes and minor eruptive vents (MEV) within a stratovolcano can be organized in time and space by the operating tectonic field, inherited structural anisotropies and the localized effect of volcano edifice shape and load. The state of stress can be inferred from the geometry of dikes and morphometry of minor eruptive vents. Intra- volcanic dikes (i.e. dikes emplaced within a stratovolcano) are considered as hydro-fractures (Gudmundsson, 2011a), whose first-order geometry is mechanically controlled by the dominant far- field stress tensor (e.g. Delaney et al., 1986; Nakamura, 1977b). From field observations, such dikes occur at shallow to sub-surface levels, being sometimes in geometric and kinematic coherence with eruptive fissures as demonstrated for the Callaqui Volcano (chapter 3.). Geometric relationships are used for kinematic interpretation of dike emplacement at the Callaqui Volcano, whereas dynamic analysis of intra-volcanic and basement dikes were separately carried out for the Tatara-San Pedro Volcanic Complex. Morphometric analysis was carried out for minor eruptive vents at the Callaqui Volcano. Eruptive vents are assumed to be fed by subsurface dikes underneath the emission vent, whose geometry might control the shape of the forming vent or group of aligned vents (Corazzato & Tibaldi, 2006; A. Tibaldi, 1995). Thus, vent morphometry, as well as dike geometry may be used as strain indicators. In strike slip domains, for instance, the maximum shortening might be depicted by the strike of dikes or aligned vents. For reduced geological time spans (e.g. <0.5 Ma) no significant block rotations are expected, therefore stress and strain axes can safely be 29 assumed to be approximately coaxial and reciprocal (e.g. Weijermars, 1991). Thus the orientation of the stress field can be estimated from the geometry of such volcanic objects. Specific objective and methodology 3: estimation of paleostress conditions of volcanic basement architecture This aim is focused in constraining the overall structural geometry and kinematics enhancing quaternary geofluids pathways (magmatic and thermal), as a frame for ongoing geohazards and geothermal research in the Transitional Southern Volcanic Zone (TSVZ). Furthermore, this work not only evaluates the stress condition for a discrete volcanic complex, but also seeks to provide key elements to better understand and decipher the role of Andean- transverse faults in the spatial organization of arc volcanism and geothermal systems. The study of the structural grain of volcanic basements is a key element to understand the permeability architecture for geofluid flow and emplacement. Fault and fracture meshes are considered to develop permeable pathways along their damage zones, organized as a consequence of the governing paleostress-tensor orientation. However, a significant portion of the basement architecture is usually covered by younger and non-deformed volcanic rocks. Therefore, an exhaustive structural mapping is needed to identify domains with relative high concentration of fault-slip data and hydro-fracture geometrical data (i.e. shape of veins and dikes). Field mapping of the basement of the Tatara-San Pedro-Pellado volcanic complex (TSPPC) has been carried out focussed on the folding pattern of stratified basement units, fault geometry and kinematics cross-cutting the basement and volcanic units, and the distribution and geometry of basement and intra-volcanic dikes. 30 1.4 Concluding remarks The south-central and southern Andes are a continuous mountain belt in which active tectonics interacts with volcanism and hydrothermal activity (Fig 1-1). From a regional perspective, margin parallel faults dominate the Southern Andes (38˚-46˚S) structural grain, whereas Andean transverse faults (ATF) appear to dominate the tectonic framework in south- central Andes (33˚-38˚S). Seismic faults within these intra-arc segments are depicted by (micro-) seismicity between 12 km depth and the surface. In contrast, low rate seismicity occurring within the forearc and foreland (back-arc) crust, takes place at deeper levels, delineating a convex-shape of the crustal seismogenic layer bottom (SLB). The first-order convexity of the southern Andes SLB might be consequence of a differential geothermal gradient across the entire orogen. The relative higher seismicity-rate (strain-rate) reported along the intra-arc results from a combination of both parameters: higher temperature and thinner brittle/frictional crustal layer. These findings contribute to baseline knowledge for seismic hazard assessment in remote locations of the Andean range. Furthermore, and a central contribution of this thesis, is showing that the nature of modern intra-arc seismic faulting is coherent with the geometry and kinematics of exhumed ‘fossil’ frictional faults. This indicates a relatively constant long-lived tectonic activity in the current volcanic front under an overall similar stress field. Volcanic-basement deformation is characterized by heterogeneous fault slip juxtaposing transtensional and transpressional domains within the intra-arc. Evidence of strain compartmentalization indicates a complex dynamic interplay among nearby tectonic blocks, driving, for instance, the development of transtensional damage zones where the loci of volcanism take place. 31 Among others, the Callaqui and TSPPC are conspicuous examples of structurally controlled Andean-type strato-volcanoes because of their causal relationship between the tectonic setting and the resulting volcanic and intra-volcanic structure. For instance, at the Callaqui Volcano, aligned and elongated eruptive vents are in geometrical coherence with nearby parallel dike swarms. Moreover, the kilometre-scale, wing-crack-like geometry of dikes and eruptive centres, implies a shear component accompanying the dominant extensional regime at the time of volcanism. Such observation indicates the consistency in time of the governing transtensional stress/strain field, yielding to an overall elongated strato-volcano anatomy, subparallel to the far-field principal stress/shortening axis. Particular evidence is reported for the TSPPC basement, were the Tatara transtensional interaction damage zone (TDZ) is described between the Melado Fault (margin-parallel strike slip fault) and the La Plata and Los Cóndores ATF. The TDZ is featured by oblique to normal faults, hosting parallel dike swarms and nesting ~55 km3 of Pleisto-Holocene eruptive products (Singer et al., 1997). The later might result from a strong coupling between long-lived faulting and permeability development for volcanism and hydrothermal activity enhancement. Finally, it is expected that for hot and thin schizosphere portions (~320˚C at ~12 km depth), frictional activity and effective fault/fracture connectivity might enhance rock permeability for geofluid flow. As a consequence, differential pore-fluid pressure and rock- strength weakening would result in a transient feedback loop between rock fracturing and fluid flow. Such findings are of great interest when facing geothermal exploration challenges. Further local studies should be conducted in geothermal fields, by combining natural seismicity with detailed structural mapping. 32 1.5 Ongoing and future projects This thesis and related paper/manuscripts constitute baseline material for further specific researches in the Southern Andes. In particular, the available data set obtained by the natural seismicity experiment, (Southern Andes Intra-Arc Seismicity - SAIAS) contains much data for constraining the geometry and kinematics of certain key individual faults within the northern termination of the LOFS and ATF within the 38˚S and 40˚S. For instance, the activity reported here at the Caburgua Lake (i.e. Caburgua Fault) has significant implications for seismic hazard assessment in the region. Among them, the maximum expected earthquake magnitude could be computed as well as the recurrence rate of relative large quakes. Ongoing transdisciplinary collaboration with other geoscience research teams will be done matching paleoseimology analyses and numerical simulation. Similarly, the results obtained for the TSPPC are accurate for precise numerical modelling using the Boundary Elements Method, in order to constrain slip senses and approximate rates for co-, post- and inter-seismic stages of the earthquake subduction cycle. On the other hand, an essential pending issue is to better constrain the absolute age of faulting, diking and hydrothermal activity in this region. Supplementary geochronological studies should be carried out, which involve major challenges for the development of geochronological technics to obtain precise absolute ages of young rock (i.e. Late-Pleistocene to Holocene). Local seismicity experiments should also be carried out to better image the permeability structure related to geothermal resources. 33 1.6 Publications and conference participations Publications resulting from this research 2016 Sielfeld, G., Cembrano, J., Lara, L.E. Transtension driving volcano-edifice anatomy: Insights from Andean transverse-to-the-orogen tectonic domains, Quaternary International, 438, 33-49 (Q1) doi:10.1016/j.quaint.2016.01.002 2018 Sielfeld, G., Lange, D. and Cembrano, J. Intra-arc crustal seismicity: Seismo-tectonic Implications for Southern Andes Volcanic Zone, Chile. Tectonics. doi: 10.1029/2018TC004985 2019 Sielfeld, G., Brogi, A., Cembrano J., Ruz, J., Stanton-Yonge, A., Pérez-Flores P. and Iturrieta P. Oblique-slip tectonics and geofluid flow: a case study from south-central Andes. Submitted to Tectonophysics (in review) Abstract and conference participations associated with this research 2013 Sielfeld, G. and Cembrano, J. Oblique-to-the-orogen fault systems and it causal relationship with volcanism and geothermal activity in Central Southern Chile: Insights on ENE and NW regional lineaments. Chile. AGU Fall Meeting, San Francisco, USA. 2014 Sielfeld G., Cembrano J., Lara, L.E. 2014. Transtension controlling volcanic morphology: Insights from oblique-to-the-arc tectonic domains AGU Fall Meeting, San Francisco, USA. 34 2016 Sielfeld, G., Lange, D., Cembrano, J. Intra-arc crustal seismicity: evidence of parallel and transverse-to-the-orogen fault systems. 2 Colloquium of Geophysical Signatures of Earthquakes and Volcanoes. Universidad de Chile. Presented as poster. Digital files. 2016 Sielfeld, G., Lange, D., Cembrano, J. Intra-arc crustal seismicity: evidence of parallel and transverse-to-the-orogen fault systems. Cities on Volcanoes 9 (CoV9), International Association of Volcanology and Geochemistry of the Earth’s Interior (IAVCEI). Puerto Varas, Chile. Presented as poster. Digital files. 2017 Sielfeld, G., Lange, D., Cembrano, J. Intra-arc crustal seismicity of Southern Chilean Andes (38˚040˚S). Integrated Plate Boundary Observatory Chile (IPOC) Workshop, Annual Meeting. Geomar, Kiel, Germany. 2017 Sielfeld, G., Lange, D., Cembrano, J. Intra-arc seismicity: geometry and kinematic constraints of active faulting along northern Liquiñe-Ofqui and Andean Transverse fault systems (38˚-40˚S, Southern Andes). AGU Fall Meeting, New Orleans, USA. 2018 Sielfeld, G., Cembrano, J., Lange, D. The Caburgua microseismic sequence: crustal intra- arc faulting in Southern Andes, XI Congreso Geológico Chileno (Concepción). Presented as poster. Digital files. 35 2018 Sielfeld, G., Lange, D., Cembrano, J. Intra-arc Crustal Seismicity: Seismo-tectonic Implications for the Southern Andes Volcanic Zone, Chile, XI Congreso Geológico Chileno (Concepción). Presented as a talk. Digital files. Publications and abstracts resulting from side projects 2014 Zellmer, G.F., Freymuth H., Cembrano, J.M., Clavero J. E., Veloso E.A., Sielfeld, G.G. Altered mineral uptake into fresh arc magmas: insights from U-Th isotopes of samples from Andean volcanoes under differential crustal stress regimes. Geol. Soc. of London. Special Volume: “Orogenic Andesites and Crustal Growth”. doi:10.1144 / SP385.9 2015 Sanchez-Alfaro, P., Sielfeld, G., Van Campen B., Dobson, P., Fuentes, V., Reed A., Palma- Behnke, G., Morata, D. 2015. Geothermal barriers, policies and economics in Chile – Lessons for the Andes. Renewable and Sustainable Energy Reviews. 51, 1390-1401. doi:10.1016/j.rser.2015.07.001 2019 Gonzalez-Vidal, D., Palma, J.L., Lange, D., Sielfeld, G. (2019?). Precursors of Villarrica volcano (Chile) eruption occurred on March 3rd, 2015, from ambient noise. Intended to be submitted to Journal of Volcanology and Geothermal Research 2013 Cembrano, J., Perez-Flores, P., Sanchez, P., Sielfeld, G. Tectonics, magmatism and fluid flow in a transtensional strike slip setting: The northern termination of the Liquiñe-Ofqui fault system, Chile. AGU Fall Meeting, San Francisco, USA. 36 2013 Perez-Flores P, Sanchez P, Sielfeld G., Cembrano j. 2013. Tectonics, magmatism and fluid flow in a transtensional strike-slip setting: The northern termination of the dextral strike-slip Liquiñe-Ofqui fault System, Chile. AGU Meeting of Americas, Mexico. 2015 Lira, E., Sielfeld, G., Bosch, A., Castellón, R., Cumming, D., Figueroa, R., Gottreux, I., Gutierrez, A., Iturrieta, P., Maringue, J., Muñoz, J.J., Pérez, N., Santibañez, I., Yañez., G. Reactivación de Fallas de larga vida transversales al orógeno andino. Perspectivas de la Falla Pichilemu. XIV Congreso Geológico Chileno (Coquimbo). Presented as poster. Digital files. 2018 Gómez, B., Sielfeld, G. Rol de las fallas transversales al arco en la evolución tectónica cenozoicade los Andes centrales: Caracterización de la Falla El Arrayán, Región Metropolitana, Chile. XV Congreso Geológico Chileno Presented as poster. Digital files. 2018 González, D., Lange, D., Sielfeld, G. Time-lapse monitoring of the unrest period of the Villarrica (or Ruka Pillán) volcano, Chile, occurred in 2015 from coda-wave interferometry, XV Congreso Geológico Chileno. 2018 Sanchez de la Muela, A., Hammond, J., Mitchell T., Cembrano, J., Sielfeld, G., Pearce, R., Marshall N. Understanding the tectono-magmatic system of the Transitional Southern Volcanic Zone (TSVZ), Chilean Andes, from seismicity analisys. XV Congreso Geológico Chileno Presented as a talk. 37 2 Intra-arc Crustal Seismicity: Seismo-tectonic Implications for the Southern Andes Volcanic Zone, Chile 2.1 Introduction Crustal faulting is an end-member effect of large scale tectonic loading. At active magmatic arcs in an oblique convergence setting, permanent heat flow enhances margin-parallel rock weakening and consequential transpressional partitioning (Tikoff and Teyssier, 1994; De Saint Blanquat et al., 1998). Upper crust seismic faulting represents a snap-shot of instantaneous strain, regarded as stick-slip frictional instability of pre-existing faults (Scholz, 1998). This mechanism refers to sudden “slippage” (earthquake) after an interseismic period of elastic strain accumulation or “stick” (Brace and Byerlee, 1966; Byerlee and Brace, 1968), commonly occurring below the critical temperature of Quartz plasticity (~300˚C; Stesky, 1975). Furthermore, crustal earthquakes within magmatic arcs are either purely tectonic, spatially associated with major intra-arc strike-slip faults (e.g. Weller et al., 2012; White, 1991; White and Harlow, 1993; La Femina et al., 2002) or have a volcanic component, associated with magma flux, degassing and volcanic processes (e.g. Manga and Brodsky, 2006; Jay et al., 2011; Patanè et al., 2011; Mora-Stock et al, 2012). Therefore, intra-arc seismicity may reflect the geometry and kinematics of faults and delimit the crustal seismogenic zone, contributing to the understanding of the tectonic and thermo-mechanical state of orogenic belts with active magmatism. In most subduction settings, long-term oblique convergence slip vectors are partitioned into margin-parallel and margin-orthogonal components (e.g. McCaffrey, 1992). A significant portion of bulk transpressional deformation is commonly accommodated in intra-arc regions, as 38 margin-parallel shear zones (e.g. Fitch, 1972; Beck, 1983; Kimura, 1986; Jarrard, 1986; Beck, 1991; McCaffrey, 1996). In the upper fractured intra-arc crust, deformation is accommodated as strike- slip fault systems for about 50% of active oblique convergence settings (e.g. Jarrard, 1986; White, 1991; Bommer et al., 2002). Resulting high strain domains are time-dependent of the interplay among inherited geological anisotropies, fault growth and frictional wear of country rock (Scholz, 1987; Cowie and Scholz, 1992; Joussineau and Aydin, 2009; Mitchell and Faulkner, 2009). Simultaneously, crustal fluid ascent, differentiation, emplacement and crystallization mechanisms are largely controlled by the geometry, permeability and kinematics of high strain domains (Clemens and Petford, 1999; Cox, 1999; Rowland and Sibson, 2004; Cembrano and Lara, 2009; Acocella and Funiciello, 2010; Yamaguchi, et al., 2011), such as demonstrated for the Sunda, Central-American and Andean magmatic arcs (Sieh and Natawidjaja, 2000; La Femina et al., 2002; Corti et al., 2005; Sielfeld et al., 2017; Iturrieta et al., 2017). Driven by the crustal stress-field and thermal state, fluid migration within such regions (e.g. Nakamura, 1977; Delaney et al.,1986) may be enhanced by rock-dilatancy/fluid-transport mechanisms triggered by shallow seismicity (e.g. Sibson et al., 1975; Scholz, 2002; Rowland and Sibson, 2004). In the Southern Andes Volcanic Zone (SVZ), in particular along the margin-parallel Liquiñe- Ofqui Fault System (LOFS), stratovolcanoes, minor eruptive centers and geothermal features, are spatially related to the geometry and kinematics of long-lived faults (Cembrano and Lara, 2009; Sánchez-Alfaro et al., 2013). Additionally, for the long-term geological record, margin-parallel faults (i.e. LOFS) coexist with second order transverse-to-the-arc oriented faults (Andean Transverse Faults, ATF; Kaizuka, 1968; Katz; 1971; Yañez et al., 1998; Melnick et al., 2009; Aron et al., 2014; among others), developing a causal relationship with crustal fluid-flow (Cembrano and Lara, 2009; Sánchez-Alfaro et al., 2013; Tardani et al., 2016; Wrage, et al., 2017; Pérez-Flores et al., 39 2016; Pérez-Flores et al, 2017). Based on kinematic and dynamic analysis of fault-slip data, Pérez- Flores et al. (2016) suggest that bulk transpressional deformation is partially partitioned into variable orientations and scales within the Southern Volcanic Zone (SVZ) of the Andean magmatic arc. Although the present-day instantaneous crustal stress state and slip partitioning effect in the SVZ remains obscure, no dense, regional-scale seismic study has been carried out in the SVZ; consequently, only local deployments associated with specific, moderate to large magnitude events, focussed on the forearc region, have been undertaken in the LOFS. (e.g. Lange et al., 2008; Legrand et al. 2010; Dzierma et al., 2012a,b). In this work, we seek to understand the interaction between crustal faulting, strain partitioning and compartmentalization in the upper crust of the intra-arc region of Southern Andes. This will shed light on how a fraction of plate boundary deformation is accommodated in the intra-arc region, during one interseismic-stage of the subduction earthquake cycle. Therefore, we installed a dense seismic network of 34 stations between Callaqui (38˚S) and Mocho-Choshuenco (~40˚S) volcanoes in southern Chile in order to obtain a high-resolution seismicity catalog occurring within an observation time-window of ~16 months (i.e. between March 6th 2014 and June 16th 2015). The study region lies in a transitional zone influenced by two overlapping tectonic loading conditions, associated with two, apparently independent, megathrust seismic cycles. As stated by Sánchez and Drewes (2016), regarding the 2010 Maule Mw 8.8 earthquake post-seismic effects, the region encompassed between 37˚S and 40˚S would have undergone compressional surface deformation from March 2010 to April 2015, with the maximum horizontal stress trending N30˚E (period of GNSS measurements). Conversely, the region south of 38°S has been undergoing interseismic re-loading related to the locking of the 1960 Valdivia Mw 9.5 earthquake rupture zone 40 (Moreno et al., 2009; Moreno et al., 2011). The main target of this experiment is to resolve, for a brief tectonic-time window, the nature and interaction of seismic faulting within (1) long-lived margin-parallel strike-slip fault systems (i.e Liquiñe-Ofqui Fault System, LOFS), and (2) long-lived orogeny -oblique faults (Andean-transverse-fault, ATF). Our results contribute to understand the overall link among intra-arc strain compartmentalization, long-lived fault geometries and seismically-active fault kinematics in the framework of an active (Quaternary) volcanic arc. 2.2 Tectonic setting of Southern Andes The Southern Andes are characterized by long-term partial partitioning of an overall transpressional deformation regime arising during interseismic phases of the subduction earthquake cycle (e.g. Cembrano et al., 1996; Arancibia et al., 1999; Lavenu and Cembrano, 1999; Pérez-Flores et al., 2016). Dextral-oblique convergence between the Nazca and South American plates trends N80˚, with an average velocity of ca. 66 mm/yr., and mean obliquity of 20° (current convergence vector estimated by Angermann et al., 1999; Fig. 2-1). The convergence vector direction has been relatively constant over the last 10 Ma (Pardo-Casas and Molnar, 1987). Late Miocene- Pliocene transpressional deformation is well documented by field structural geology combined with geochronological data (Arancibia et al., 1999; Lavenu and Cembrano, 1999; Cembrano et al., 2000; Cembrano et al., 2002; Rosenau et al., 2006; Pérez-Flores et al., 2016), as well as by geodetic and seismological studies (e.g. Wang et al., 2007; Moreno et al., 2008, Lange et al., 2008; Haberland et al., 2009; Legrand et al., 2010; Agurto, et al. 2012). The study area has been affected by two interplate megathrust earthquakes in the last 60 years. The 22 May 1960 Valdivia Mw 9.5 earthquake ruptured > 900 km of the Nazca – South American plates interphase, from Arauco Peninsula (37.5˚) to the Taitao (46.5˚S; Moreno et al., 41 2009). It is the largest instrumentally recorded seismic event (Engdahl and Villaseñor, 2002). The epicenter took place beneath the Arauco Peninsula, spatially related with forearc strike slip crustal faulting (Lanalhue fault System, LFZ) that has been documented in the past decade (Moreno et al., 2009; Haberland et al., 2006; Fig. 2-1). In turn, the 2010 Maule Mw 8.8 earthquake broke a ~500 km-long portion of the subduction interphase, extending from 38.2˚S to 34˚S (Moreno et al., 2012). The epicenter locates tens of kilometers north from the Arauco Peninsula (Fig. 2-1), however the coseismic slip is distributed in two main patches, reaching a maximum slip in the northern part of 15.7 m (from GPS measurements; Moreno et al., 2012). Both epicenters locate ~200 km and ~300 km to the WNW and NW of the center of our study area, respectively. Liquiñe-Ofqui Fault System (LOFS) and Andean-transverse Faults (ATF) The LOFS is a 1200 km-long intra-arc dextral strike-slip structure, characterized by master NNE-striking dextral strike-slip fault segments and subsidiary synthetic NE- to ENE-striking transtensional faults (Cembrano and Hervé, 1993; Cembrano et al., 1996, 2000; Melnick et al., 2006a,b). Regional to local-scale transtensional domains, such as duplexes, releasing bends and horsetail geometries have been described (Potent, 2003; Rosenau et al., 2006; Pérez-Flores et al., 2016; Fig. 2-1) and evidenced by local transtensional seismic faulting (Legrand et al., 2010). In spatial association with these domains, sub-parallel diking, ENE-striking volcanism and hydrothermal activity is a common feature (Melnick et al., 2006a; Rosenau et al., 2006; Lara et al., 2008; Cembrano and Lara, 2009; Sielfeld et al., 2017). Furthermore, tensional stress magnitudes can be larger where NE-striking faults are close to, or intersect, master fault segments, which enhances long-term crustal fluid pathways during interseismic tectonic loading conditions 42 (Iturrieta et al., 2017). A significant portion of the bulk transpressional deformation is accommodated by slip on the intra-arc Liquiñe–Ofqui Fault System (LOFS) displaying markedly slip-rates differences from north to south along different fault segments (Rosenau et al., 2006; Stanton-Yonge et al., 2016; Iturrieta et al, 2017). Calculated slip magnitudes from numerical models may be larger along pure strike-slip fault segments, ranging from ~18 mm/yr., close to southern LOFS termination (Iturrieta et al., 2017), down to 0.5 mm/yr. in the northern termination (Stanton-Yonge et al., 2016; Fig. 2-1). Thus, as a result of highly coupled plate interfaces, the forearc sliver motion, and thermal weakness at the intra-arc region (Cembrano et al., 2002; Wang et al., 2007), the overriding plate shows a variable degree of deformation partitioning along and across the Southern Andes (Arancibia et al., 1999; Lavenu and Cembrano, 1999; Rosenau et al., 2006; Pérez-Flores et al., 2016). The LOFS records long-term right-lateral ductile deformation between 6 and 3 Ma (Cembrano et al., 2000) that is overprinted by dextral brittle deformation since 1.6 Ma (Lavenu and Cembrano, 1999). Present-day activity of this fault system is consistent with northward motion (6.5 mm/yr.) of the forearc sliver (Wang et al., 2007) and available crustal seismic faulting (Fig. 2-1). A compilation of focal mechanisms between -33 and -44˚S shown in Fig. 2-1, clearly document dextral strike-slip partitioning accommodated in the intra-arc (Chinn and Isacks, 1983; Barrientos and Acevedo-Aranguiz, 1992; Lange et al., 2008; gCMT, http://www.gcmt.org). Andean transverse faults (ATF) are WNW-striking, long-lived structures recognized as conspicuous morphotectonic lineaments and fault zones striking obliquely with respect to the overall Andean orogeny strike (~N10˚E). ATFs are suggested to be older or coeval to the LOFS (e.g. Pankhrust, et al., 2006; Rivera and Cembrano, 2000; Moreno et al., 2011), and have been suggested to play a first-order role in SW Gondwana Carboniferous-Permian basin evolution and 43 the Carboniferous accretion of the Patagonian terrain (Breitkreuz, 1991; Tankard, et al., 1995, Rapela et al., 2003; Pankhrust et al., 2006). ATFs also have been thought to control the Neogene forearc basin architecture and megathrust rupture segmentation associated with the seismic subduction cycle (Glodny et al., 2008; Melnick et al., 2009). Furthermore, deformation of the ATFs might be activated from megathrust earthquakes (Arriagada et al., 2011; Farías et al., 2011; Aron et al., 2014; Stanton-Yonge et al., 2016), and may control crustal fluid migration within the intra- arc region (Katz, 1971; Rivera and Cembrano, 2000; Cembrano and Lara, 2009; Sánchez et al., 2013; Sielfeld, et al., 2017; Pérez-Flores et al., 2016; Roquer et al., 2016; Wrage et al., 2017). 44 45 Fig. 2-1: Seismotectonic map of the southern Andes. Convergence rate and direction is shown by the solid white vector (Angermann et al., 1999). Approximate coseismic rupture zones (slip>1m) of the 1960 Valdivia Mw 9.5 (epicenter in red star) and 2010 Maule (moment tensor in red-white) megathrust earthquakes are outlined by cyan and yellow dashed lines, respectively (Moreno et al., 2009; Moreno et al., 2012). Moment tensors of the gCMT catalog (www.globalcmt.org, 1976-2017) include: i) 13 subduction-type earthquakes (blue-white beach-balls Mw > 6.5) ratifying the average slip accommodated in the subduction interphase; ii) outer-rise seismicity Mw > 6.5 (beige-white beach-balls); iii) crustal events Mw ≥ 5 and depth < 20 km in green-white beach-balls. Crustal focal mechanisms from local deployments are shown in black-white (Haberland et al., 2006) and light blue-white (Lange et al., 2008). Focal mechanisms are scaled proportional to their moment magnitude (factor 0.4). Location of geothermal springs and stratovolcanoes taken from Hauser, 1997, Risacher and Hauser, 2008 and Siebert et al., 2010. Crustal faults systems (LOFS, ATF) and morphotectonic lineaments from Melnick and Echtler (2006), Cembrano and Lara, 2009, Lira et al., 2015; Pérez-Flores et al., 2016. Margin orthogonal and margin-parallel components are schematized in dashed arrows. Average slip vector of subduction thrust earthquakes are indicated in blue. The expected residual margin- parallel slip component is shown by green vectors (Stanton-Yonge et al., 2016). The rectangle in white lines represent the area shown in Figures 2-6 in this chapter and 6-1 in Supplementary Material (S.M.). Shaded relief image generated with 1 arc- second (~30 m) resolution data from the Shuttle Radar Topography Mission. Nature and historical record of intra-arc seismicity In general, intra-arc upper crustal seismicity sources are either pure tectonic or related to sub-volcanic and volcanic processes (Zobin, 2003). Tectonic wave forms, in which this paper is focused, are characterized by P and S body-wave onsets; the largest reported magnitudes are of ~7.5 (Mw) associated with the activation of long segments of arc-parallel strike-slip faults such as in Central America or in Sumatra (Reid, 1913; White and Harlow, 1993). Intra-arc volcanic seismicity, in turn, involves a wide spectra of low-frequency (long-period) quakes associated with subsurface to surface magmatic processes, i.e. shear and tensile failures within volcanic edifices, fluid pressurization, diking, volume changes, degassing, among others (Zobin, 2003; Chouet, 2003; Manga and Brodsky, 2006). For instance, the biggest recorded volcanic events are ~5.7 Ml (e.g. in 1981 Mt. St. Helens eruptive cycle, and 1991 Mt. Pinatubo eruption; Mori et al., 1991). Nonetheless, strong to mega tectonic earthquakes cause small static stress changes in the crust, which may perturb and potentially trigger eruptions at primed volcanoes (e.g. Mt Fuji’s 1707 eruption, linked to M 8.7 Hoei earthquake at the Nankai megathrust; Chelsey et al., 2012). Few and sparse local seismic networks in Southern Andes have reported shallow crustal seismicity (depth <20 km) along several domains of the Southern Volcanic Zone (SVZ) (e.g. 46 Barrientos and Acevedo-Aránguiz, 1992; Farías et al, 2010; Mora-Stock et al, 2012; Cardona et al., 2015) and the LOFS (Bohm, et al., 2002; Lange et al., 2008; Legrand et al., 2010; Agurto et al., 2012). Several of the observed events, are of magnitude 5 or greater, including two Mw 6.1 and Mw 6.2 Aysen Fjord earthquakes in April 2007, associated with a strike-slip duplex at the southern termination of the LOFS (e.g. Legrand, et al., 2010; Mora-Stock et al., 2010; Agurto et al., 2012); the strike-slip El Melado 6.0 (Mw ) earthquake in the Maule region in June 2012 (indicated in Fig. 2-1), spatially associated with the NS trending El Melado River Lineament (Cardona et al., 2015); and the Mw 6.5 Teno earthquake in August 2004, associated with dextral strike-slip on El Fierro Fault (gCMT; Farías et al., 2010; Fig. 2-1). At Lonquimay volcano (38˚22’S), one 5.3 (Mw) earthquake on February 24th (1989) was recorded at 15 km depth two months after a flank cone eruption, (Strombolian “Navidad crater” eruption on December 25th, 1988; Barrientos and Acevedo- Aránguiz, 1992; Fig. 2-1). Volcano-tectonic events distribution and moment tensor of the Mw 5.3 quake (gCMT; Dziewonski et al., 1990) indicate right lateral strike-slip faulting along the NNE- striking Lolco Fault of the LOFS (Barrientos and Acevedo-Aránguiz, 1992). In contrast, sparse and small magnitude crustal earthquakes have been recorded along the ATFs. For instance, left-lateral strike-slip to compressive focal mechanisms (Haberland et al., 2006) were reported for the NW-striking Lanalhue Fault Zone in the forearc region of southcentral Chile (LFZ; Fig. 2-1) (Melnick and Echtler, 2006, Haberland et al., 2006). Further, south in the main Andes (41.5˚S - 43.5˚S), Lange et al. (2008) observed seismic clusters associated with secondary NW left-lateral strike slip faulting cross-cutting the LOFS close to strato-volcanoes, with focal mechanisms in line with the long-term stress fields estimated from faults-slip data inversions along the volcanic arc (Lavenu and Cembrano, 1999; Iturrieta et al., 2017). 47 Similarly, at the northern termination of the LOFS, the strike-slip Callaqui-Alto Biobio earthquake (Mw = 5.5) occurred on December 2006, close to a fault intersection. Tele-seismically detected (Fig. 2-1, NEIC - gCMT), the epicenter is located in between the NW- striking Biobio- Aluminé Fault (BBAF) and the NNE-striking Lolco-Lomin Fault segment of the LOFS (Fig. 2-1). Hypocenter uncertainty does not allow to correlate this event to either the LOFS or the BBAF. More importantly, all above mentioned events are compatible with overall ENE-trending shortening, and NNW-trending extension axes. 2.3 Data Processing and Methods We deployed a passive seismic network of 34 stations between 2014 March and 2015 June (~16 months) (Fig. 2-3 and Fig. 6-1). The local network consisted of 8 broad-band (Trilium 120) and 25 Mark L4-3D 1 Hz three-component sensors registering ground motion in continuous mode at a 100 Hz sample rate. All instruments were equipped with EarthDataLoggers (section 6.1.2, Figs. 6-2 and 6-3 in S.M.). Horizontal separation among the closest stations varies from 6 to 18 km, covering an area of ca. 200 km by 100 km along the magmatic arc (38˚- 40˚S; Fig. 3.b, Table 6-1 and Fig. 6-1 in S.M). Homogenous and dense station-separation allow detection localization of micro-seismicity for the entire zone. In order to assure constant operation in the high mountainous region, including through wintertime, stations were equipped with 90 W solar panels and 65 Ah deep-cycle battery (Text S1, Figs. S2 and S3). Most stations ran continuously, other than 2 high altitude stations (>2000 m a.s.l) that shut down for approximately 3 weeks after heavy snowfall. To further densify the network, we incorporated continuous vertical seismograms of sixteen permanent broadband stations of the Andes Volcanological Observatory (OVDAS), installed on main stratovolcanoes (Fig. 6-1 and Table 6-1). 48 Crustal seismicity level in the region is very low; in fact, no event at all has been listed in the CSN (Centro Sismológico Nacional de Chile) catalog for that period. Therefore, we manually inspected the vertical records of the complete dataset (16 months) using overlapping windows of nine minutes with the Passcal Quick Look software (PQL). Then, P- and S- waves onsets of local events with S-P-phase time differences of less than 10 s were manually picked. Overall, the detection and localisation were optimized on the creation of a seismicity catalog down to the smallest possible magnitude events. On average S-P-phase time difference smaller than 3 s is characteristic for shallow quakes occurring within the study area (see waveform example in Fig. 2-2). A small number of low and very low period events were detected, but not located; these are then not considered for further analysis. For a total of 652 detected events, P and S onsets were carefully identified using the software package SEISAN (Ottemöller et al., 2014). For each arrival we manually picked the first possible, the best and last possible arrival times. The error in P and S determination was then estimated by the differences between the first and last possible arrival time and was then converted into weights. Furthermore, we picked P-phase polarities for the stronger events. After phase picking, we used the HYP program, SEISAN modified version of HYPOCENTER location program (Lienert, et al., 1986; Lienert and Havskov, 1995), as a first and linear hypocenter determination approach. The final catalog was selected upon the basis of epicentral distance smaller than 100 km from the network border and hypocentral misfits (RMS, root mean square) smaller then 1 s, from which the 96.9% has RMS smaller than 0.5 [s]. The final catalog consists of 356 crustal events with 2,716 P- and 2,323 S-phase arrivals. The residual events, not considered here, belong mostly to interplate seismicity. 49 Figure 2-2 Raw waveform example of a typical crustal event (8.8 km depth, Mw=1.4 on 18.06.2014, 20:04 UTC) recorded by station LS2S (Fig. 2-3) at an 11.7 km epicentral distance. Vertical (Z) and horizontal components (NS and EW) are showing impulsive P and S phase arrivals. A S-P time difference of 1.78 [s] indicates a short distance to the source (14.5 km with our velocity model). Construction and validation of a minimum one-dimensional velocity model Travel times were used to compute 1-D layered velocity models using VELEST program (Kissling et al., 1995). VELEST inverts 1-D velocity models from phase arrival times, using an initial start-up model (Kissling, 1988; Kissling et al., 1995). Generated results are 1-D Vp and Vs velocity models, station corrections (all delays are < |0.5| s), origin-time and hypocentral coordinates (to,x,y,z). The following criteria were used for selecting 95 well-constrained earthquakes as inversion input: i.) minimum of 8 P- and 4 S-phase arrivals; ii.) observations within the network (greatest azimuthal gap (GAP)≤180˚); iii.) event residuals (RMS) smaller than 0.3 s. 50 We implemented a brute-force search with a large number of plausible Vp input models (124.000) and a constant Vp/Vs ratio of 1.78, in order to find the minimum 1-D velocity model with the smallest RMS (e.g. Lange et al., 2007, 2012). Then, we followed a staggered approach (Husen et al., 1999) using the best Vp model (i.e. with the smallest RMS) and 60 initial Vp/Vs ratios (inverting each 0.1 steps in the range 1.5 < Vp/Vs < 2.1) to determine a minimum 1-D Vs model. The best 1,000 P-, and all (60) S-velocity models are plotted in Fig. 2-3, and compared to the minimum 1D Vp and Vs known for the entire forearc (Haberland et al., 2006). The minimum 1D Vp and Vp/Vs models are used to determine the final probabilistic hypocenter origin time, coordinates and uncertainty (Lomax, 2000; http://alomax.free.fr/nlloc/). Figure 2-3 a) Computed 1-D Vp and Vs velocity models. Thick lines represent the minimum misfit 1D Vp and Vs models. The input range of the Vp velocity models is indicated by grey dashed lines. Dashed-green lines indicate the minimum Vp and Vs velocity model obtained for the forearc in this latitudinal range (Haberland et al., 2006). The right panel shows the Vp/Vs ratio. b) Stations distribution and P- and S-wave station corrections corresponding to the minimum 1-D Vp and Vs models. Station names are labelled. Reference station (Vp delay = 0 s) is indicated with a black star (LS4S). Stratovolcanoes are indicated with pink triangles. Villarrica volcano (Vi) is shown as reference for Fig. 2-5. 51 Hypocentre Determination and Moment Magnitudes For the final hypocenter determination, we re-localized the seismicity with the NonLinLoc program (Lomax et al., 2000). The program locates seismicity using a non-linear Oct-tree grid- search approach, based on a probabilistic density function (PDF) for the inverse problem of spatial (x,y,z) and origin-time (t) hypocenter localization (Tarantola and Valette, 1982; Moser et al., 1992; Wittlinger et al., 1993; for a complete description of the method refer to 6.1.3 in S.M). The technic provides comprehensive uncertainty and resolution information. The errors in the observations (phase time picks) and in the forward problem (travel-time calculation) are assumed to be Gaussian. This assumption allows the direct analytic calculation of a maximum likelihood origin-time given the observed arrival times and the calculated travel times between the observing stations and a point in the xyz space. Thus the 4D problem of hypocenter location reduces to a 3D search over x,y,z space (Lomax et al., 2000). The Gaussian estimators and resulting confidence ellipsoid provide location uncertainties, in case the complete non-linear PDF has a single maximum and has an ellipsoidal form. The maximum likelihood location is chosen as the preferred location. For each event, we estimated the three-dimensional spatial confidence (68%) from 3D ellipsoids fitted to 3D PDF scatter clouds. Seismic moment (M0) as well as moment magnitude (Mw) were estimated for all events with more than 4 amplitude readings (265 events). The SEISAN AUTOMAG routine (Ottemöller and Havskov 2003; Ottemöller et al., 2014) use spectral analysis for making the attenuation and instrument corrected displacement spectrum. This analysis determines the flat spectral level and corner frequency for each record, from which the events seismic moment is calculated. 52 Focal Mechanisms Focal mechanisms yield information about the faulting style. The initial motion of the P- wave depends on whether the ray left the source from a compressional (i.e. upward first motion at the receiver), or dilatational quadrant (downward first motion at the receiver), easier observed and picked in the vertical component of the receiver. Local events with high quality record, located inside the network (GAP≤180˚) and with at least 8 P-, and 4 S-wave onsets, and a maximum location misfit (RMS) of 0.3 s were considered. These events had depths ranging from 4 to 15 km, and, moment magnitudes between 1.5 and 3.4 (Table 6-2). Selected events were first analysed with FPFIT program (Reasenberg & Oppenheimer, 1985) which use take-off angles (obtained from maximum likelihood hypocentres) to estimate the azimuth and take-off angles at the hypocentre, and determines, from the P-wave first arrival polarities, the double coupled force equilibria of the source. In addition to P-wave polarities, S- to P-amplitude ratios provide constraints to the inversion of the source radiation pattern (Kisslinger, 1980). Vertical maximum P- and S-wave (Pv and Sv) amplitudes were picked in the vertical channel. For each station, horizontal S-wave maximum amplitude (Sh) were also picked in the synthetic rotated transverse- to-the-ray-path channel. Take-off angles, P-wave polarities and the Sv/Pv, Sv/Sh and Sh/Pv amplitude ratios were used to compute double couple focal mechanisms using the FOCMEC program (Snoke et al., 1984). Finally, we compared solutions obtained from FPFIT and FOCMEC. We only accepted focal mechanism obtained with the FOCMEC approach, which do not differ significantly from the FPFIT solutions (less than 20˚ of difference in primary and auxiliary fault planes). A total of 36 events remained after the selection criteria. 53 2.4 Results Nature, Magnitudes and Temporal Distribution of Intra-arc Seismicity Because of the dense station spacing and long-time deployment, we monitored the volcanic arc crustal (micro-)seismicity on a regional scale with very high resolution, for both detection and hypocentral coordinates. Detection and location quality of the 356 crustal events is homogeneous throughout the study area with a magnitude completeness (Mc) of Mw 1.4. Events range from Mw 0.6 to Mw 3.6. Fig. 2-4 shows our data fitting the Gutenberg-Richter (G-T) relation for magnitudes equal or bigger than Mc. The estimated b-value for our experiment is of 1.11 with a maximum magnitude of 3.6 (Mw) in the forearc at 39 km of depth and 3.4 (Mw) in the intra-arc at 9.4 km of depth. Figure 2-4 Magnitude-frequency histogram for the 265 events with calculated magnitudes of our catalog. Columns indicate number of events for each 0.2 units-of-magnitude bin. Crosses represent cumulative number of events per bin. Data fits the Gutenberg-Richter distribution up to the magnitude of completeness (Mco), with b-values close to 1.1. The shallow intra-arc 1-D velocity structure (<9 km) is characterized by relatively low P- wave velocities (4.9 km/s in average) compared to Vp estimations for the whole forearc from 54 Haberland et al., 2006 (Fig. 2-3). However below 9 km depth, P-wave velocity increases rapidly to 6.2 km/s, demarking a high velocity transition depicted by a seismogenic zone (82% of localized seismicity lies between -1 – 14 km). From that depth down to 35 km, a steadily velocity increment up to 7.3 km/s is estimated. At 40 km depth Vp reach 8.3 km/s and Vp/Vs drops from 1.82 to 1.59, demarking another first-order velocity discontinuity. Deeper than 40 km velocities are poorly resolved because of limited ray coverage. Altogether, our ~16-month-long (467 days) experiment recorded 356 crustal events covering an earth surface area of 14,071 km2. The SAIAS catalog (Southern Andes Intra-Arc Seismicity) is listed in Data Set S1 (in S.M.). Activity rate is of 0.762 events per day with an average magnitude of 1.25 (Mw), as is shown in Fig. 2-5a and 2-5b. Background level seismicity is characterized by scattered individual events or minor sequences (< 4 events per day) commonly organized in clusters or alignments. Temporal distribution of seismicity (Fig. 2-5a) depicts five Andean-transverse seismic sequences with more than 5 events per day, which are described later on this section. A total of 293 shallow crustal events are located within the intra-arc region, completing the 82.3% of registered quakes. Our results yield to a total seismic moment of 9.7*1014 [Nm] for the 16 months of continuous measurement (Fig. 2-5.c). This is equivalent to a daily seismic moment of 1.476*108 [Nm/km2] which equals a daily moment magnitude of Mw - 0.62 units per square kilometer. Spatial Distribution and Kinematics of Intra-Arc Seismicity Most reported intra-arc events (78.8%) are spatially associated with morphostructural alignments, surface mapped faults and/or ellipsoidal clusters (Fig. 2-6.a and Video 6-1, link in S.M.). As a first-order observation, bulk distribution of seismicity follows the overall N10˚E strike 55 of the LOFS. Detailed observations reveal second-order seismicity clusters involving NE- and NW striking seismic sequences (Fig. 2-6.a, 2-6.b and 2-6.c). The latter, which are hitherto unreported, reveal active deformation in unexpected, and some of them, populated areas. The remaining 21.2% localizes at shallow depth, spatially associated with stratovolcanoes and/or to geothermal features (e.g. 62 events occurred within Villarrica volcanic edifice). Figure 2-5 Temporal properties of the seismicity catalog. a) Number of events per day. b) Total daily moment magnitudes (Mw); Events without magnitude are plotted in grey at 0.3 units. The experiment average Mw is indicated by a dashed blue line. c) Cumulative 56 seismic moment. Green line represents the eruption of Villarrica on 3rd of March 2015. Seismic sequences above the background natural seismicity level are highlighted with grey dashed lines and labelled in panel ‘a’. Hypocenters beneath the main arc-axis locate between -1 and 16 km. Depths are characterized by a bimodal distribution, defining two levels of relative higher activity: one at 9±2 km and the other at 1±1 km of depth (see histogram in Fig. 2-6.d). Within these depth ranges, relative large-magnitude events tend to occur in the deeper level, whereas very shallow seismicity moment is weaker. In contrast, within the forearc between latitudes 38.5˚S and 39.5˚S, crustal hypocenters locate at significantly deeper-crustal levels, reaching depths of up to 40 km (e.g. epicenters to the west of the Villarrica Lake in Fig. 2-6.a and 2-6.c). In general, these events have larger uncertainties and locate outside of the network in areas of predominate 2-D structure (such as a magmatic arc). However, local earthquake tomography studies in the forearc and intra-arc regions between 39˚ and 40˚S do not resolve significant lateral velocity changes for the upper crust (Dzierma et al., 2012b), so the minimum 1-D velocity model with station corrections (Fig. 2-2) is a good approximation of the 3-D structure (Kissling, 1988). Therefore, we used maximal hypocenter depths from events close to our network (GAP<200) together with the more regional catalogue from Dzierma et al. (2014) to define the seismogenic layer bottom (SLB). Dzierma et al. (2012 a, b; see for instance cross sections P1 and P2 in Fig. 2-6 in Dzierma et al., 2012a) show how crustal seismicity observed in forearc locates much deeper than in the intra-arc, describing a convex-like distribution of events. On latitudinal cross-section view (Fig. 2-6.c), hypocenters delineate a ~120 km long convex geometry for the crustal SLB, which is shallower near the arc-axis, where the LOFS and ATFs are found. Moreover, intra-arc seismicity can be organized into three geographic domains with distinctive geometrical patterns, as shown in Figures 2-7, 2-8 and 2-9 and in Video 6-1 (S.M.): 57 (1) Northern domain (Fig. 2-7, 37.9˚S – 38.8˚S). This region contains 51 crustal events with small hypocenter uncertainties (±0.2, ±0.5 and ±0.4 km) in the NS, EW and vertical components, respectively. From south to north, 8 hypocenters locate on the north-eastern flank of Llaima Volcano and surrounding the Sierra Nevada (labelled Ll and SN in Fig. 2-7), on a plausible NNW- oriented cluster. Two focal mechanisms, including the July 31st 2014 (Mw =3.4) event –largest intra- arc event recorded during the experiment- are coherent with a dominant strike-slip setting with a NE-trending maximum subhorizontal shortening. The Mw 3.4 event occurred at the eastern tip off a ENE-oriented volcanic fissure in the northeastern flank of Llaima Volcano. A single extensional focal mechanism is observed beneath the Sierra Nevada volcano summit. This solution indicates ENE-oriented, steeply-dipping normal fault planes, whose geometry is similar to that of the ENE–striking dike swarm in Sierra Nevada western ridge (indicated as a narrow volcanic fissure in Fig. 2-7). At Lonquimay Volcano (Lo in Fig. 2-7), seismicity splits into two ~N20˚E and N30˚E striking bands, consistent with the northernmost branches of the LOFS (dashed blue lines in Fig. 2-7; modified from Pérez-Flores et al., 2016). These features occur along strike of prominent morphologic structures, as river incisions and aligned Holocene scoria cones (minor eruptive vents in Fig. 2-7). Two computed focal mechanisms indicate strike-slip and reverse faulting consistent with NE- to ENE- striking horizontal compression axes, respectively. By integrating surface mapped fault traces and the distribution of hypocenters, ENE- to NE-oriented fault surfaces might dip steeply east, shaping a splay fault array at the northern termination of the LOFS. (2) Central domain (Fig. 2-8, 38.8˚ - 39.3˚S). In this segment, we find oblique-to-the-arc striking seismicity clusters, which we describe from south to north: 58 i. NW- oriented cluster alignment involving the Villarrica Lake (VL), the Villarrica Volcano (Vi) and the Lanin Volcano (La) individual ~NW oriented clusters (Fig. 2-8.a) is consistent with the trend of the LFZ (e.g. Haberland, et al. 2006). Along this alignment, seismicity progressively deepens west- and eastwards away from the arc-axis, dipping with an angle of ~30˚ in-line with a rise of the seismogenic layer bottom (SLB) beneath the volcanic arc (latitudinal cross section of Fig. 2-8b). Towards the west, the forearc hypocenters reach depths of up to ~40 km, whereas beneath stratovolcanoes, depths are less than 13 km. Towards the east, seismicity is sparser, but clearly deepens down to ~17 km, with the exception of one event at 35 (±12) km depth beneath the Lanin volcano. This observation (convex SLB in sectional view) is supported by 107 crustal events located by Dzierma et al., 2012a (events with centered black dots in Fig. 2-8 a and b). Crustal activity clearly splits from interplate seismicity constrained by Dzierma et al. (2012a, b) who describe a ~30˚ east-dipping and ~20 km thick subducting slab geometry (Fig. 2-8b). In this overall NW- striking cluster alignment, intense activity (46 events with Mw between 1.1 and 2.5) took place at ~9 km depth underneath the Villarrica Volcano (Fig. 2-8c) from the 18th to the 23th of April – i.e. forty-five days after the March 3rd, 2015 Villarrica’s strombolian eruption, represented as the green dashed line in Fig. 2-5a and 2-5b. The NW-oriented seismicity cluster and accompanying computed focal mechanism indicate a dominant sinistral strike-slip regime with E-W±20˚ oriented maximal horizontal shortening axis (Fig. 2-8c). 59 ii. A N60˚W-striking micro-seismic sequence in the Caburgua Lake (Fig. 2-9 and Video 6-2) reveals a sub-vertical faulting process that took place within 03 successive marked steps. This is documented by peak activity occurring on November 13 to 15 (~68 hours), December 9 and December 16, 2014 (see temporal distribution in Fig. 2-5.a). Hypocenters distribute at all depth between 9.9 km and -1km depth (i.e. hundreds of meters below the surface in the mountainous sector), being the two more frequent depths ranges: 9.2 to 8.6 km (n=16) and 0.3 to -0.1 km (n=28). Similar to the Villarrica Sequence, narrow seismic alignment and computed focal mechanism delineate a 20 km long transpressional N60˚W left-lateral strike-slip fault with E-W ±20˚ oriented maximum horizontal shortening axis (Fig. 2-9). iii. A N50˚E-striking elongated cluster at the north-western flank and basement of Sollipulli Volcano (here called ‘Sollipulli cluster’), occurs in close spatial relationship to main fumaroles and surface geothermal activity (Fig. 2-9). A brief sequence, consisting of 11 NE- aligned events, occurred from the 12th to the 15th of October 2014, with the greatest magnitude event of 1.9 (Mw) on November 14th, 2014 09:09:36.7 UTC. A single focal mechanism indicates strike-slip faulting with N-S oriented maximum horizontal shortening axis. In combination with the overall cluster geometry, this would indicate left-lateral strike-slip faulting in a N50E-striking fault. Towards the north, deeper activity is clustered between 12 and 18 km in a NNW plunging cylinder-like cluster. In this sub-domain, cluster orientation can be associated with the LOFS. The latter activity occurring 13 km south from Llaima, and 14 km northwest from 60 Sollipulli stratovolcano summits, has been also reported by OVDAS and analyzed by Mora-Stock et al. (2012). This cluster occurs beneath the town of Melipeuco, in the Allipén valley (indicated with pink diamond in Fig. 2-9). Our computed focal mechanisms (n=04) in this secondary cluster reveal predominant strike-slip, but also reverse (01 event) faulting with NE-SW (03 events) to NW-SE (01 event) shortening axes (Fig. 2-9). (3) In the Southern domain, between 39.5˚ and 40˚ S, events distribute along two ~20 km long sub-parallel N10˚E-trending bands (Fig. 2-8c). Hypocenters often coincide spatially with surface mapped master branches of the LOFS. A small N-striking elongated cluster can be observed at Liquiñe (Fig. 2-8c). Two focal mechanisms reported here suggest normal faulting, with NS striking fault planes solutions, or strike-slip with north-south oriented maximum sub-horizontal shortening axis. 61 62 Figure 2-6 Southern Andes crustal seismicity. A) Mapview of local seismicity. Hypocenter location uncertainties for each event are indicated with error bars. Main volcanic and geothermal features (Sernageomin 2002; Hauser, 1997) are plotted. Holocene stratovolcanoes: Callaqui (Ca), Copahue (Co), Caviahue Caldera (Cv), Tolhuaca (To), Lonquimay (Lo), Llaima (Ll), Sierra Nevada (SN), Sollipulli (So), Villarrica (Vi), Quetrupillán (Q), Lanín (La) and Mocho-Choshuenco (MC). Quaternary minor eruptive vents are tele- detected and compiled in this study. Digital elevation model (DEM) from STRM30 (https://lta.cr.usgs.gov/SRTM1Arc). Central traces of N74˚W-striking swath cross-sections [W-W’ to Z-Z’] presented in Fig. 2-10 are indicated with black lines. Swath width to scale (gray bands). B) Longitudinal cross section along 71.5˚W. Only events with distances of less than 120 km to the profile are shown. The gray line is the average topographic elevation from the DEM. C) Like “b” but on a latitudinal cross section at 39˚S. d) Histogram (1 km bins) showing hypocenter depths. Figure 2-7 Crustal seismicity in the northern domain. The four events with centered black dots and absent uncertainty bars are from the seismicity catalog of Dzierma et al. (2012a). Events align along N~30˚E oriented fault branches of the LOFS and long-lived ATF (dashed when inferred; modified from Pérez-Flores et al., 2016). The N30˚W ATF Biobio-Aluminé fault (BBAF) is indicated with purple lines. Multiple minor eruptive vents (MEV) frequently are organized on ENE- to NE-oriented alignments. 63 Figure 2-8 a) Seismicity distribution of SAIAS catalog (with uncertainty bars) together with the regional catalog from Dzierma et al., 2012a (shown without uncertainty bars and centered-black dots). Hypocenters in red denote crustal events, showing NW-SE aligned clusters, e.g. the Villarrica Lake [VL]- Villarrica Volcano [Vi] – Lanin Volcano [La] cluster alignment and Caburgua Lake sequence [Cb]. Geological and seismic features follow the same symbol criteria than figure 2-7. b) Latitudinal cross section with 25 km swath width to both sides. Vertical exaggeration is 0.5. Our data (events in colored dots with uncertainty bars) is confronted to the SFB-Kiel project data from 2009 for the fore-arc (events in colored dots and black centered dots). Crustal seismicity depicts an upwelling shape for the bottom of the crustal seismogenic layer (red dots) at a theoretical 340˚C isotherm (e.g. Stesky, 1975; Scholz, 2002). For reference, deeper events of Dzierma et al. (2012a) are plotted delimiting the interplate seismogenic zone (Wadati-Benioff zone). A referential Mohorovicic discontinuity is outlined based on receiver functions (Yuan et al., 2006) and our local velocity model (white dashed line). c) Inlet indicated in ‘a’. Hypocenter distribution and computed focal mechanism for shallow events occurred within the Villarrica volcanic edifice – i.e. a NW oriented cluster southeast of the summit and another sparse cluster north of the summit - and along major N10˚E oriented fault branches of the LOFS. Note different depth color coding for panels a and c. Inlet of southernmost portion of figure 2-9 is indicated. 64 Figure 2-9 Close-up map view with seismicity of the central domain (location of this map is shown in Figure 2-8.a). Computed focal mechanisms are indicated in dark-green and white, whereas moment tensors for the Mw 5.2 on 24/02/1989 at 15 km of depth (gCMT) is plotted in light-green and white. The Melipeuco town is indicated with a pink diamond. Other symbols follow the same criteria as previous figures. 65 2.5 Discussion In the south-central Andes, along the LOFS northern termination, intra-arc seismicity enables the identification of current deformation/stress states and associated faulting. Intra-arc seismicity can clearly be discriminated from crustal forearc seismicity based on the following observations: (i) its shallowest typical hypocentral depth, ranging from ~-0.2 to ~12 km (Fig. 2- 6.b,c,d and 2-8.b and 2-10; (ii) its predominant strike slip kinematics; and (iii) its geographical occurrence on or close-to long-term crustal faults (LOFS and ATF) and/or Quaternary eruptive centers and geothermal areas (see Fig. 2-6 for regional context and Fig. 2-7, 2-8 and 2-9 for local spatial relations). In turn, forearc crustal seismicity is mostly localized on ATF (e.g. Pichilemu Fault, Lanalhue Fault Zone) and might be continuous to the intra-arc, as the LFZ follows the Vl-Vi-La NW- striking seismic cluster alignment. This point will be further addressed later. As it is widely accepted, the b-value is a constant for the Gutenberg-Richter (G-T) relation that measures the ratio of large earthquakes over small ones. Therefore, the b value is related to the fractal dimension of faulting (e.g., Aki, 1981; Turcotte, 1997). Globally, the b-value is frequently found between 0.8 and 1.2 for a wide variety of active tectonic regions (e.g. Utsu, 2002). Our estimates yield to a b-value of 1.11 (data fits the G-T relation up Mco; Fig. 2-4), which is coherent with a tectonic origin of the intra-arc seismicity. Additionally, the tectonic nature is confirmed by the dominant stick slip behavior of active intra-arc faulting, deduced from clear and distinctive P- and S- wave onsets for most recorded events (Byerlee and Brace, 1968). Recently, similar b-values and faulting mechanism behavior have been documented for crustal seismicity in the pre- cordillera of northern Chile (Salazar et al., 2017). Although, crustal composition, fluid presence/absence and crustal thickness might differ from northern to southern Chile, we can 66 support that the first order control for crustal seismicity derives from tectonic forces associated with the overall oblique convergence setting. The frictional-plastic transition in the intra-arc region Based upon our observations, the thinner and shallower intra-arc seismogenic zone of the SVZ may be the most favorable crustal domain to accommodate frictional strain in the overriding South American plate. This is consistent with an earlier observation made by Barrientos et al. (2004) for Central Chile (33˚- 35˚S) indicating a higher rate of crustal seismicity within the arc, in comparison to the observed activity in the forearc and back-arc regions. This idea may also be in harmony with an enhanced thermo-mechanical weakness of a heated crust, as originally proposed (Fitch, 1972, Tikoff and Teyssier, 1994, De Saint Blanquat et al., 1998). Furthermore, because seismic velocities decrease with increasing temperature and deformation (Griffin and O’Reilly, 1987; Sibson, 1984; Sibson, 2002), a heated and fractured crust might explain the relatively low velocity structure in the upper intra-arc crust (Fig. 2-3a) when it is compared to other 1D crustal velocity models calculated for Southcentral and Central Chile (Haberland et al., 2006, Lange et al., 2007). Thus, arc-orthogonal geothermal gradients may significantly influence the brittle architecture of the overriding plate, enhancing faulting where the crustal seismogenic layer is thinner. In quartzofeldespatic crust, greenschist metamorphism or quartz plasticity onsets at around ~300˚C (Stesky, 1975). This sub-solidus crustal boundary marks an inflexion point of crustal strength (Scholz, 1988; Sibson, 2002). Thus, in continental margins with an active volcanic arc, the vertical transition from a frictional faulting domain to quasi-plastic shearing may be significantly deeper in regions tens of kilometers away from the arc axis. Moreover, in situ studies 67 in northern Japan combining local seismicity and direct temperature measurements of geothermal-boreholes in granitic crust, indicate that the SLB in this region corresponds to the 380˚C isotherm (Suzuki et al., 2014). They correlate this temperature with depths for the thermal conduction/convection transition. Whereas the maximum strength zone would be consistent with the 340˚C isotherm, where the brittle-ductile transition is defined (Suzuki et al., 2014). Based on our data, quasi-plastic mechanical behavior of the crust would be expected for depths below the convex envelope of the crustal seismogenic layer bottom, as seen in section (data in Fig 2-8b, Fig. 2-10 and schematized in Fig. 2-11). This feature is conspicuous between latitudes 38.5˚S and 39.5˚S where crustal seismicity is organized within a NW-oriented seismic volume, consistent with the NW-striking Lanalhue Fault (e.g. Haberland et al., 2006), NW-aligned clusters (VL-Vi-La) and with the NW oriented Caburgua crustal seismic sequence (Fig. 2-8a, 2-8b and Video 6-1). A similar shape of the SLB, can be observed to the kilometric-scale in active geothermal fields of northern Japan, where high resolution passive seismic mapping, define local and abrupt shallowing of the SLB at very high geothermal gradients (Suzuki et al., 2014). 68 Figure 2-10. Cross-sections (perpendicular to the LOFS and SVZ) showing the local seismicity, together with the regional earthquake catalogue from Dzierma et al. (2012). Location and swath of the profiles are shown in Fig. 2-6. Local events are shown with uncertainties. Events close to the network (GAP<=200°) are indicated with colour circles. Seismicity far outside the network (GAP>200°, poor constraint depths) are shown in white circles, and are not used for interpretation. Events from the earthquake catalogue from Dzierma et al. (2012 a, b) are plotted as black circles. Strato-volcanoes are indicated with light-blue triangles and seismic stations are shown with inverted triangles (black=broadband, white=short-period). Note that vertical and horizontal scales are equivalent. Slab geometry (grey line) inferred from interplate seismicity (Dzierma et al., 2012a) is dashed where no data is available and is projected from southernmost cross-sections. Topographic relief is projected from SRTM 30 data. Northernmost swath cross-section (W-W’) is 20 km wide, whereas the other three (X-X’, Y-Y’ and Z-Z’) are 24 km wide. Similar crustal geometries have been observed for the Central Andes. For example, Farías et al. (2010) associated the west-dipping hypocenter distribution in Central Chile (33˚-35˚S) with 69 the shape of the continental margin ~400˚C isotherm, which was earlier defined at a more regional scale by Yañez and Cembrano (2004). Farías et al. (2010) interpreted this structure as an east-vergent, low angle thrust that would play a major role in the Andean mountain building. If this constitutes the main crustal architecture of the Chilean Southern Andes, the forearc seismicity should be decoupled from the intra-arc seismicity and most of the hypocenters should be distributed close to the base of the SLB. The observed seismicity should then define a low angle, west-dipping seismogenic plane and dip-slip focal mechanism would be expected. Nonetheless, activity in the forearc between 38.5˚S and 39.5˚S (ATF) occur in a wide depth range (0-40 km; Fig. 2-8b) and focal mechanisms are mostly strike-slip (Haberland et al., 2006), indicating a prominent NW oriented Andean-transverse deformation system. The latter could be defined as a NW oriented crustal volume that includes the Lanalhue Fault (Glodny et al., 2008, Haberland et al., 2006, Melnick et al., 2009), the VL-Vi-La seismic cluster alingment and the Caburgua seismic sequence. If the upper crust is detached by a subhorizontal convex layer (e.g. SLB), then the partitioning would not show predominantly in the magmatic arc because the tectonic forces in the forearc would be accommodated close to the SLB and not in the whole brittle-crust section, as it is revealed by ours and others seismic experiments (Dzierma et al., 2012 a, b; Haberland et al., 2006). So, what is the source, then, to the conspicuous observed strike-slip faulting? We consider that the partitioning of convergence between subduction and strike-slip movement of the forearc slivers constitute a simple and more straight-forward explanation to the observables, and that the low angle detachment alternative is at least equivocal in this segment of the Andes (38˚S-40˚S). Recent local seismicity studies in northern Chile forearc (Bloch et al., 2014; Salazar et al., 2017) demonstrated an analogue and marked convex shape of the SLB boundary around the 70 West Fissure Fault System (WFFS) and Northern Chile volcanic arc. By contrasting hypocenters to heat-flow density models, Salazar et al. (2017) observed that crustal seismicity is clearly organized above the 350˚C isotherm (Springer, 1999). The latter, down to ~60 km depth in the forearc-slab boundary, progressively shallows up to ~15 km beneath the active volcanic front, describing a convex geometry in line with the maximum depth of crustal events. In turn, in the southern Central Andes (32˚ - 33.4˚S), Nacif et al. (2017) show similar hypocentral depths beneath the volcanic front (< 12 km). Integrating EHB Bulletin Catalog data, this author, indicates also how foreland/back-arc seismic activity locates progressively deeper to the east, down to reach the continental Moho (~ 50 km depth), about ~550 km away from the trench. In contrast with Farías et al., (2010) interpretation of an east-vergence major crustal thrust, and in agreement with Salazar et al. (2017), we suggest that a first-order control of the convexity and thinning of the SLB in Northern and Southern Chile volcanic arcs is related to a critical isothermal envelope limiting the frictional deformation field in depth. The geometry of this rheological transition might be caused by thermal diffusion of active magmatic heat sources along the arc, independent of inherited fault geometries, but probably from fluid content and availability. Additionally, reflection seismic studies in Northern Chile define two sub-parallel west-dipping bright spots between 15 and 40 km of depth (Yoon et al., 2003; Yoon et al., 2009). These structures are consistent in geometry and position with the SLB defined by Bloch et al., (2014) and Salazar et al. (2017), as well as with the 350˚C isotherm (Springer, 1999). Towards the arc front, where activity is shallower (15-20 km depth) these structures have been correlated to the Altiplano low-velocity zone, the one has been proposed to demark the brittle-plastic transition in this region (Wigger et al., 1994; Yuan et al., 2000). In summary, for the Southern Andes, the SLB might delineate a major rheological boundary between upper and mid-crustal levels, similarly as was proposed by 71 Koulakov et al. (2006) and Bloch et al. (2014) for the Andes orogeny in northern Chile. For this segment of the Andes, we assume that the SLB temperature should be within the temperature range of 300˚-380˚C, considering a quartzofeldespatic crust mainly formed by Tertiary granitoids and volcano-sedimentary rock successions (refer to Fig. 6-1 in supplementary information). However, the SLB shape might not only be controlled by the geothermal gradient, but by the strain rate and the presence of crustal fluids, diminishing rock strength to failure. Long-term geologic data shows that continental margin high-strain domains are mostly localized along the Miocene-Present day volcanic arc and/or at long-lived ATF (e.g. Rosenau et al., 2006; Glodny et al., 2008; Cembrano and Lara, 2009; Pérez-Flores et al., 2016; among others). If the current strain rate is indeed higher in the volcanic arc domain, the SLB should be deeper than that of the forearc and foreland regions given a constant geothermal gradient, which is something not supported by our data and that of previous work (e.g. Nacif et al., 2016; Salazar et al., 2017). Therefore, although strain rate in the volcanic arc is higher, the effect of heat flux and associated higher geothermal gradient largely overcomes that of strain rate. ATF can provide key evidence in favor of the role of differential geothermal gradient on the depth of seismicity because they run across the forearc and arc regions. For instance, the maximum seismicity depth recorded along the strike of the LFZ goes from ~40 km in the forearc to ~12 km within the arc (Haberland et al., 2006; Dzierma et al., 2012a; in addition to our data). This strongly supports the hypothesis that differential geothermal gradient is one first order control on the maximum depth for seismicity on Andean crustal faults. Regarding the influence of crustal fluids, it is well known their effect on reducing effective stresses and promoting failure. Regions with enhanced fluid pressure at all crustal levels will tend to have a deeper SLB. If fluid pressure is not evenly distributed with depth, the effects are difficult 72 to generalize, because they will depend on local conditions. In the Southern Volcanic Zone, there is widespread evidence for upper crustal fluid flow in the form of magma and/or hydrothermal fluids/hot springs (i.e. NE-oriented electrical conductor; Brasse et al., 2008). This would probably have an effect on enhancing faulting and seismic activity in the uppermost crust (~0-2 km); effects at higher depths are difficult to constrain at regional scales. A local example of this, however, will be discussed when addressing crustal fluid-related seismicity below. Strain partitioning and compartmentalization Intra-arc seismicity provides fundamental insights on how transpressional deformation partitions, transfers and compartmentalizes within the overriding plate. The results of our brief time-window (~16 months) experiment, however, is reasonably consistent with most of the main tectonic features already documented through different approaches/methodologies (i.e. structural geology, numerical modelling and geodetic studies) on the long-term pattern of crustal strain within intra-arc regions. Tikoff and Saint Blanquat (1997) have stated that syn-magmatic strike-slip partitioning may accommodate within margin-parallel strike-slip mylonitic and cataclastic domains. In turn, a combination of strike-slip and reverse faulting in the forearc and foreland/back-arc regions may accommodate both margin-parallel and margin-orthogonal components during one interseismic stage of the subduction earthquake cycle. However, few studies (La Femina et al., 2002; Lange et al., 2008; Stanton-Yonge et al., 2016; Pérez-Flores et al., 2016; Salazar et al., 2017), which include detailed observations, yield a more complex picture regarding the way by which the margin-parallel component accommodates within the intra-arc. In fact, it might be entirely accommodated by margin-oblique faults (e.g. bookshelf faulting in Central Andes volcanic arc; La Femina et al., 2002); or strain might distribute within variable 73 orientations and with a wide kinematic range, as deduced from focal mechanisms in Northern Chile, with statistically dominant oblique-slip kinematics, with substantial strike-slip component (Salazar et al., 2017). The observations include both the long-term and present-day interplay between margin-parallel and transverse fault zones, which in turn are spatially associated with volcanism and hydrothermal activity (e.g. Cembrano and Lara, 2009; Sánchez-Alfaro et al., 2013). The continental continuous crustal deformation (displacement) model for Latin America (VEMOS2015 model with spatial resolution of 1˚ x 1˚; Sánchez and Drewes, 2016) inferred from GNSS (GPS+GLONASS) measurements gained after the 2010 Maule megathrust earthquake, indicates how the effect of the Maule earthquake changed the surface kinematics of a large continental area (30˚-45˚S), particularly in the Andean region. The model indicates, that before the earthquake, the strain field showed a strong west-east compression between 38˚S and 44˚S (roughly parallel to the convergence vector) whose magnitudes diminish with distance from the subduction front. After the earthquake, and until the end of the measurements in 2015, the larger horizontal compression was documented between 37˚S and 40˚S, with a N30˚E oriented axis. The authors predict that the present-day deformation (valid for the period 2010-2015) caused by the Maule earthquake, might extend between 30˚S and 45˚S (Sánchez and Drewes, 2016). However, no increase in seismicity was observed along the magmatic arc after the Maule earthquake (Lange et al., 2012; Mora-Sock et al., 2011). Similar observations have been made for other subduction regions (e.g. at Sunda subduction zone the 2004, 2005, 2007 and 2012 earthquakes where not followed by higher seismicity rate along the Sumatra Fault, although predicted by Coulomb stress models (e.g. McCloskey et al., 2005; Pollitz et al., 2006). As a general approach, the N30˚E horizontal shortening derived from VEMOS2015 would be in kinematic coherence with the numerical modelling implemented by Stanton-Yonge et al. (2016), however 74 the far field stress tensor differs from the N30˚E horizontal shortening. The latter study, computes and compares the expected displacements and kinematics on long-lived crustal faults in Southern Andes (LOFS and ATF) for both, co- and interseismic stages of the subduction earthquake cycle. Beside local anomalies induced by post-seismic deformation and heterogeneous locking distribution, the interseismic model fits well GPS data (Wang et al., 2007; Moreno et al., 2011 among others). The model yields right-lateral slip with a maximum rate of 3.5 mm/yr on the Eastern Master of the LOFS (involved in our study area) and oblique sinistral- reverse slip for NW striking ATF with the potential to accommodate a maximum slip rate of 1.4 mm/yr. In turn, ENE striking fault would display dextral slip along strike at a maximum rate of 0.85 mm/yr. This partitioning pattern in the intra-arc region is in agreement with the faulting geometries and with several strike-slip and compressional focal mechanisms, with NE to NS maximum horizontal shortening here presented (Figs. 2-7, 2-8 and 2-9). Depending on the epicentral distance and position, for a coseismic phase, in turn, we would expect an inversion of the strain pattern and fault kinematics here described. As presented by Stanton-Yonge et al. (2016), coseismic deformation field resulting from the 2010 earthquake induced normal-sinistral slip along the LOFS with a maximum slip at its northern termination. Accordingly, NW-striking faults (in the latitudinal range of the earthquake rupture area) would switch to dextral-normal slip, and ENE oriented faults, to sinistral slip. Unfortunately, no coseismic crustal earthquake has been reported for such arc-parallel or arc-oblique faults, apart from the post-seismic, NS-striking right-lateral strike-slip El Melado 6.0 (Mw) earthquake (Fig. 2-1; Cardona et al., 2015). Nonetheless, Cifuentes (1989) reports a Ms ~7.5 coseismic crustal earthquake (June 6th 1960, i.e. 15 days after the 1960 Valdivia megathrust M 9.5 earthquake) on top of the surface trace of the LOFS southern termination. Previously, Kanamori and Stewart 75 (1979), proposed right-lateral slip on a N80˚E oriented fault plane for this event. Although this solution is poorly constrained, it suggests a NW- oriented horizontal coseismic compression axis and NE- oriented horizontal coseismic extension axis, which are in accordance with the coseismic strain field proposed by Stanton-Yonge et al. (2016). However, expected coseismic activity within the intra-arc seismogenic zone might occur in compatibility and dependence of the coseismic strain field and intra-arc faults geometries. Spatial and spatio-temporal analyses of our data reveal variable faulting styles taking place predominantly within the uppermost 12 kilometers of the magmatic arc (schematized in Fig. 2-11). Predominant NNE and subordinate NE-striking faults of the LOFS, in addition to the NW-striking ATF, shape the active intra-arc schizosphere seismogenic fabric. For instance, high angle, east-dipping mapped faults (Lolco Fault, Pérez-Flores et al., 2016), and slightly east-shifted N20˚E to N30˚E striking hypocenter alignment, configure a splay fault array at the northern termination of the LOFS (Fig. 2-7). In turn, the central zone of the study area is characterized by Andean transverse faulting. This is evident from isolated events and clusters aligned on a >100 km long NW striking band, extending from the Lanin Volcano towards the forearc (i.e. VL-Vi-La in Fig. 2-8a). We propose that such Andean transverse strain domain is controlled by a long-lived (pre-Andean) NW-striking faults with sinistral-reverse kinematics as the dominant shortening pattern deduced from focal mechanism of the Vi cluster. On a smaller scale, kinematically compatible activity on tens-km long ATF is prompted on narrow NE- (e.g. Sollipulli cluster) and NW- (e.g. Caburgua cluster) striking intra-arc seismic fault zones. Temporal evolution, geometry and kinematics of micro-seismicity at the Caburgua Lake, could be related to a swarm-like behavior (Fig. 2-9). The southernmost zone reveals a spatial distribution of crustal seismicity in association with major branches of the LOFS. In particular, the Liquiñe, Villarrica and Caburgua 76 clusters, might be consistent with fault geometries and kinematics of second order faults of the LOFS and ATF with Holocene neotectonic activity, as suggested from geomorphic analysis by Melnick et al. (2006b) and Astudillo et al. (2018). Detailed and focused paleoseismicity studies in such key sites (clusters) are needed in order to better understand the LOFS and ATF Quaternary slip rates. Kinematic solutions estimated from seismic source radiation patterns (i.e. computed focal mechanisms on Figures 2-7, 2-8 and 2-9 and on Table 6-2) confirm a prevailing ENE to WNW- instantaneous maximum subhorizontal compression, and local NNE-SSW to N-S compression. The former findings are largely consistent with the long-term deformation recorded by structural geology analysis, and are slightly rotated from the few gCMT (Mw ≥5) moment tensors (Fig. 2-1, 2-7 and 2-9). The shortening axes are in turn more coherent with the N30˚E post-seismic shortening axes proposed by the VEMOS2015 model (Sánchez and Drewes, 2016). However, local NNE-SSW to N-S compression solutions, as well as few normal faulting focal mechanisms (i.e. vertical maximum shortening) on subsidiary faults of the LOFS, such as the ones at the Sierra Nevada Volcano and Liquiñe clusters (Fig. 2-7 and 2-8.c, respectively), might be associated with heterogeneous strain accommodation on damage zones neighboring major faults. It is known, from self-similarity and scaling-parameters studies on fault zones (system that is similar to a part of itself) that faults and related damage zones are created by scale-invariant mechanisms (Jensen et al., 2011; Savage and Brodsky, 2011; among others) and that collateral damage zones may host diverse geometries and kinematics because of rotating blocks, media anisotropies (e.g. healing of previous faults generating buttress boundaries) and/or by stress variations on fault intersections (Kim et al., 2004; Gratier, 2011; Iturrieta et al., 2017). All above-mentioned is consistent with an overall strain partitioning along the arc strike compartmentalized into margin- 77 parallel and Andean-transverse fault domains. Major along-strike changes can be visualized on cross-section of Fig. 2-6b and Video 6-1. Small-magnitude focal mechanisms are expected to be associated either with small displacement along main fault branches, or alternatively to restricted surfaces on subsidiary faults. The former is in line with the natural varying seismic style and asperity sizes along fault strike (e.g. Sibson, 2002). The latter may explain the local clock-wise rotated shortening axes due to slip on subsidiary faults. The experiment resolution is not high enough to specify the source nature of such small magnitude events. However, by considering fault outcrop scale observations, it is reasonable to assume slip on smaller-scale subsidiary faults, as those analyzed by structural geology studies such as those of Lavenu and Cembrano (1999), Rosenau et al. (2006), and Pérez-Flores et al. (2016). 78 Figure 2-11 Conceptual model of crustal seismogenic layers along a magmatic arc. The cartoon includes high frictional strain domains (LOFS and ATF; olive-gray volumes), where expected fault geometries and kinematics respond to the regional shortening axis (black vectors). Strato-volcanoes are represented with the red triangles. Based on receiver functions analysis, Yuan et al., (2006) resolve the continental Moho beneath the Southern Andes. The Moho is described as a continuous envelope that is deeper beneath the magmatic arc and orogenic front (~45 to 40 km), becoming progressively shallower towards back- and forearc regions (30 – 20 km). The Vp and Vs velocity increment that we see at 35 and 40 km in our minimum 1-D velocity model (Fig. 2-3a) is in line with the regional Moho depth from Yuan et al., (2006). Fluid-related seismicity and the Villarrica 2015 eruption No clear volcano-tectonic activity has been determined by the present study. However, clustered seismicity, as the one reported between Llaima and Sollipulli volcanoes, might be fluid- related (Fig. 2-9, northern and deeper cluster). In particular, this site of activity has been previously reported by OVDAS and analyzed by Mora-Stock, et al. (2012). The authors suggest a 79 strong fluid-depend seismic component, as deduced from low frequency events. In particular, this cluster represent the deepest intra-arc activity here reported. This particular cluster - up to 6 km deeper in comparison to the more frequent ~9 km maximum depth (SLB) along the intra- arc strike -, might be related to deeper-seated hydrofracturing, as a consequence of effective stress reduction, rock-strength weakening and failure (e.g. Cox 2010). Usually, individual moderate shallow earthquakes occur as localised shear dislocations which do not lead to displacements over the entire surface of an existing geological fault; in turn, seismic faulting may act as a pumping mechanism whereby individual seismic events are capable of moving significant quantities of crustal fluids (Sibson, 1975). Along the south-central Chile arc- strike, an important fraction of the seismicity (98 events, i.e. 27% of our catalog; Fig. 2-6d) locates in the uppermost 4 kms of the crust. For instance, the brief timespan, intermitent and very shallow micro-seismicity observed at the Caburgua NW-striking cluster, might be related to episodic fluids circulation, that could be retaled to sporadic permeability enhancement and seismic swarm-like behavoir. No evident precursory seismic activity can be linked to the Villarrica strombolian eruption that occurred on March 3rd 2015 (OVDAS, 2015; Naranjo, 2015). Furthermore, we report only 6 events in a 40-day window before the eruption, absence of co-eruptive seismicity, and only 11 events in the following 40-days (Fig. 6-4). We carefully rechecked the seismograms for 20 days before and after the eruption for small magnitude events, but could not find events which were not previously detected during the catalog creation. Normally, volcano-tectonic pre-eruptive seismicity respond to spatio-temporal evolution of magma-filled cracks expected to be caused by: 1) hydro-fracturing (Aoki et al., 1999; Gudmundson, 2011); 2) faulting linkage between offset dikes (Hill, 1977) or 3) lateral micro-faulting in wall-rock as a mechanical response to stress 80 changes induced by dike inflation (Roman, 2005; Roman and Cashman, 2006). Since no clear seismicity is found prior to the Villarrica 2015 eruption, none of the processes could be associated with the eruptive cycle. Curiously, forty-five days after the Villarrica eruption, an intense seismic sequence took place at around 9 km depth (See Fig. 2-5a and Fig. 6-4 to see the increment of number in time and Fig. 2-8c for detailed observation of the cluster distribution). Strike-slip focal mechanisms and hypocenters describe a NW oriented ellipsoidal cluster, located 2 km to the south-east of the eruptive vent. We speculate that in this position, aseismic (or not detectable by our network) feeder dikes may have been injected. We interpret this particular seismic sequence as an effect of transient, thermo-dependent seismogenic properties of the upper intra-arc crust, where post-eruption micro-cracking and faulting might occur, induced by the reduction of confining stresses due to cooling and contraction of the rock mass. Similar magma-wall-rock behavior has been suggested for small double, and non-double couple earthquakes in Reykjanes, Hengill and Krafla volcanic systems of the accretionary plate boundary in Iceland (Foulger and Long in Gasparini et al., 1992). After the post-eruptive seismic sequence at Villarrica in April 2015, seismicity would have recovered the regional background activity level (Fig. 6-4). However, no records are available after the 13th of June 2015, the completion date of the experiment. The lack of notable seismicity rate changes after the eruption in March 2015 indicates that seismicity is not a reliable precursor for Villarrica volcano, at least in this eruptive cycle. 2.6 Summary and Conclusions In this study, we present moderate to small magnitude seismicity ranging from Mw 3.6 to Mw 0.6, of a 16 month-long natural seismicity experiment. It is the first network to cover the northern termination of the Liquiñe-Ofqui Fault System at a regional scale, and provides high 81 resolution for local earthquake detection and localization. The seismicity catalog consists of 356 crustal events, coherent with the bulk long-term tectonic construe from structural geology analysis for this segment of the Andean orogeny. Combined interpretations from geometric and kinematic analysis can be summarized into three major outcomes. 1. Frictional strain in the intra-arc region is compartmentalized into geometrically and kinematically distinctive latitudinal subdomains. In the northernmost portion of the LOFS (37.8˚- 38.8˚S), seismic faulting is consistent with the long-term record of brittle deformation. Main NNE- oriented branches of the LOFS give way to a splay fault array, along which minor eruptive centers, stratovolcanoes and geothermal surface expressions are placed. A central zone (38.8˚-39.6˚S) is dominated by Andean transverse active faults, alternating NW- (e.g. Caburgua sequence, VL-Vi-La seismic cluster alignment) with NE-oriented faulting. In the southernmost sector of the study area (39.6˚-40.1˚S), margin-parallel faults accommodate most of the strike-slip component of the partitioned bulk strain. 2. The latitudinal convexity (as seen in WNW-ESE cross sections) and 30˚ west dip of the SLB in Southern Andes between latitudes 38.5˚S and 39.5˚S, is in-line with a geothermal gradient exerting a first-order longitudinal control on the depth of the frictional-plastic transition. From a geometric and linear extrapolation, we suggest that the geothermal gradient is at least three times higher in the arc-axis than in the forearc. Consequently, a higher geothermal gradient and thinner fractured crust in the arc-axis may enhance a feedback loop among frictional strain, permeability increase and convection conditions for shallow fluids to flow and/or emplace. 3. The Villarrica volcano strombolian eruption on March 3rd, 2015 did not have a significant pre-, syn- or post- eruption seismicity. However, an intense sequence, took place 45 days after 82 the eruption at ~9 km depth in a NW-elongated cluster from the summit towards the southeast. We argue that this activity may have been induced by feeder dike cooling and contraction. 2.7 Acknowledgments This research project has been supported by FONDAP project 15090013 CEGA “Centro de Excelencia en Geotermia de los Andes” and Fondecyt project 1141139 to J.C. Gerd Sielfeld acknowledges the support by CONICYT’s national Ph.D. scholarship 21140749. Seismometers and EarthDataLoggers for the seismic network were provided by the Geophysical Instrument Pool Potsdam (GIPP, GFZ Potsdam, loan: LOFZ). We thank Christian Haberland for his technical support. We thank all Chilean landowners, companies and institutions for allowing us to install the seismic station on their property. We are grateful for the help provided by the field crews (P. Pérez-Flores, P. Guzman, P. Iturrieta, A. Stanton-Yonge, N. Pérez-Estay, P. Azúa, D. Tardani, M. Lizama, J. Riquelme, J. Puratic, F. Rodriguez, R. Rivera, M.P. Lira, K. Muñoz, O. Iturrieta, V. Vicenzio, E. Lira, A. Órdenes, A. Salazar). We are thankful to the OVDAS crew by providing their seismicity catalog, which allow us to improve our detection quality and calibrate magnitudes. G.S. also thanks GEOMAR and Heidrum Kopp by allowing productive internships for the processing of the seismic data. We also thank Martin Thorwart for providing the seismicity catalog of Dzierma et al. (2012a and 2012b). Finally, we thank G. Yañez, P. Iturrieta, A. Stanton-Yonge, P. Salazar, J. Crempien, N. Pérez-Estay, P. Pérez-Flores, P. Sánchez-Alfaro, F. Aron, J. Paredes-Mariño and R. Gomila for their helpfully comments and discussions during the development of this research. Becky Pearce, from the University College London, kindly proofread the final manuscript version. The complete catalog is available in the Supporting Material. 83 3 Transtension driving volcano-edifice anatomy: Insights from Andean transverse-to-the-orogen tectonic domains 3.1 Introduction and problem statement Evolution of continental magmatic arcs and associated crustal deformation are widely dependent on active convergence at destructive plate margins (e.g. Jarrard, 1986; Scheuber and Gonzalez, 1999). Therefore, in subduction settings, long and short-term tectonic activity plays an essential role in the generation, segregation, transport and emplacement of melts and magmas within the overriding lithosphere (Valerio Acocella & Funiciello, 2010; Annen et al., 2006; Clemens, 1998; Clemens & Petford, 1999; Zellmer et al., 2005). Magma long-term ascent and emplacement mechanisms within magmatic arcs are largely controlled by the interplay among inherited structural anisotropies, regional stress field, crustal deformation, magmatic pressure, and the viscous resistance to magma flow (Saint Blanquat et al., 1998; Delaney et al., 1986; Dumond et al., 2005; Petford et al., 2000; Tibaldi & Lagmay, 2006). Low viscosity magmas are expected to propagate as sheet-like intrusions oriented perpendicular to the minimum principal stress axis at any depth within the lithosphere (Emerman & Marrett, 1990). Furthermore, most magmas emplaced at shallow crustal levels and giving rise to volcanic eruptions migrate through existing fractures controlled by the regional tectonic setting and hydrofractures that are driven either by magmatic overpressure (Gudmundsson, 2011) and/or by differential stress regime (e.g. Acocella & Neri, 2009; Nakamura, 1977; Shaw, 1980). Although extensional tectonics is considered as a fundamental requirement for volcanic output at any arc-tectonic setting (Acocella and Funiciello, 2010), many authors have 84 stated that structures related to brittle deformation formed prior to magmatism play a role on controlling the spatial distribution and geometry of minor eruptive vents and related feeder dikes (e.g. Brogi et al., 2010; Corazzato & Tibaldi, 2006; Delaney et al., 1986; Emerman & Marrett, 1990; Hou, 2012; Lara et al., 2006; Nakamura, 1977; Shaw, 1980; Tibaldi, 1995). Therefore, by integrating volcanic eruptive centres location, dike orientation and cone morphometry, we can get new insights on the interplay among magma migration, tectonic setting and volcanic architecture (Fig. 3-1). The interplay among volcano anatomy, regional differential stresses, edifice-load and erosional pattern has only been poorly constrained. This study aims to a better understanding of such causal interactions. We here postulate that stratovolcanoes emplaced on top of inherited basement faults parallel or close-to parallel to the shortening axis in regional transpressional settings may achieve an elongated anatomy driven by parallel dike propagation. Feeder dike predominant propagation and eruption direction may become organized along either (a) trends parallel or close-to-parallel to the maximum horizontal shortening at distal zones of the main volcano edifice (Andersonian dikes), and/or (b) dilate and propagate through favourable oriented long-lived basement faults beneath the volcano-edifice (non-Andersonian dikes). Predominant orientation may prevail in time enhanced by elongated volcano construction and normal-to-the- volcano elongation de-confinement. The aim of this research is then to contribute to the understanding of how dikes and minor eruptive vents (MEV) within a stratovolcano can be organized in time and space by the operating tectonic field, inherited structural anisotropies and the localized effect of volcano edifice shape and load. An optimal candidate to address these questions is the Callaqui Stratovolcano, emplaced on a transitional segment between Central and Patagonian Andes, 85 where major geological and tectonic changes take place. By focusing on dikes and MEV of the Quaternary Callaqui Stratovolcano (38ºS, Andean Southern Volcanic Zone), structural and morphometric analyses, complemented with remote sensing data, we provide a consistent interpretation for the orientation of the stress/strain fields during the time of magma propagation, emplacement and eruption. Classical field mapping methods were applied, as well as detailed description of dike geometry and MEV morphometry (e.g. Tibaldi, 1995). Results reveal a combined effect of regional stress-field, pre-existing fault zones and gravitational/erosional control on the loci and nature of the volcanic-scale strain, dike emplacement and monogenic eruptions along the fissure-like Callaqui Stratovolcano (CSV). Geometrical analysis indicates an overall right-lateral transtensional tectonic regime during the time of volcanism, being effective in the long-term shaping of an elongated anatomy of the edifice. The understanding of the building processes of tectonic-driven volcano morphologies is crucial to minimize hazards related to volcano instability at populated areas, such as in the Southern Volcanic Zone of Chile and Argentina. Also it contributes to the understanding of structural controlled volcanic system that may hold active geothermal systems. 3.2 Tectonic and geological setting The tectonics of the Southern Andes magmatic arc (33ºS and 46ºS – also known in literature as Southern Volcanic Zone; López-Escobar et al., 1995) is governed by dextral-oblique convergence between the Nazca and South American plates (Pardo-Casas and Molnar, 1987; Angermann et al., 1999). In this Andean segment, the convergence vector strikes N80º with an average velocity of ca. 66 mm/y (Fig. 3-2) that has remained relatively constant over the last 20 Ma (Pardo-Casas and Molnar, 1987; Somoza, 1998; Angermann et al., 1999). 86 In the Southern Andes, partial strain partitioning is evidenced by compressional tectonics in the fore-arc and foreland regions, and by pure strike-slip to transpressional tectonics in the intra-arc region (Blanquat et al., 1998; Cembrano et al., 2000; Corbett & Leach, 1998; Dewey & Lamb, 1992). The latter is represented by the long-lived Liquiñe-Ofqui Fault System (LOFS) consisting of a 1200 km long, N10º striking, right-lateral strike slip fault system, spatially related to Quaternary volcanism (e.g. Cembrano et al., 1996; Fig. 3-2). The first-order architectural grain of the Andean orogenic belt consists of margin-parallel folds and fault systems (e.g.Charrier, 2007; Costa et al., 2006; Fig. 3-2). These structural elements shape the landscape, organizing from west to east the main orogenic morpho-structural units into: Coastal Range, Central Depression, Principal Cordillera and the foreland fold and thrust belt. However, second-order structures (transverse fault systems), cross-cutting the main morphotectonic domains have been claimed to play a key role in the subduction earthquake segmentation (Melnick et al., 2009) and in the emplacement of regionally aligned volcanic chains (Cembrano and Lara, 2009; Lara et al., 2006). Along the arc, stratovolcanoes are commonly grouped into 2 or more edifices that are commonly aligned transverse-to-the-orogen. In Southern Chile, two main groups of volcano alignments are observed: ENE-WSW and NW-SE, both displaying a significant component of fissural volcanism. The ENE-trending systems are dominantly basaltic to andesitic (e.g. Callaqui–Copahue–Mandolehue; López-Escobar et al., 1995) and the NW–SE systems, frequently include dacitic to rhyolitic rocks (e.g., Puyehue-Cordón Caulle, Lara et al., 2004). 87 Figure 3-1 Traditional models for differential stress field dictating spatial distribution and geometry of feeder dikes within (a) stratovolcanoes and (b) scoria cones. (a) Nakamura's (1977) dike orientation method to estimate volcano-scale crustal stress field over stratovolcano life span (~1000 years). Monogenic flank vents (grey dots) and underlying radial fissure-shaped magma conduit (red dashed line) are indicated. (b) Selected Tibaldi (1995) and Corazzato and Tibaldi (2006) parameters (co-axial in the cartoon) to determine the magma-feeding fracture orientation (i.e. dike, dashed red line): (i) Coeval vents alignment, ii) azimuth of cone base maximum elongation axis, iii) azimuth of maximum crater elongation axis, and (iv) and crater-rim depressed points alignment. In Andean type volcanism, two fissural eruptive styles have been reported spatially associated with transverse-to-the-orogen fault systems; (1) co-seismic misoriented fissural volcanism (e.g. Cordón Caulle volcanic chain; Lara et al., 2006, 2004) and (2) inter-seismic optimally oriented fissural volcanism. (1) Large subduction earthquakes arising from interplate slip may result in elastic rebound towards the trench, triggering extensional co-seismic strain in the fore-arc and/or the intra-arc region (Lara et al., 2004; Arriagada et al., 2011; Aron et al., 2013; Pritchard et al., 2013). Tectonic relaxation within the arc may trigger short-lived tensional to transtensional tectonics promoting volcano subsidence, open fracturing and even fissure eruptions (e.g. Cordón Caulle eruption, Lara et al., 2004; 2010 Maule Earthquake related co- seismic fractures in the main Cordillera, Arriagada, et al., 2011; and Andean volcanoes subsidence, Pritchard et al., 2013); (2) The steady state of the overall inter-seismic ENE-WSW 88 crustal shortening within intra-arc regions enhances fissural volcanism parallel to the shortening axis. This evidence can be observed in volcanic complexes such as the Tatara-San Pedro-Pellado (e.g. Singer et al., 1997), the Callaqui–Copahue–Mandolehue (Melnick et al., 2006), and on the distribution of flank vents on stratovolcanoes (e.g. Llaima, Villarrica, Osorno; Cembrano and Lara, 2009; López-Escobar et al., 1995; Lara et al., 2012). 89 Figure 3--2 Tectonic setting of central-southern Andes. Convergence vector taken from Angermann et al. (1999). Chiloe sliver northward motion is taken from Wang et al. (2007). Triangles represent current volcanoes front organized in Southern Volcanic Zone (ZVS; northern, transitional, central and southern) and Austral Volcanic Zone (AVZ; López-Escobar et al., 1995). White lines indicate main crustal fault zones and dashed white lines transverse-to-the-orogen faults. Focal mechanisms taken from Lange et al., (2008) and http://www.globalcmt.org. Blue rectangle is an inset of Regional geological map of Fig. 3-3. Callaqui Stratovolcano is highlighted with a red triangle. 90 Regional geology and basement nature This study is focused on the Andean Principal Cordillera (38˚S), in a transitional region between Central (33ºS-38ºS) and Patagonian (38ºS-48ºS) Andes (Fig. 3-2, sensu Groeber, 1921). Significant changes take place in the basement architecture, intra-arc tectonics and volcanic nature over this region. The Central Andes are characterized by (1) a predominant non- partitioned setting evidenced by compressional tectonics; (2) the absence of a continuous surface expression of a seismically active intra-arc fault system; (3) a thick Meso-Cenozoic volcano- sedimentary cover resulting in a higher (3 km mean elevation) and wider (up to 800 km) Cordillera; and (4) broader volcanic arc with heterogeneous composition (Stern et al., 2007). The Patagonian Andes, in turn, exhibits (1) a partially partitioned deformation style, represented by (2) an active margin parallel intra-arc fault system, the LOFS (Cembrano et al., 1996); and (3) a lower (1.5 km mean elevation) and narrower (~300 km) Cordillera. Here, Quaternary volcanism is spatially and genetically linked to the LOFS (Fig. 3-2; Cembrano and Lara, 2009; Stern et al., 2007). Thus, there is a different causal link between arc volcanism and crustal deformation in the Patagonian Andes with respect to the Central Andes. In this transitional region, the ~90 km long and N~60 E trending Callaqui-Copahue- Mandolehue (CCM; Fig. 3-3) lineament has been claimed to be an inherited (Oligo-Miocene) transverse-to-the-orogen transfer zone (Radic, 2010). Clusters of numerous Quaternary eruptive centres have been emplaced along-strike the CCM, becoming the longest transverse-to-the- orogen volcanic chain in the Southern Volcanic Zone. In addition, the Callaqui-Copahue- Mandolehue chain decouples active deformation from the intra-arc strike-slip LOFS to the south and the back-arc Copahue–Antiñir transpressional fault system to the north (e.g. Folguera et al., 2004; Melnick et al., 2006). 91 During the Oligo-Miocene in Central-Patagonian Andes, continental rifting controlled and developed syn-tectonic volcano-sedimentary basins dominated by pyroclastic rocks of the Curamallin Fm. (Carpinelli, 2000; Jordan et al., 2001; Radic, 2010 and references therein). Between 36ºS and 39ºS the extensional basin system was compartmentalized at least in three diachronic NS aligned sub-basins with an asymmetric half-graben architecture (Radic, 2010). Radic (2010) states that the half-grabens had alternated polarities and were limited by transverse-to-the-basin accommodation zones striking NE or NW as a mechanical response of differential extension within the system (Rosendahl, 1987). Late Miocene tectonic inversion folded and thrusted the Curamallin basin rocks, constituting the basement architecture of many volcanic complexes of the Southern Volcanic Zone (CSVZ, 37ºS-41.5ºS) and the Transitional Southern Volcanic Zone (TSVZ, 34.5ºS-37ºS; López-Escobar et al., 1995). Pliocene to Quaternary transpressional deformation reactivated the older transverse faults, which serve as magma pathways for arc volcanism (e.g. Chillan, CCM and Lonquimay-Tolhuaca volcanic complexes; Radic, 2010). Studies on the Copahue Volcano and Caviahue caldera (Melnick et al., 2006) – in the centre of the CCM volcanic complex – revealed a heterogeneous strain pattern and long-lived volcanism related to three regional structural systems: the CCM tectonic-lineament, the transtensional LOFS horsetail (specifically the Lomín Fault) and the Copahue–Antiñir fault system. The latter defined as a right-lateral transpressive fault system (Folguera et al., 2004) constrains to the east the geometry and position of the Caviahue Caldera (Fig. 3-3). In addition, the NW-striking, transverse-to-the-orogen Biobio-Aluminé Fault System (BAFS), interacts at an oblique angle with the northern horsetail of the LOFS (Fig. 3-3; Potent, 2003; Rosenau et al., 2006, among others). The CCM volcanic morpho-structure in turn, is apparently cut by the BAFS at its south-western tip (i.e. south-western flank of the Callaqui 92 Volcano). Close to this area, the Mw 5.5 Ralco earthquake took place on 31 December 2006, at 17.7 km depth. The location and focal mechanism of the event (http://www.globalcmt.org) is probably imaging the Neogene activity and left-lateral kinematics of the BAFS (Fig. 3-3). The Callaqui Stratovolcano The Late-Pleistocene–Holocene Callaqui Stratovolcano (CSV) in Southern Chile (38ºS) represents a world-class candidate to gain insights into the nature of the interplay between upper crustal deformation and volcanic morphology in volcanic arcs with predominant strike-slip kinematics. A conspicuous N~60ºE trending, 16 km long ridge formed by sub-parallel dike swarms, volcanic fissures and aligned flank vents characterizes the overall volcano morphology (Moreno et al., 1984; Stern et al., 2007; Fig. 3-4). Based on its distinctive ENE morphological elongation, the CSV has been thought to be built on top of a pure tension gash in the basement (Rosenau et al., 2006). However, the new field and remote sensing data documented in this paper suggests a more complex structural history. The CSV is mid-size (ca. 110 km3) when compared to neighbour stratovolcanoes located in the Central SVZ (Völker et al., 2011). It was built on top of Early-Pleistocene volcanic layers of the Pacha Sequence, part of the Plio-Pleistocene Cola de Zorro Formation (Moreno et al., 1984; Fig. 3-4). The stratovolcano geology has been organized into four volcanic units, the oldest of which is 540 ± 140 ka old (K/Ar; Moreno et al., 1984) and hosts a number of dikes. Two recently obtained 40Ar–39Ar plateau ages of 146 ± 30 ka and 77 ± 28 ka in the basaltic groundmass (see Table 3-1) constrain the age of the unit 2, which forms the core of the present stratocone, only covered by postglacial lavas of units 3 and 4. 93 Partially ice-capped, the 3164-m-high basaltic–andesite Callaqui volcanic edifice includes well-preserved monogenetic pyroclastic cones, fissures, and lava flows, which constitute stratigraphic units 3 and 4 (Moreno et al., 1984; Fig. 3-4). Only few historical eruptions are confirmed (1751, 1864 and 1937) including a small phreatic ash emission noted in 1980 (Stern et al., 2007; Global Volcanism Program ). In turn, periods of intense fumarole activity and water-vapour emission are often observed near the main summit (http://www.ssd.noaa.gov/VAAC/vaac.html). Late Holocene debris avalanche deposits have been recognized in the north-western flank (Polanco & Naranjo, 2008), right below the steepest slopes and probably sourced in a scar open to the north-west. Recently, local seismic activity has been reported at the SE and WNW of the volcano edifice (http://www.sernageomin.cl). Mainly volcano-tectonic events indicate rock fracturing at shallow depth (4–10 km). Few long-period events have been reported during 2014 associated with internal fluid dynamics (http://www.sernageomin.cl). However, the most significant earthquake (Mw 5.5) in the region (southern Callaqui basement, Fig. 3-3) occurred on December the 31st of 2006 at 17,7 km depth. The latter displays a strike slip focal mechanism with a shortening axis oriented WSW–ENE (http://www.globalcmt.org). Morphological observation suggests that this event can be linked either to a NE lineament, which may be the surface expression of a NE–SW right-lateral strike slip subsidiary fault of the LOFS; or image a NW–SE left-lateral branch of BAFS (Fig. 3-3). Earthquake location is not precise enough to correlate it with one or the other fault. However, such focal mechanism is consistent with the overall ENE crustal shortening arising from oblique subduction. 94 Table 3-1 Summary of 40Ar/39Ar incremental heating experiments *GM = groundmass material free of phenocrysts. Preferred ages in bold. All uncertainties reported at 2s precision. **Samples analysed at Servicio Nacional de Geología y Minería, Chile. Calculated relative to 28.03 Ma Fish Canyon sanidine (Renne et al., 1994). 95 Figure 3-3 Regional geological map of Andean intra-arc zone between latitudes 37º20’S and 38º40’ (compiled from Pérez-Flores et al., 2016; Sernageomin, 2003; and data of this study). Main intra-arc fold and faults (inferred faults included) are indicated to illustrate the volcanic basement architecture (Stratovolcanoes summits in green triangles). Long-term finite strain diagrams in the bottom, result from fault plane kinematic analysis (striae data inversion) from depicted fault zones in the area (this study and taken from Pérez-Flores et al., 2016). Geothermal activity is indicated by hot springs with outflow temperatures higher than 50 ºC (light-blue dots). Crustal seismicity is indicated by compilation of shallow earthquakes (<35 km depth; yellow dots) obtained from international catalogs (http://www.globalcmt.org). Both presented focal mechanisms (Ralco and Lonquimay) are coherent with ENE–WSW shortening. Minor eruptive vents are highlighted in pink. A yellow dashed line surrounds the Callaqui–Copahue–Mandolehue transfer zone. Blue rectangle is an inset of Fig. 3-4. 96 3.3 Methodology A combination of structural field mapping and remote sensing analysis constitutes the basis of this work. All field measurements were taken with a Freiberger compass with a magnetic deviation of 4.8º. A high-resolution multispectral Quickbird image (0.6 m/pixel) was used to perform spatial and geometrical analysis of the Callaqui dike swarm and eruptive vents. Measuring dikes Pleistocene intravolcanic dikes were systematically mapped using both, fieldwork reconnaissance and satellite images interpretation. Strike, dip and thickness data were collected for a total of 78 dikes recognized in the field (Fig. 3-5) within an area of 10 by 2.5 km and altitudes between 1800 and 2200 [m a.s.l.] (Fig. 3-6). Mapping was carried out on Callaqui's south-western (Fig. 3-5b and d) and north-eastern flanks, where better outcrops were found. A detailed field mapping was conducted at the Quillaycahue Valley headwaters in order to recognize better the nature of geometrical variations of dikes (Fig. 3-5). Commonly dikes on the outer borough of the Callaqui ridge show along-strike thickness and strike variations (e.g. Fig. 3-5e and g). Thus, several measurements along-strike were needed to get a representative average dike thickness and strike. Similarly, geometrical variations were taken in account, measuring features such as step-overs, inflexions (bending) en echelon arrays and irregular dike tips. A suit of dozen field-mapped dikes identifiable on the Quickbird image (e.g. dikes of Fig. 3-5a, b and f), where used as an initial approach in order to identify and map dikes remotely from the satellite image. A combination of texture, colour and geometry patterns result in a basic criterion, which allowed the remote sensing of a total of 129 dikes. 97 Field mapped dikes were weighted by their thickness, whereas satellite-image detected dikes were weighted by using their outcropping lengths. Length of non-continuous dike outcrops (dikes partially eroded or covered by later lava flows), were counted as the sum of the outcropping segments. A weight factor was obtained for each measured dike by arbitrary breaks based on natural clustering (thickness or length), observed from field and remote sensing data (Table 3-2). Dike geometry database – in Stereonet 6.3.3 – was constructed by adding singular dike orientations the times the weight-factor is. Table 3-2 Weight factor for field and remote detected dikes Morphometry of minor eruptive vents Analyses of cone morphologies allow for indirectly identifying the orientation of the underlying feeder dike of cones and fissures (Tibaldi, 1995). A morphometric survey was taken from minor eruptive vents (MEV) considering Tibaldi (1995) and Corazzato and Tibaldi (2006) 98 parameters that indicate the geometry and kinematics of the underlying magma-feeding fracture (Fig. 3-1b). Used parameters are indicated in increasing reliability order, as demonstrated by analogues models varying underlying topography slope (Corazzato and Tibaldi, 2006; Tibaldi, 1995): i) alignment of the centre points of coeval vents; ii) azimuth of maximum crater and/or iii) cone base elongation axes (when possible); and, iv) crater-rim depressed points alignment (in general at opposite sides of the crater rim). Additionally, we calculated the degree of ellipticity expressed as 𝑔" = 𝐷%&'⁄𝐷%() (1) where Dmin and Dmax are the minimum and maximum diameters, respectively. Values of ge express a mean of relative differential strain, ranging from 1 (cylindrical cone construct under uniform strain pattern) to 0, indicating an extreme differential strain field (ideal limitless fissure length). To discard a significant gravity influence during the cone and fissure construction process, measurements were considered valid only for vents emplaced above a gentle-slope topography (i.e. slope smaller than 10º) and showing quasi-continuous crater rims. Hence, only the late Pleistocene MEV were included, rejecting those with evident gravitational effects (partially collapsed along major slope direction) and intense glacial erosion. Numerous aligned post-glacial MEV reflect fairly well the local strain state during it eruption phase. From these data, finite strain (~stress) field orientation was obtained for the Callaqui volcano Late-Pleistocene to Holocene evolution, and finally discussed within the tectonic setting. 99 Coeval cone alignment, maximum crater elongation axes azimuth and depressed points azimuth were used to calculate principal morphometric orientations, whereas crater ellipticity was considered as a reference value for differential strain during eruption. We proceeded in a similar way for the dike geometrical analyses; morphometric orientations were the basic input data in Stereonet application for morphometric analysis. 100 Figure 3-3 Callaqui Stratovolcano map (updated from Moreno et al., 1984). Intravolcanic dike swarm and minor eruptive vents distribution are improved in this study by field mapping and remote sensing data (resolution of Quickbird image is 0.6 m/pixel). Observed morphological scarps are presented 101 3.4 Data analyses and results Dike geometry and spatial distribution Callaqui intravolcanic mafic dikes are organized into a 13-km-long and 2-km-wide, sub- parallel dike swarm, trending ENE and cropping-out between 1800 and 2200 m a.s.l. Because dike compositions are restricted to basalts and andesites with no evident compositional organization with respect to the overall structure, we collectively refer to them as mafic dikes. Exhumed dikes cross cut volcanic products of early Callaqui units (1 and 2) and were covered by late Callaqui sequences (Fig. 3-4) and debris deposits, constraining a relative age range between 146 ± 30 ka and 77 ± 28 ka. The general ENE–WSW swarm strike contributes notoriously in the building and preservation of the ENE–WSW oriented Callaqui ridge. Along this morphology, differential erosion enhanced dike exhumation constituting the main mapped area (e.g. see in Fig. 3-5b WSW portion of the Callaqui ridge, view to the ENE). However, the sub-parallel Callaqui dike swarm exhibits considerable along-strike structural differences. First, dikes closer to the main summit and central portion of the Callaqui ridge, are wider and maintain a constant strike, when comparing to those outcropping in the WSW or ENE tips of the Callaqui ridge (Fig. 3-6). A second main difference is based on the average strike of dikes observed for different geographical areas, classifying them in the following domains: a) Central ridge domain: This area is basically consistent with the main ridge morphology. Here, mafic dikes dominantly striking between N40º–60ºE (black stripes, petals and planes, in map and diagrams of Fig. 3-6, respectively), presenting the major exposed lengths and widths; this domain consists of hundreds of dikes, which contribute to the first-order morphology of the 102 Callaqui ridge (Fig. 3-5a and b). This zone is partially covered by snow and ice nearby the summit, thus dike measurements are restricted to the middle and lower portions of the ridge, between 1800 and 2200 [m a.s.l.] and an average slope of 8.5º. In these domains were found the thickest dikes of the system, which reached up to 20m thick (e.g. dike in the background of Fig. 3-5a). Dikes of this domain show roughly constant strikes, and as they are bigger, they can be followed for hundred meters along their strikes (Fig. 3-5a). b) Distal domains: these consist of two distal (“away from the summit”) domains described, one on the ENE, another on the WSW termination of the Callaqui ridge (Figs. 3-5d, e, f and 3-6). Since individual dikes start bending their strikes when moving away from the central ridge domain, the boundary between both domains is transitional (white dashed line in Fig. 3-6). Distal dikes in general show a complex geometrical nature with predominant strikes in the N60º to N88ºE range and significant along-strike thickness changes. Systematic clockwise strike-variations were documented from both field and remote sensing analysis (see stereoplots on Fig. 3-6). Progressive along-strike thickness narrowing is also observed as moving away from the central ridge domain, until they disappear at distances of more than 7 km from the main volcano summit. Fig. 3-5a and c were taken on the same dike in the north- eastern flank (panel 3-5a is indicated with an arrow in Fig. 3-6b), but with 400 m of separation along an ENE trend. In picture 3-5a the dike thickness is of 16 m and in 3-5b, is ~3.5 m. Concordantly, when comparing with the central ridge domain, thinner and fewer dikes are present per unit area. Generally, dikes are exposed along irregular topography with an average 10º slope and get covered by colluvium and forest from ~1000 [m a.s.l.] downhill. 103 Dikes in the distal domain show significant geometry variations, such as sinusoid forms, step- overs and en echelon arrays (e.g. of en echelon dike at Quillaycahue valley headwaters Fig. 3-5e). Field data from the Quillaycahue valley (green dot in Fig. 3-6) are presented in the planes and pole contour diagram in Fig. 3-6. For sinuous dikes several measures were made, thus the corresponding diagram shows some ~NS strikes. In general dikes are dipping >70º to either the SE or the NW. Combined field and remote sensing data analyses of the whole Callaqui dike swarm orientations (global rose diagram, Fig. 3-6) yields to a general dike-swarm-geometry with two extensional tail cracks at both distal domains (i.e. similar to a wing crack damage zone; Kim et al., 2004). Wing cracks trend N60º–88ºE, and are connected through the N45º–60ºE trending central ridge domain (Fig. 3-6). Morphometry of minor eruptive events (MEV) Flank cones and fissures (Fig. 3-7) were described in order to infer the orientation and kinematics of underlying feeder dikes by using Tibaldi (1995) and Corazzato and Tibaldi (2006) criteria. Based on boundary and dike-cross-cutting relationship, as well as glacial erosion and a single 40Ar–39Ar age of a related lava (~77 ± 28 ka), it is possible to constrain the age of the analysed MEV from Late-Pleistocene to Holocene. Thus morphometric analyses are separated into two diachronic systems: a) glacial-eroded vents, and b) post-glacial vents. a) Glacial-eroded vents consist of ten elliptical craters mostly preserved and aligned along the south-western ridge (Fig. 3-7a), with an average N60ºE direction. Only eight out of ten craters 104 were taken into account, considering substratum slopes smaller than 10º and the state of erosion (as basic criteria, an 80% of crater rim preservation was required for the vent to be taken into account for further analysis). Crater dimensions vary from hundred to seven hundred meters wide (Table 3-3). Average grade of ellipticity is 0.56, with a maximum value equals to 0.71 and a minimum of 0.37. A single coeval vent alignment, striking N61ºE, was measured for craters “a, b, c, d” and “e” (Fig. 3-8). With the exception of crater “f”, vents maximum crater–diameter axes trend from N61ºE to N79ºE. Crater-rim depressed points alignment was only possible to be measured for craters “a, b” and “c”, whose orientations range from N66ºE to N71ºE. The ca. 12º clockwise obliquity between coeval vents alignment and directions of maximum crater elongation and crater-rim depressed points, may define an en echelon array displayed by craters “a, b, c” and “d” (Fig. 3-8A). Because maximum elongation axes of the eroded vents are not parallel to the overall vent alignment, a possible right lateral displacement component adds to the predominant NW–SE extension direction. If this is the case, feeder dikes propagated under bulk dextral transtension and not under pure extension. This is also in agreement with the overall dike geometry described before. However, this result could be ambiguous when considering the influence of erosion on the craters. Random errors may arise from the measurement of elongation and crater-rim depressed point trends. b) Post-glacial eruptive vents are related to unit 4 (last eruptive phase) of the Callaqui volcano, consisting of scoria cones and andesitic–basaltic lava flows (Fig. 3-7b and c). Cone and fissure structures are well preserved, best represented by: i) two parallel fissures in the south-western ridge and, ii) an isolated talweg scoria cone with six associated lava emission centres, in the north-eastern distal zone of the edifice (coincident with the ENE distal domain of the dike swarm). 105 i) Two parallel-juxtaposed N56ºE striking eruptive fissures (upper and lower) in the WSW ridge are composed of a dozen of aligned scoria cones. The upper fissure consists of a 586 [m] long N56ºE-striking feature, hosting elongated craters -less than 30[m] in diameter- which are nested and aligned in a bottoniere array (Fig. 3-7b). This indicates distinct contemporaneous eruptions along the same eruptive fracture until it gets sutured (Ollier, 1988). Distances among coeval vents are in the order of ten to twenty meters. The lower fissure, in turn, significantly more eroded, consists of three aligned vents. Both parallel fissures breach out towards the south–southeast in their central portion, where drainage crosscuts the eruptive structures (Fig. 3-8B). ii) Beside the NE ridge, on the WNW oriented Quillaycahue glacial-valley, a scoria cone is emplaced on top of the upper section of the talweg (Fig. 3-7c). Pyroclastic deposits and lava flows are down-streams filling the Quillaycahue valley (Fig. 3-8C). The Quillaycahue cone maximum basal diameter is ~400 [m] long, whereas its main elongated crater is 140 [m] long (Table 3-3). The maximum diameter azimuth (N75ºE) is sub-parallel to the measured cone base diameter (N72ºE), as listed in Table 3-3. Furthermore, the crater rim depressed points azimuths, including the Arrayán fissure, are consistently aligned, striking N75ºE (Fig. 3-8C). Downstream the Quillaycahue glacial valley, six lava emission vents – here named Boldo, Litre, Peumo, Pehuén, Lingue and Raulí – are aligned with the Quillaycahue cone along a WNW direction, parallel to the valley axis. Conversely to the observed coeval vents alignment and main topographic control (glacial incision), strong morphometric data of the well-preserved Quillaycahue scoria cone (i.e. co-axial ENE trending: azimuth of crater maximum elongation, 106 crater-rim depressed points alignment and azimuth cone-sale maximum elongation; Fig. 3-8C and Table 3-3), suggests a more complex building history including both tectonic (ENE oriented feeder dike underlying the Quillaycahue cone) and topographic controls on the Holocene monogenetic eruption. 107 108 Figure 3-4 Field pictures of mapped dikes. Green arrows in the Quickbird image in the centre of the figure point to the position and direction of each photograph shot. (a) Picture looking SW along a basaltic dike striking N60ºE with a mean thickness of 16 m (black arrow points to a person as scale reference) in the north-eastern Callaqui ridge. (b) View to the ENE of a group of 6 parallel dikes striking N60ºE, in the southwestern Callaqui ridge. Callaqui summit and summit-fumarole in the background. (c) Same dike as “a” but in a further ENE position where strikes N62ºE with a smaller average thickness. Several normal-to-the-dike-strike columnar joints are visible. (d) View to the SW of a group of 7 dikes dipping moderately to the SE at Callaqui's southwestern ridge. (e) View to the south of a basaltic dike (N80ºE/85ºS) in Callaqui north-eastern distal zone. Dikes exhibit an en echelon array (pointed with a yellow arrow) consistent with right-lateral shearing. Backpack in the right-lower corner as scale reference. (f) View to the south of a N78ºE striking dike in Callaqui's southwestern distal zone. (g) View to the NE of a group of ENE to E–W striking dikes. Cross cutting relationships allow differentiating three generation of dikes intruding a red tuff layer. Dikes shows local thickness variation (e.g. jog in the centre of the picture), en echelon structure (pointed with yellow arrow), and parallelism and splay of two dikes of different generation (left side of the picture) Figure 3-5 Quickbird image (0.6 m/pixel) with the distribution of the Callaqui parallel dike swarm. Dikes and groups of dikes (thickness not to scale) are represented as black stripes in central ridge domain, and as red stripes in distal domains (roughly enclosed by white dashed line). (a) Dike distribution in southwestern ridge. Yellow arrow indicates position and direction where picture 5b and 7a were taken. (b) Dike distribution in north-eastern ridge in close spatial relation with visible minor eruptive vents. Up-left: rose diagram of total weighted dikes, including south-western and north-eastern ridges. Black petals represent average trend of dikes from the central ridge domain. Red petals represent dikes of distal domains. Up-right planes and pole-plane contour of field measured dikes in Quillaycahue valley headwaters (green dot in ‘b’). Strikes ranging from NS to E–W are consistent with dike sinuosity and strike variations in this distal domain 109 Table 3-3 Input data for the morphometric analysis 3.5 Discussion The fissural Callaqui Stratovolcano, active at least since ca. 500 ka years BP, consists of four volcanic stages (Moreno et al., 1984). Whereas the early Callaqui stage is less-well recognized due to intense glacial erosion and tapering of younger units, the last three output stages reveal an outstanding first-order tectonic control during the Late-Pleistocene–Holocene volcano- building process (e.g. Nakamura, 1977; Moreno et al., 1984; Rosenau et al., 2006; Melnick et al., 2006; Radic, 2010). Pleistocene–Holocene magma migration within the upper crust, rooted into specific ENE oriented pathways, was likely enhanced by a combined effect of long-lived activity of optimally-oriented inherited basement faults – i.e. CCM volcanic morpho-structure (Melnick et al., 2006; Radic, 2010), Quaternary intra-arc NE–SW shortening (Cembrano and Lara, 2009; Lavenu and Cembrano, 1999; Rosenau et al., 2006), and the permanent building up of an elongated volcano anatomy. The latter, constructed by a mass-balance among diking, fissural 110 volcanism and steep flank erosion, increases the differential strain field and promotes further parallel diking and volcanic out-put. These factors result in a positive feedback loop of fracture opening, dike propagation and diachronic elongated volcano construction/erosion. However, geometrical and morphometric data examined in this research document more complex volcano-building evolution based on the following observations. • Dike swarm tail-crack geometry, plus the en echelon nature of individual dike segments (in distal domains, i.e. Quillaycahue valley) and the apparent en echelon array of glacial-eroded MEV, documents a coeval right lateral strike slip component during the predominant NNW– SSE-trending extension, and consequent dike propagation and eruption along the Callaqui edifice. Two different dike orientations, systematically organized into two domains, describe an “S” shaped array of the dike swarm. Within the central ridge domain, dikes strike N50º– 60ºE, whereas at distal domains, dikes strike in the range of N60º–88ºE; at these domains dikes bend progressively into an orientation free of volcano load effect, which we interpret as close- to the regional shortening axis. Taking in to account scale magnitudes, the overall “S” geometry of the Callaqui dike swarm, could be compared with the “Y” shaped distribution of fissures documented at both edges of the elongated Mount Cameroon Volcano (Kervyn et al., 2014). That volcano, considered one of the most voluminous continental volcanoes on Earth, is characterized by tectonic faulting consistent with the elongation of Mount Cameroon along the Cameroon fracture line (Déruelle et al., 1987). The latter suggests a similar structural setting to the one of the Callaqui Stratovolcano. Nevertheless, the main difference between the “S” and “Y” distribution of dikes/fissures is accounted for by the ridge-parallel shear component here documented for the Callaqui Volcano. 111 • Dikes located at the central ridge domain follow a similar orientation than the underlying long-lived CCM discontinuity, which has been claimed to be a regional fault spatially and genetically linked with the locus of the Plio-Quaternary volcanism on top (Callaqui, Copahue, Caviahue Caldera and Mandolehue volcanic structures). Following published analogous models (Mathieu et al., 2011; Van Wyk De Vries & Merle, 1998) we infer that the high recurrence of subvertical dikes striking parallel to the main N50º–60ºE ridge and CCM orientation is reinforced by a positive feedback loop of eruption following extension, the latter promoting more volcanic output and further normal-to-the-elongation flank failure. Volcanism would be originally located along pre-existing fault zones (i.e. CCM) favourable oriented for right-lateral transtensional reopening during long-term continental deformation (i.e. Quaternary inter-seismic periods). • On the other hand, in distal zones, dikes are organized oblique-to-the-CCM and main Callaqui ridge. A progressive clockwise bending of dikes is observed along-strike, as going away from the central ridge domain. We infer that in distal zones the edifice load is negligible, causing dikes to orient close-to to the local shortening axis, whose orientation may be determined from these results. Thus, distal dike orientations may represent the local-scale and short-term crustal 112 Figure 3-6Field pictures of some minor eruptive vents named in Table 3-3 and in Fig. 3-8. (a) View to the east–northeast of southwestern Callaqui ridge. Glacial eroded craters “d” and “f” are visible as well as dikes cross cutting crater “d”. Postglacial fissure “i”-with nested vents-is pointed out by the arrows. (b) View to the east of the “Upper fissure i” with six nested craters. (c) Oblique aerial view to the east of the Quillaycahue cone and downslope lava emission centres (Arrayán, Boldo and Peumo MEV are visible) 113 shortening axis. If this is correct, the shortening direction would be slightly different from the one obtained by Lavenu and Cembrano (1999) for this portion of Southern Andes. Fig. 3-9 synthetizes the kinematic control on fissural volcanism type at Callaqui Volcano. Figure 3-7 Map with the distribution of minor eruptive vents. (A) Glacial eroded vents. Rose diagrams of morphometric parameters considered for this group of craters are shown. Yellow arrow indicates position and direction where picture 5b was taken; (B) Southwestern fissures “i” and “j”; fissure “i” is characterized by several nested vents indicated with green arrows; and (C) Quillaycahue scoria cone and lava emission vents. Coeval vents alignment considering the scoria cone and downhill lava-flow pits. Maximum crater elongation and crater-rim depressed points alignment rose diagrams are measured for Quillaycahue cone only. Yellow arrow indicates position and direction where picture 7a was taken 114 Figure 3-8 Kinematic model for the overall right-lateral transtensional setting driving the Callaqui stratovolcano elongated anatomy. At the south-western ridge, topographically controlled SE oriented breaching of dominantly N60ºE trending fissures, could be wrongly interpreted as NW–SE oriented feeder dikes. In fact, morphometric evidence indicates N60ºE oriented feeder dikes. Similarly, the monogenetic Quillaycahue scoria cone in east north-eastern area and associated WNW–ESE aligned lava emission vents would primarily suggest a WNW–ESE oriented feeder dike. However, strong morphometric evidence at Quillaycahue edifice indicates an ENE–WSW structural controlled feeder dike (Fig. 3-8C and Table 3-3). This suggests that the glacial valley incision play a significant role in facilitating the interception of concealed ENE–oriented-dike with the surface, lacking of a local dike reorientation due the topographic/gravity differential load. Our work suggests that the Andean building process is strongly dependent on the volcanic output gathered and accumulated during the inter-seismic deformation phase of the 115 subduction seismic cycle. Long-lived favourable oriented crustal discontinuities, such as ENE striking transverse-to-the-orogen fault systems, are expected to enhance magma migration within transtensional domains (e.g. Brogi et al., 2010). The right lateral transtensional kinematics here reported for the Callaqui Stratovolcano, constrains the volcanic output along predominant ENE–WSW oriented eruptive fissures. This feature is mechanically consistent with the symmetric transtensional horsetail-like tip of the LOFS at this region (Fig. 3-3; Pérez-Flores et al., 2016; Potent, 2003). We propose here that Plio-quaternary reactivation of optimally-oriented (i.e. close to parallel to the main horizontal shortening axis) inherited basement fault would enhance the development of discrete transtensional domains across the first-order intra-arc fault system (i.e. LOFS). These domains could evolve into linking damage zones (Kim et al., 2004), for instance the CCM transfer zone, claimed to decouple active deformation from the intra-arc strike-slip LOFS to the back-arc Copahue–Antiñir fault system (Folguera et al., 2004; Melnick et al., 2006). Along this transtensional domain, successive Pleistocene–Holocene eruptions enhanced by transtensional kinematics, contributed to the overall ENE–WSW oriented CCM volcanic chain building process, and in particular, to the elongated architecture of the Callaqui Volcano. This may reflect the combined effect of regional/far field tectonic stresses and orientation of pre-existing fault systems in the control of dike propagation and eruption style on a particular Stratovolcano. In the Southern Andes, a similar geological setting can be observed at (a) the ENE–WSW oriented Tatara-San Pedro-Pellado volcanic chain characterized by abundant ENE oriented dikes and MEV (36ºS, Singer et al.,1997 and Hildreth et al., 2010); and (b), at the Llaima – Sierra Nevada volcanic chain (39ºS), where both MEV aligned in a ENE trend at Llaima, and ENE–WSW oriented dikes at the ridge-like Sierra Nevada Volcano are observed (Bouvet de Maisonneuve et al., 2012; Naranjo & Moreno, 2005; Sernageomin, 2003; See Fig. 3-3). 116 For reduced geological time spans (e.g. <0.5 Ma) no significant block rotations are expected, therefore stress and strain axes can safely be assumed to be approximately coaxial and reciprocal (e.g. Weijermars, 1991). Thus, finite strain inferred from Quaternary dike geometry and/or cone morphometry can be a first order approach to understand the local stress state during magma propagation, shallow magma emplacement and eruption. Regional stress models for dike propagation and emplacement best fit within volcanic environments absent of prominent relief, such as rift zones (Acocella & Neri, 2009; Mathieu et al., 2011; Pinel & Jaupart, 2003). However, within tall stratovolcanoes the propagation of feeder dikes -intravolcanic dikes, i.e. emplaced within older units of the same volcanic system- also depends on the localized effect of the size (load) and shape (topography) of the edifice (e.g., Acocella and Neri, 2009; Van Wyk De Vries and Merle, 1998). For example, in volcanic arcs with strike-slip kinematics, volcano-load combines with the tectonic stress field resulting in local stress reorientation, as also documented by geometry and kinematics of local faults in analogue models (Van Wyk De Vries and Merle, 1998; Mathieu et al., 2011; and reference therein). Stress field modification may influence the nature of strike-slip faults and shear zones, promoting an increase in the extensional component of pre-existent faults and strengthen the development of extensional pull-apart and/or releasing bend structures. According to Van Wyk de Vries and Merle (1998) volcano-scale stress field reorientation due to volcanic built-up process and progressive edifice loading may turn into a feedback loop, promoting more extension and consequently more magma extrusion. The effect of topography and gravity loading on shallow magma intrusion and the development of preferential volcanic system building orientation had also been documented for other volcanic environments (e.g. Kilauea volcano in Hawaii; Fiske & Jackson, 1972). These authors conclude that the predominant stress field controls the development of rift zones. At the Callaqui 117 Stratovolcano, strain geometry and kinematics suggest that the effect of the elongated volcanic load drives successive shallow emplacement and eruptions through similar to constant oriented pathways. Additionally, erosion effect along steeper flanks would perpetuate or enhance the elongated anatomy of the ridge-like volcano, adding an extra extensional component within the volcano life span. Moreover, the internal structure of long-lived stratovolcanoes (i.e. 100 ka–1 Ma built over successive and overprinting intravolcanic diking events) would be a complex network of dikes and eruptive conduits, defining the first order architectural frame of the edifice. We hypothesize that swarms of successive subparallel dikes are initially dominated by regional differential horizontal stress (Lister & Kerr, 1991) and/or pre-existing fractures e favours the construction of elongated stratovolcanoes, rather than cone shaped volcanoes. The latter is the case of shield volcanoes or conical stratovolcanoes, where stresses focused above the centre of shallow magma reservoirs promote the development of a central conduit/vent systems (Pinel & Jaupart, 2003). As stated by previous studies, elongated volcanoes develop steeper and unstable slopes in normal-to-the-volcano-elongation flanks, promoting lateral collapse and volcanic avalanches in the direction on the least horizontal stress (Lagmay & Valdivia, 2006; Mathieu et al., 2011; Tibaldi, 2001). This erosional pattern may result in volcano scale de-confinement and local attenuation of the minimum principal horizontal strain axis (Sh min; i.e. increase of local differential stresses). Finally, avalanche hazard along the steep flanks oriented normal-to-the-volcano- elongation, are a likely consequence of the interplay between volcanic and tectonic edifice destabilization. Field data is consistent in documenting the overall ENE–WSW internal and external architecture of the Callaqui Stratovolcano. As indicated in Fig. 3-4, main scarps observed 118 along the Callaqui ridge, also follow this preferential trend, facing both normal-to-the-ridge slopes. Thus, preferential direction of erosion, collapse and avalanche flow (i.e. normal to the ridge trend) reinforces the elongated shape of the volcano edifice. This general observation is consistent with the direction and nature of the volcanic avalanche reported by Polanco and Naranjo (2008) for the northern flank of the Callaqui volcano. However, further analysis has to be conducted to minimize natural hazards over populated localities and infrastructure lying on NW and SE flanks and basement of the Callaqui volcano (Ralco and Pangue dams). 3.6 Conclusions 1. Dikes represent the position and spatial distribution of magmas intruding the crust, whose orientation is the result of complex mechanisms predominantly driven by the regional and local stress field and the orientation of pre-existent weakness zones (inherited faults). 2. Favourably oriented inherited fault zones and the operating stress field during the time of magmatism control transverse-to-the-orogen volcanic systems. 3. Significant differential stress fields, over favourable oriented transverse-to-the-orogen magma- pathways, are key factors to define transtensional tectonics dictating the overall feeder dikes geometry and distribution, and consequent volcano edifice up-building anatomy. Parallel andesitic–basaltic dike swarms constitute the fundamental architecture of elongated stratovolcanoes. 4. Tail crack geometry of Callaqui's parallel dike swarm and the en echelon array of individual Pleistocene dikes in distal domains are evidence of a right lateral component accompanying the predominant extension. 119 5. Progressive-forming elongated volcanoes may increasingly lead to normal-to-the-volcano- elongation steeper flanks, triggering avalanches and flank erosion. Consequently, this may reinforce along-strike eruptions. 6. Local scale gravitational effects along misoriented glacial or fluvial incisions, control successive oblique-to-the-incision dike surface-interceptions and vent topography-controlled breaching along unexpected directions, not necessarily following their underlying feeder dikes orientations. 7. A geologically stable differential stress field facilitates parallel dike swarm propagation and extrusion through long-lived juxtaposed parallel pathways, contributing to long-term morpho- structural stability of elongated volcano-architectures. 8. A positive feedback loop results among long-term inter-seismic shortening, transtension within favourable oriented inherited faults, parallel dike propagation, diachronic volcanic out-put, elongated volcano anatomy construction, and lateral steeper flank collapse (avalanches). The latter may result in local de-confinement, perturbing the balance between volcanic output and erosion. This may be reflected in a local increment of the extensional component and further dike propagation and volcanic output. 3.7 Acknowledgements The authors are grateful for the funding provided by FONDECYT Projects 1060187 and 1141139 to J.C. and L.L. and by FONDAP Project 15090013 Centro de Excelencia en Geotermia de los Andes (CEGA). Gerd Sielfeld acknowledges the financial support by CONICYT doctoral grants. We also thank Georg Zellmer, for his constructive discussion and overall contribution and Gloria Arancibia, for her feedback and revision of an early version of this manuscript. Authors are very grateful at the comments, corrections and suggestions made by the reviewers Dr. Andrea Brogi and Dr. Joël Ruch, and editors of this special volume Dr. Asfawossen, Dr. Min-Te Chen and Dr. 120 Daniela Kroehling; they significantly improved the quality of the manuscript. Finally, we also thank Pablo Azúa, Natalia Varela and Pedro Guzman for their assistance during the field campaigns, and Pamela Pérez-Flores and Ashley Stanton-Yonge for discussing and reviewing the content of this paper. 121 4 Oblique-slip tectonics and geofluid flow: a case study from south-central Andes. 4.1 Introduction and problem statement In active continental margins, arc magmatism is structurally connected with intra-arc crustal deformation. The causal relationship among frictional faulting, permeability evolution and crustal fluid flow has been debated through diverse perspectives including: transdisciplinary research on the existing interplays between volcanism and shallow-crust active tectonics (e.g. Valerio Acocella et al., 2018; Cardona et al., 2018; José Cembrano & Lara, 2009a; Lupi & Miller, 2014; Payne et al., 1996; Richard H. Sibson, 1989; Singer et al., 2014) and exhaustive-multiscale examination of paleo-permeability in exhumed intra-arc fault systems (e.g. Gomila et al., 2016; Jensen et al., 2011b; Richard H. Sibson, 1996; Veloso et al., 2015). It is accepted that fault-fracture meshes and collateral damage develop differential permeability structure, where the major permeability vector is parallel to the intermediate stress (s2) direction (e.g. Rowland & Sibson, 2004b; Richard H. Sibson, 1996; Wibberley et al., 2008). It is also accepted that for diverse natural fault-systems, strain partitioning might distribute heterogeneous frictional damage, materialized in diverse fault geometries and slip sense, configuring the dynamics for crustal fluids plumbing system (e.g. Pérez-Flores et al., 2016; Roquer et al., 2017; Rowland et al., 2012; Richard H. Sibson, 1996).Thus, for volcanic provinces ultimately hosting economic geothermal resources, the permeability structure is locally constrained by discrete strain compartmentalization pattern, as a direct effect of the local stress-strain field (e.g. Taupo Volcanic Zone; Rowland et al., 2012; Rowland and Sibson, 2004). Such tectonic constrains are expected to influence the evolution of 122 Andean Southern Volcanic Zone (SVZ), which has been claimed to hide over 3350 MW economically feasible geothermal resources (Lahsen et al., 2010; Sanchez-Alfaro et al., 2015; Soffia & Clavero, 2010). Recent published works on the eruptive threat of Laguna del Maule Volcanic Field (LMVF) at the Transitional Southern Volcanic Zone (TSVZ; Fig. 4-1) shed light on the active deformation pattern in the region (Cardona et al., 2018; Feigl et al., 2014; Fourmier et al., 2010; Singer et al., 2014). For instance, exceptionally high rate of surface deformation at LMVF has been measured from interferometric analysis of synthetic aperture radar (InSAR; Fournier et al., 2010; Feigl et al., 2014). These authors found uplift rates according to accelerated inflation. In particular, Feigl et al. (2014) reported uplift rates exceeding 280 mm/yr between 2007 and 2012. On the other hand, Cardona et al. (2018) determined several crustal seismic clusters in the arc region between 35.8˚S-36.2˚S latitudes. For instance, they found that in the southern-east corner of LMVF, the NE-striking Troncoso Fault -earlier recognized by Muñoz and Niemeyer (1984)- showed high rate of small to moderate magnitude seismicity consistent with overall dextral strike-slip faulting. Simultaneous, WNW striking faulting with sinistral kinematics, has been defined by seismic clustering along the Laguna Fea morphotectonic lineament (Cardona et al., 2018; Fig. 4-2). Both active faulting processes, are in kinematic accordance with the far-field NE-striking shortening and NW-striking extension. The strongest reported crustal earthquake in the region is the June 6th 2012 Mw 6.2 event on the N10˚E striking Melado Fault (Fig. 4-2; Cardona et al., 2018; gCMT, http://www.gcmt.org), tens of kilometres west of the LVMF. The nature of the faulting process is constrained by hundreds of aftershocks organized as a ~N10˚E striking cluster (Cardona et al., 2018) and by a dextral strike-slip moment tensor reported by NEIC (National Earthquake Information Centre, USGS) and GCMT catalogs. The latter consist of a NS-, and an EW-striking 123 vertical nodal planes with NE oriented compressional axis (Fig. 4-1 and 4-2). So, the Melado Fault geometry and kinematics is defined by the aftershock distribution and the ~NS nodal plane of the GCMT moment tensor. Because of the relative strong magnitude of this event, and it’s kinematic coherence with other Mw >5 crustal events in the region (e.g. Mw 6.5 Teno on 28/08/2004 and Mw 5.6 El Fierro on 11/06/1980 earthquakes; gCMT; Fig. 4-1 and 4-2), the NE- oriented compressional axis is expected to represent the current instantaneous s1 orientation within this segment of south-central Andes intra-arc. The paleostress regime, in turn, encircling southern Andean Pliocene to Holocene arc volcanism is not well constrained, nor detailed structural mapping of the volcanic front basement. Mountain regions, such as the Andes, are built upon an anisotropic geological and structural grain through polyphasic deformational stages and thus record heterogeneous fault- slip data (Perez-flores et al., 2016; Piquer et al., 2016; Sielfeld et al., submitted, 2016; Veloso et al., 2015). In this study, we investigate the evolution and role of successive stress states in the development of crustal fault networks in consistency with arc volcanism and hydrothermal activity. We tackled this problem by describing the basement and intra-volcanic architecture of the Tatara-San Pedro-Pellado Volcanic complex (TSPPC) in south-Central Andes (36˚S). Hitherto, exhaustive geological studies have been undertaken in the TSPPC, mostly examining the geochemical signature, the volcanic stratigraphy, and the paleomagnetic and geochronological evolution of the magmatic system, as well as the erosional and exhumation processes of the volcanic complex and its basement (Davidson et al., 1988; Dungan et al., 2001; Frey et al., 1984; Nelson et al., 1999; Singer et al., 1997b; among others). However, very little has been documented on the tectonic evolution of this conspicuous composite volcanic complex. 124 We here describe the architecture of the TSPPC basement and the distribution of regional to local scale intra-arc faults and dikes. Our results suggest that the volcanic system has been constructed on top of a ca. 9 km long and ca. 4 km wide ENE oriented transtentional interaction damage zone (e.g. Peacock et al., 2016), which is documented in a tectonic domain controlled by ENE- to WNW-striking normal to oblique-slip faults. This interaction damage zone, is flanked to the west by the Melado strike-slip fault and to the east by a diffuse deformation zone, depicted by discrete NS to NE strike-slip splay faults and the left-lateral NW-striking Los Cóndores Fault, first reported in this work. We postulate that such geometrical and kinematic array strongly organizes stress and strain accommodation during the time of volcanism, arising from the overall transpressional Southern Andes. Such transtensional elements play a key role in ENE-striking dike and fractures organization for volcanism and geothermal reservoir enhancement. The latter might be associated with heat sources and permeability structure of the ENE- oriented Mariposa high enthalpy geothermal system (MGS) inferred from magnetotelluric modelling (Hickson et al., 2011; Reyes-Wagner et al., 2017). We carried out paleostress analysis by using heterogeneous fault-slip data and hydraulic fracture geometrical data. Our aim is to constrain the overall structural geometry and kinematics enhancing quaternary geofluids pathways (magmatic and thermal), as a frame for ongoing geohazards and geothermal research in this region. Furthermore, this work not only evaluates the stress condition for a discrete volcanic complex, but also seeks to provide key elements to better understand and decipher the role of Andean-transverse faults (ATF; e.g. Pérez-Flores et al., 2016) in the spatial organization of arc volcanism and geothermal systems. In particular, because of a lack of structural-geology knowledge in this region, many fundamental questions remain unresolved, such as: (1) why is the SVZ segmented, as described by Stern (2004)?; (2) why does 125 the TSVZ segment have an anomalous width in comparison with segments to the north and south?; (3) Why does it hold a relative long-lived activity? We hypothesize that there is a first order causal relationships between crustal faulting and fluid migration in this region. Relative wider segments of volcanic chains, such as the Transitional Southern Volcanic Zone of the Andean magmatic arc, are characterized by complex and wider, transverse long-lived fault arrays. In the TSVZ segment, arc-oblique faults might dominate over margin-parallel faults, making the overall arc-architecture more complex than segments to the south, which are dominated by the margin- parallel Liquiñe-Ofqui Fault System (Cembrano and Lara, 2009). 4.2 Tectonic setting South-central Andes magmatic arc tectonics (33ºS and 46ºS) is governed by stresses transferred from dextral-oblique convergence and interplate coupling between the Nazca and South American plates (Angermann et al., 1999; Moreno et al., 2012; Fig. 4-1). In this segment of the Andes, the convergence vector strikes N78º with an average velocity of ca. 66 mm/y (Fig. 4-2) that has remained relatively constant over the last 20 Ma (Angermann et al., 1999; Pardo-Casas & Molnar, 1987; Somoza, 1998). Partial strain partitioning has been evidenced by compressional tectonics in the fore-arc and foreland regions, and by transpressional to transtensional tectonics in the intra-arc region (M. D. Saint Blanquat et al., 1998; José Cembrano et al., 2000a; Corbett & Leach, 1998; Dewey & Lamb, 1992; Folguera et al., 2004; Pérez-Flores et al., 2016; Piquer, Berry, Scott, et al., 2016; Sielfeld et al., 2016, 2018). South of latitude 38˚S, intra-arc tectonics is governed by the margin-parallel ca. N10˚ striking Liquiñe-Ofqui Fault System (LOFS; e.g. Jos Cembrano et al., 1996), whereas from 38˚to 33.5˚S, the tectonic framework is represented by discrete ~NS- and oblique-to-the-arc striking faults, i.e. Andean Transverse Faults (ATF). For instance, the El 126 Melado and Troncoso faults at ca 36˚-36.5˚S (Cardona et al., 2018), the El Fierro Fault System at ca. 35˚S (Farías et al., 2010) among others are in spatial association with giant ore deposits such as Rio Blanco-Los Bronces (33.1˚S) and Teniente (34.1˚S) porphyry coppers (Piquer, Berry, Scott, et al., 2016; Yañez et al., 1998). In this segment (33.5˚ - 38˚S) the Andean orogen strikes between N17˚~N20˚(i.e. Maipo orocline; Farías et al., 2010; Yañez et al., 1998; Yáñez & Cembrano, 2004), limited to the north by the southern edge of the central Chile-Argentina flat slab region (Martinod et al., 2010), which results in the northern end of the SVZ (López-Escobar et al., 1995). Other major transverse crustal features, such as the Melipilla geophysical anomaly (Yañez et al., 1998) also occur in this transitional region. Regarding Andean arc segmentation (López-Escobar et al., 1995; Stern, 2004) the TSPPC belongs to the Transitional Southern Volcanic Zone (TSVZ), which is developed on the physical transition from a continental crust up to 60 km thick at 33˚S, to a thinner crust that remains ~30- 38km thick south of latitude 36˚S (Wes Hildreth & Moorbath, 1988; Tassara & Echaurren, 2012). The continental crust thickness beneath TSPPC is estimated in ~46km from 3D density model and seismic information (Tassara & Echaurren, 2012). The TSVZ constitutes one of the largest magmatic provinces of the Andean Southern Volcanic Zone, with up to 60 km wide in the arc front and >200 km wide considering vast back- arc volcanism (C. Stern et al., 2007; C. R. Stern, 2004). It preserves an almost uninterrupted volcanic history since the Late Pliocene (e.g. Frey et al., 1984; W. Hildreth et al., 1999; Wes Hildreth et al., 2010; Singer et al., 1997a). Synchronic and successive phases of volcanic landscape construction and erosion (Singer et al., 1997) encompasses a complex tectonic evolution, strongly influenced by the structural grain heritage ( a Lavenu & Cembrano, 1999b; Radic, 2010b). The TSVZ hosts volcanic products ranging from basalts to rhyolites, erupted from composed 127 volcanoes, minor eruptive vents and calderas (W. Hildreth et al., 1999; Wes Hildreth et al., 2010; Singer et al., 1997b). Figure 4-1 Seismotectonic map of the south-central Andes. Shaded relief image generated with 1 arc-second (~30 m) resolution data from the Shuttle Radar Topography Mission. Convergence rate and direction is shown by the solid white vector (Angermann et al., 1999). Margin orthogonal and margin-parallel components are represented as dashed arrows. The average slip vector of subduction thrust earthquakes is indicated in blue. The expected residual margin-parallel slip component is shown by green vectors (Stanton- Yonge et al., 2016). Approximate coseismic rupture zones (slip>1m) of the 1960 Valdivia Mw 9.5 (epicentre in four-tips red star) and 2010 Maule (moment tensor in red-white) megathrust earthquakes are outlined by cyan and yellow dashed lines, respectively (M. Moreno et al., 2012b; M. S. Moreno et al., 2009a). Moment tensors of the gCMT catalog (www.globalcmt.org, 1976-2017) include: i) subduction-type earthquakes (blue-white moment tensor Mw > 6.5) ratifying the average slip accommodated in the subduction interphase; ii) outer-rise event Mw > 6.5 (beige-white beach-balls); iii) crustal events Mw ≥ 5 and depth < 20 km in green-white beach- balls. Crustal focal mechanisms from local deployments are shown in black-white (C Haberland et al., 2006). Focal mechanisms are scaled proportionally to their moment magnitude by a factor of 0.4. Location of geothermal springs and stratovolcanoes is taken from Hauser (1997), Risacher et al. (2011), Siebert et al. (2010). Crustal faults systems and morphotectonic lineaments from (Cembrano and 128 Lara, 2009; Lira et al., 2015; Melnick and Echtler, 2006; Perez-Flores et al., 2016). The rectangle in white lines represent the area shown in Figure 4-2. Figure 4-2 Volcano-tectonic setting of Transitional Southern Volcanic Zone. Pleistocene-Holocene eruptive products are indicated as pink polygons, summits of main strato volcanoes with white triangles and caldera with grey ellipses. Active faults are in blue line and morphotectonic lineament with white-grey dashed lines. The equal area stereogram at the bottom right compiles 43 10 km-long lineaments measured in this image. Crustal focal mechanisms are from gCMT, whereas position of the Mw 6.2 Melado earthquake was relocated by Cardona et al., 2018. 129 Regional geology The Tatara-San Pedro-Pellado volcanic complex overlies by unconformity a heterogeneous and deformed basement (González and Vergara, 1962; Muñoz and Niemeyer, 1984; Fig. 4-3). This unconformity surface outcrops between 1800 to 2000 m a.s.l. Basement rocks mainly consist of Meso- Cenozoic folded stratified units, Cretaceous and Miocene granitoids and non-folded Pliocene pyroclastic strata. Geochemical evidence of non-outcropping Precambrian through Triassic crust at depth, has been reported by Nelson et al. (1999). The Mesozoic back arc sequence of the Andes at these latitudes includes fossiliferous marine sedimentary rocks representing transgression-regression cycles (e.g. Nacientes del Teno Fm., Baños del Flaco Fm.; Klohn, 1960); and successions of continental red sedimentary rocks composed of breccias, conglomerates and sandstones constituted of volcanogenic debris (e.g. Río Damas Fm., the informal Brown-reddish clastic unit, BRCU; Charrier et al., 1996; Flynn et al., 2003; González and Vergara, 1962). These units are intruded by the ~79 Ma granodioritic, El Indio Pluton. (Drake, 1976; Nelson et al., 1999). Mesozoic basement units record regional-scale east-verging thrusting, folding and block rotation (R. Charrier et al., 1996; Kay et al., 2005; Mescua et al., 2013). In turn, the Oligo-Miocene cover consists of volcanic and volcano-sedimentary sequences of the Cura-Mallin and Trapa-Trapa formations (which correlates to the Abanico Fm. north of 36˚S; e.g. Charrier et al., 2010) deposited on top of an eroded surface made of Mesozoic rocks (unconformity 1 in Fig.4-3). These units represent filled intra-arc basins (hemigrabens), which have been inverted by Late Miocene - orthogonal to the arc - shortening (e.g. Burns et al., 2006; Radic, 2010b). They display ~NS striking folds which are describe later. 130 All above mentioned units are intruded by two ca. 6 Ma old plutons: the Risco Bayo and Huemul granitoids (Fig. 4-3). Risco Bayo pluton ranges from granodiorite to Qz-monzonite and is exposed north of the TSSPC, mostly in the Río Colorado drainage; the Huemul leucogranite, in turn, directly underlying the TSPPC, outcrops in both the northern and southern flanks of the San-Pedro – Pellado saddle (Davidson et al., 1988; Singer et al., 1997). Evidence for Pliocene – Early Pleistocene volcanism (e.g. Campanario Fm., Drake, 1976; Cola de Zorro Fm., González and Vergara, 1962) is only locally found. These units display sub-horizontal bedding and uncomformably overlie the above mentioned Eocene-Miocene basement units (unconformity 2 in Fig. 4-3). 131 Figure 4-3 Geological map of the Tatara-San Pedro-Pellado volcanic complex. Compiled from Hildreth et al., 2010; Sernageomin, 2003; Singer et al., 1997. The nature, geometry and kinematics of regional faults and folds have been further constrained in this study. 132 Tatara-San Pedro-Pellado volcanic complex The Tatara-San Pedro-Pellado volcanic complex represents an Early Pleistocene-Holocene group of composite volcanoes, lava flows, dike swarms and minor vents, earlier thought to have been built upon an eroded Caldera (Río Colorado Caldera; Davidson et al., 1988). The TSPPC locates in the central western margin of the Transitional Southern Volcanic Zone (TSVZ; Stern 2004). From west to east, the main summits represent two massive volcanic edifices, The Tatara- San Pedro and the Pellado-San Pablo peaks (Fig. 4-2). These edifices are aligned in an ENE- orientation with more than 20 eroded minor eruptive vents (i.e polygenetic and monogenetic vents) located in the flanks and extending to the east, over a lava plateau (Frey et al., 1984; Singer et al., 1997; Hildreth et al., 2010). The volcanic history of the Complex comprises periods of volcanism separated by stratigraphic lacunae implying significant periods of erosion (Singer et al., 1997). Silicic to intermediate, ENE striking, subvertical dikes have been mapped in the northern side of the complex (Singer et al., 1997). The volcanic products erupted from at least three central vent regions include lavas and pyroclastic deposits compositionally ranging from basalts to rhyolites (mean compositions are within the basaltic-andesitic range, i.e. 52%-56% SiO2; Singer et al., 1997). Detailed description of the volcanic chrono-stratigraphy can be found in Singer et al. (1997) and Hildreth et al., (2010). For this study, eruptive units play a key role for constraining the relative age of faults, fractures and dikes. Therefore, these units have been organized as follows (Figs. 4-3 and 4-4). 133 Early Pleistocene Eruptive unit (Plev: 0,781 – 1,806 Ma) This unit distributes as isolated lava flows and pyroclastic piles mostly to the north and east of the TSSPC, covering an area of ca. 79 km2 and displaying variable thickness of tens to hundreds of meters. The eruptive products range from andesitic to rhyolitic compositions with minor basalts and basaltic-andesites. Some eroded eruptive structures are moderately preserved such as the El Murcielago rhyolitic dome. This unit uncomformably overlies Tertiary strata of the Curamallin Fm. and intrusive rocks of the Risco Bayo and Huemul plutons (Singer et al., 1997; Nelson et al., 1999; Hildreth et al., 2010). Middle Pleistocene Eruptive unit (Plmv:126 - 780 ka) This unit covers 151 Km2 all over the TSSPC, representing the most abundant eruptive products in the area. It displays variable thickness depending on the distance from the emission source. This unit includes moderately preserved polygenetic and monogenetic vents such as El Zorro Volcano and El Murcielago rhyolitic dome (indicated in Fig. 4-3). It is mostly composed of lavas and pyroclastic material ranging in composition from basalt to andesite and lesser evolved eruptive products (Singer et al., 1997 and Hildreth et al., 2010; Hickson et al., 2011). Late Pleistocene Eruptive unit (Pllv: 11,7 - 125 ka) This unit is mostly made up from basaltic to rhyolitic lavas and pyroclastic deposits erupted from the Tatara and Pellado volcanoes, as well as from minor eruptive centres distributed between the TSSPC and the LMVF (Hildreth et al., 2010). The unit covers ~93 Km2, displaying a highly 134 variable thickness: from isolated lava flows (4-20 m) in distal zones, to thick lava packages (>100 m) in proximal zones. (Singer et al., 1997 and Hildreth et al., 2010; Hickson et al., 2011). Holocene volcanics (post-glacial; Hv < 11,7 ka) This unit includes the products of San Pedro volcano and tens of minor eruptive centres from LMVF covering ca. 64 Km2 on an E-W elongated surface. The eruptive material are principally lavas and pyroclastic deposits ranging from basalts to rhyolites (Dungan et al. 2001; Hildreth et al., 2010). Geothermal Features Four surface geothermal areas surround the Pellado volcano between 1960 and 2400 m a.s.l. (steaming soils in Figs. 4-3 and 4-4): Pellado fumarole, La Plata hot-spring, El Valle and Los Hoyos steaming grounds. All geothermal areas occur at the edges of a conductive zone discovered with a magnetotelluric survey (Hickson et al., 2011). Gas geochemical survey carried out by Hickson et al. (2011) on three out of the four manifestations indicates that the surface geothermal activity is typical of neutral high temperature geothermal reservoirs, inferring up to 292˚C by using hydrogen-argon geothermometer (Giggenbach et al. 1991; in Hickson et al., 2011). Davidson et al. (1988) earlier suggested that these features represent the vestiges of a caldera-related hydrothermal system. This system has caused pervasive argillic alteration, which can be detected by white-yellowish soils at Pellado volcano flanks. 135 4.3 Field mapping and data analysis technics A combination of structural field mapping and stress field analysis from fault-slip data and dikes constitutes the basis of this work. Intensive field work was carried out between 2014 and 2016 summer seasons. Mapping was organized in order to recognize the basement architecture and the intra-volcanic TSPPC structural style. We used the ‘dip direction – dip’ convention for plane geometry mapping and, right hand rule (RHR) convention for striae and linear structural elements (database S1). Folds were mapped by stratigraphic correlation of volcanoclastic successions and systematic stratification (S0) measurements. A total of 3 east-vergence asymmetric fold axial planes were defined after stratigraphic correlation and S0 integration (Fig.4-3). Geographic distribution and field pictures of folds are presented in Fig. 4-3, 4 and 4-5. Faults, veins and dikes were systematically examined, to characterize cross-cutting relationships and relative age among structural elements, detailed geometry, and when correspond, fault sense movement. A representation of such structures is synthetized in the map of figures 4-3 and 4. Field pictures of different structural elements are summarized in figures 4- 6, 7, 8 and 9. Fault-slip data was obtained by using the kinematic criteria described in Petit (1987). Structural stations (ST; Table 6-3) with abundant striated fault surfaces and over n=12 kinematic indicators were hierarchized on the basis of fault core thickness and/or slickenside quality (i.e. variety and consistency among kinematic indicators). After first observation of dynamic solutions (pair-stereograms and stress tensor per structural sites are presented in Fig. 6-5) we could define structural blocks, identified by common stress-state in geographic coherence within the overall 136 brittle architecture. Fault-slip data sets consist of a list of fault plane dip-direction and dip, plus striae trend, plunge and sense of movement for each representative structural element. Kinematic indicators observed include mineral growing fibres, coherent slipping steps with mineral or cataclastic filling, a variety of Riedel and hybrid-type fractures and tension cracks (see field photographs in Fig. 4-6). In sum, our data base consists of 899 fault-slip data as collected from 35 structural stations (ST), distributed over and area of ~210 km2. From these, only 33 structural sites with a total of 885 fault-slip data were considered for kinematic and paleostress analysis (the two remaining ST have less than 12 fault-slip data). An additional data base of dilatational fractures (i.e. dikes and veins, examples in Fig. 4-6, 8 and 9)- was generated. A total of 96 basement, and 54 intra-volcanic dikes at TSPPC were mapped ranging from Late Pliocene to late Pleistocene (<103 ka) in age respectively (Figs. 4-4 and 4-5). Commonly, subparallel dikes outcrop in swarms of up to 20 individual bodies, cropping out between 1600 and 2600 m a.s.l. Locally present hydrothermal-vein data was collected commonly within fault zones (Fig. 4-6). Descriptions were made combining both, field and satellite images technics. Strike, dip and thickness data were systematically measured in the field. When observable, dikes length was estimated from satellite images as well as along-strike variations (e.g. sinuosity) and dike segmentations. Kinematic analysis of fault-slip data was early processed with the FaultKin program (Allmendinger et al., 2012; Marrett & Allmendinger, 1990). This algorithm calculates the principal axes of average incremental strain for scale-invariant faults by using the Bingham distribution of shortening and extensional axes (P&T) of each fault-slip record. Data was grouped and analysed following a 3 steps criteria: 1) relative age; 2) filling material; and 3) geographic location. Mapped 137 fault planes as well as computed P and T axes were commonly found to be heterogeneous, thus clustering and fault-slip data differentiation was carried out applying dynamic analysis. Dynamic analyses for paleostress conditions were conducted separately for fault-slip and tensional fractures data. We used the Multiple Inverse Method (MIM, Yamaji, 2000) for fault-slip data clustering and paleostress state determination. We first analysed separately each structural site with more than 12 complete fault-slip data (fault geometry, striae orientation and sense of movement) and then correlated them by state of stress and geographic proximity. The inversion method is described in detail in the following section. Paleostress analysis of dilatational fractures was done using the GArcmB software (Yamaji, 2016; Yamaji & Sato, 2011; described later). We used the prefix “paleo” for all results derived from the abovementioned technics in order to differ them from current stress field estimated from active tectonic observations (e.g. natural seismicity, geodesy). The Multiple Inverse Method (MIM) for heterogeneous fault-slip analysis Classic stress inversion methods are only applicable in limited cases in nature, where homogeneous fault-slip data can be obtained. If within a certain area, faults do not slip in response to a unique deviatoric stress, fault slip data are called heterogeneous, therefore biased results are expected from classical stress-inversion technics (Yamaji, 2000). For example, curved fault striae left by the 1995 Hyogo-ken- Nanbu (Kobe) earthquake, interpreted as variations of coseismic slip direction result in heterogeneous fault-slip (Otsuki et al., 1997; Yamaji, 2000; Yoshida et al., 1996). Even more, orogenic belts, built upon thousands seismic cycles, are 138 expected to record heterogeneous and polyphasic slip elements within long-lived fault surfaces (e.g. Veloso et al., 2015; Yamaji et al., 2006). The Multiple Inverse Method (Yamaji, 2000) is based on Angelier's (1984) inverse technic that assumes that fault slip in the direction of the shear stress applied on the fault surface. Among other technics, MIM allows to separate deviatoric stresses from heterogeneous fault-slip data, constraining common stress tensors to a certain combination of a fault-slip data set. The MIM algorithm calculates and sorts for a given data set the orientation of principal stress axes and computes the relative stress-ratio for a combinatorial amount of subsets containing k-number of fault-slip data elements (Yamaji, 2000). The generated number of subsets (C) for a total of N structural elements is given by the binominal coefficient: ,𝐶 = ,! . (eq.4.1) .!(,1.)! where N! is the factorial, N! = N(N-1)(N-2)….2x1. In this study, we used k=5 for each structural site data set. For the global stress analysis including the 899 fault-slip data (database S1; last diagram in Fig. 6-5), k was set to 3, else the amount of combinations exceeds the software analytical capacity. Then the best-fit stress tensor, and hence the orientation of the principal stress axes [s1, s2 and s3], is obtained for each subset. The software calculates the optimal stress tensor for each subset of k structural elements applying classical inverse methods based on the Wallace-Bott hypothesis (Bott, 1959; Wallace, 1951). Stress-state solutions are plotted in paired equal-area stereograms showing the orientation of maximum and least axial stresses (the left diagram shows s1 and the right, s3; see for example solutions in Fig 4-4.). Solution are individualized by a 139 ‘tadpole’ symbol, where the head represents s1 or s3 axis position and the tail indicates the trend and plunge for s3 or s1, respectively (Yamaji et al., 2011). Solutions are color-coded in accordance to their calculated stress ratio [f = (s2-s3)/( [s1 - s3); eq.2], where f ranges from 0 (violet) to 1 (red; colour-bar is presented in between paired-diagrams); f- values are indicators of the stress ellipsoid shape, where f = 0, corresponds to a prolate shape (i.e. uniaxial compression, where s2=s3 ) and f = 1, to an oblate ellipsoid (i.e. radial compression s1=s2), thus triaxial differential stresses have intermediate f values. Therefore, a state of stress is completely depicted by a single tadpole symbol on the left or right equal area stereogram, however the combined use of both is useful for discerning the principal orientation of the state of stress of ordinary fault-slip data set (Yamaji et al., 2011). In order to evaluate the dominant f-value among diverse stress-rates solutions, f histograms where calculated for each individual data set. So, knowing the dominant f value, stress tensor where manually picked on the highest concentration area of tadpole clusters. In case of bimodal f behaviour two stress-tensor solutions were defined and separately presented in the fault-slip data solution-table of Fig. 6-5 in S.M. The stress tensor solution (s1axis is denoted with a blue triangle and s3, with a red star in diagrams of Fig. 4-4 and 6-5) for each analysed structural station is plotted on a slip vector stereogram where optimally oriented faults to be active under the calculated stress state are coloured in red or purple. The latter colour-code is indicative of the misfit angle (i.e. red: 0˚-10˚, purple: 11˚-20˚, and blue: 21˚-30˚), defined as the difference between calculated and theoretical fault-activation angle for a certain stress tensor. 140 Dilatational fracture inversion: mixed Bingham distribution for 3d orientation data fitting The orientation distribution of dilatational fractures is affected by the state of stress around them at the time of their formation and by the pressure of the fluid that opened them (Faye et al., 2018; Jolly & Sanderson, 1997). This is often represented as a multimodal distribution of the poles to dilatational fractures. Therefore, clustering of fracture orientations is required to interpret stress conditions at the time of fracture formation (Yamaji, 2016; Yamaji & Sato, 2011). The GArcmB software (Yamaji, 2016) fuzzy clusters 3D orientation data of dilatational fractures by searching for the mixed Bingham distribution that best fits the dataset (Yamaji, 2016; Yamaji and Sato, 2011). The genetic algorithm employed follows and improves Jolly and Sanderson’s (1997) method, that estimates the orientation of the minimum principal stress (σ3) from the maximum concentration of fracture poles (Yamaji and Sato, 2011). Unlike Jolly and Sanderson’s (1997) method, the present method allows paleostress analysis of dilatational fractures that were formed by polyphasic tectonics by fitting the dataset to a linear combination of Bingham distributions, with their own concentration axes (k1, k2) and parameters, each interpreted as representing a different stress state (Yamaji, 2016). The method allows clustering for K values 1, 2, …, 5. For each K value evaluated, a mixed Bingham distribution is fitted to the N orientation data by maximizing the logarithmic likelihood function L (Yamaji, 2016). The program should be run for several iterations for a given data set and K value, because the results of the computation depend on the configuration of the randomly generated 0th generation (Yamaji, 2016). Once L is maximized, K is evaluated using Bayesian 141 Information Criterion (BIC). The optimal number of clusters (Kopt) is the K value that minimizes the BIC function: 𝐵𝐼𝐶 = −2𝐿(𝑋9:;) + (6𝐾 − 1)log (𝑁) (eq. 4.2) where Xopt is a function of K and represents the Mixed Bingham distribution that best fits the data, L(Xopt) is the log likelihood function, N number of data, K number of parameters (clusters) (Yamaji, 2016). Once the Kopt value is chosen, stress conditions are estimated from the equal area projection and the concentration axes and parameters (Yamaji, 2016; Yamaji and Sato, 2011). Principal stresses orientation axes (σ1, σ2, σ3) are obtained from the minimum, intermediate and maximum concentration axes of each Bingham component respectively (Yamaji, 2016; Yamaji and Sato, 2011), thus indicating the state of stress at the time of fracture formation. Furthermore, stress ratio is obtained from the concentration axes. To interpret the results as stress conditions, the following assumptions are taken: (1) host rock is isotropic; (2) far field stress is invariant during the formation of the cluster; (3) fractures will form only if the fluid pressure exceeds the normal stresses; and (4) fractures making a cluster were formed intermittently from fluids (thermal o magmatic) with various pressures (Faye et al., 2018; Yamaji & Sato, 2011). A total of N=150 dike orientation data was analysed using the method described above (database S1). We subdivided the total fracture data (dike orientation) into two groups, older basement dikes (N=96) and younger intra volcanic dikes (N=54). We ran » 10 iterations of the GArcmB program for each K value (K =1, 2, …, 5). 142 4.4 Results Our structural observations can be synthetized as follows: a) middle to late Miocene regional scale folding, b) brittle deformation leading to paleostress analysis,c) Paleostress analyses from diking. Results are sub-grouped upon chronological order estimated by cross-cutting relationships among geological objects. A summary of the spatial distribution of folds, faults, dikes and veins are shown in the maps of figures 4-3 and 4-4; whereas their appearance in selected field pictures of figures 4-5 to 4-9. Middle to Late-Miocene east-verging folding Subparallel, ~NS-striking asymmetric folding with east vergence were recognized and measured affecting Oligo-Miocene volcanic and volcano-clastic rocks (i.e. Cura-Mallin and Trapa-Trapa formations; OMvc in Fig. 4-3) and unconformable underlying Mesozoic sedimentary strata (Muñoz and Niemeyer, 1984; González-Ferrán and Vergara 1962; Charrier et al., 2010). Geometrical data (S0) from Mesozoic units are not considered for Tertiary folding analysis because these units might be affected by Late Cretaceous deformation and block rotation (Mescua et al., 2013). In turn, bedding geometry of Oligo-Miocene strata yields to qualitative identification of three inclined cylindrical folds: Ñirales Anticline, Las Yeguas-Verdugo Anticline and Saso Anticline (from W to E, respectively; Figs. 4-3 and 4-4). These are characterized by west dipping axial planes (~70˚) defined by steep east-dipping (~60˚~90˚), and gently west-dipping flanks (20˚- 50˚; refer to pole concentration contours in Fig. 4-5). Fold geometry alternates tight antiforms with open synforms that are plunging ca. ~8˚ to the north (more field pictures are presented in Fig. 6-6). We estimate fold wavelength of 8 km in average, however smaller scale- and parasite-folds (m scale) wavelength have been observed close to the major hinge zones. A regional Miocene-Pliocene angular unconformity obliterates continuous folded forms (Fig. 4-5c). In the western and central zone of the study area, folded units underlie unconformable the TSPPC, whereas to the east, Pliocene pyroclastic units (Campanario Fm.) overlie the Miocene-Pliocene unconformity (Fig. 4-3). 143 Figure 4-4 Structural map with main structural features. Stereograms contain fault slip data and stress tensor solutions of 17 selected structural stations (other solutions and pair stress-state stereograms are presented in Fig. 6-5). 144 Figure 4-5 Field pictures and bedding analysis for Oligo-Miocene volcanic and volcanoclastic strata (OMvc). a) Oblique picture at Ñirales ridge on south-western flank of Tatara-San Pedro edifice. White-dashed line indicates unconformity between folded Oligo-Miocene strata and Middle-Pleistocene lava related to Tatara eruptive sequences (TSPPC). Bedding and fold axes are denoted with blue symbols. White solid line traces approximately the A-A’ profile view in panel ‘b’ and the white arrow head the position and direction of photograph in panel ‘b’. b) Tight asymmetric antiform of Ñirales Fold with ~N10˚W/70˚W oriented axial plane. Projection of two bedding surfaces are shown in blue-dashed lines. c) Steeply east-dipping OMvc pyroclastic strata underlying subhorizontal and topography-parallel lavas (TSPPC) at Molino Gully. Panel ‘d’ presents the pole concentration contour of measured beddings (S0) for Ñirales and Las Yeguas- Verdugo anticlines; and pole contours for all observed axial planes. Frictional deformation and paleostress fields from fault-slip data analysis The TSPPC basement is dominated by normal to oblique-slip faults, which outline an E-W ENE- oriented depressed tectonic block limited to the west by the NS- striking Melado dextral strike-slip fault (Cardona et al., 2018) and to the east by a complex array of strike-slip to oblique- slip faults (Fig. 4-4). Brittle deformation is manifested as sets of mesoscopic faults, fractures, veins and dikes (Fig. 4-6). Geometric and dynamic analyses of frictional faults have been carried out for 31 structural sites with a total of 899 fault-slip data (Fig. 6-5; refer to section 6.2.1 for dataset). Analyses yield to an overall transtensional regime within the proximity of the TSPPC (Fig. 4-4) 145 surrounded by transtensional oblique-slip to strike-slip regimes in distal zones of the TSPPC. The distribution of calculated stresses has been organized by structural blocks on the basis of fault geometry and kinematic similarity. For instance, substantial NW, NS- and NE-strike-slip to oblique-slip faults localize in blocks surrounding the central structural block, where dominant ENE- to WNW-striking normal faults are kinematically consistent and parallel to extensional elements such as dikes and veins (Fig. 4-6). Figure 4-6 Field picture compilation of fault, hybrid fractures and veins cross-cutting volcanoclastic rocks (OMvc). a) Conjugate faults. Striated surfaces with limonite patches can be observed in the centre of the picture. b) Hybrid fault-fracture with hydrothermal mineral fibres (zeolite) growing in the direction of slip. c) Hydrothermal veins in an E-W oriented en echelon array. d) Vertical to subvertical veins within the Castillo Gully normal fault (ST. 18). The Western Block (WB) was defined considering detailed mapping and dynamic analysis along the Castillo Gully, Ñirales Ridge and San Pedro Gully (ST. 21-24 in Fig. 4-4). Here volcanoclastic rocks are affected by NNE to ENE striking transtensional dextral oblique-slip faults. Heterogeneous fault slip data analysis yields to diverse stress states, however s1 is consistently plunging between 24.5˚ and 57.4˚. Because of faulting does not affect volcanic rocks of the 146 Tatara volcano (Pllv), this deformation phase is roughly constrained in age between the Late Miocene and Middle Pleistocene. The Central Block (CB), in turn, is defined by kinematically coherence among fault-slip data sets of structural stations located to the north (i.e. ST. 11-13) and south (i.e. ST. 25-28) of the Pellado-San Pablo volcanic edifice (Fig. 4-4). Affected basement rocks range from deformed volcanoclastic sequences (OMvc), leucogranites (Mgh) to subhorizontal pyroclastic layers (Pc). Here, oblique-slip to dip-slip faulting is characterized by steep plunging striation dominated by normal, dip-slip, kinematic indicators. Particularly dense fault and fracture arrays have been mapped at El Valle Steaming Ground site (ST.11) and at the Little Graben site (ST.28). The later, as the name suggests, constitute an area of ca. 10.000 m2 characterized by ENE striking normal faults hosted within the 6.2 Ma Huemul leucogranite. On some fault planes, active upwelling of warm/cold (~20˚C) water has been found. The Little Graben structure models the overlying lavas of Cordón Guadal (Plmv) and Pellado volcano (Plmv and Pllv; Fig. 4-4). Just two structural sites include faults affecting volcanic products of the TSPPC (Fig. 4-7 and 4-8). From these, only ST. 25 at the Pellado Fumarole (Fig. 4-8) has enough fault-slip data for applying the multiple inversion method. Dynamic analyses for stations within the CB, yield to stress tensor solutions characterized by inclined to subvertical s1, plunging between 45.5˚ and 72.2˚(Figs. 4-4 and S1). Because of the large amount of fault-slip data, the kinematic and geometric similarity among the observed faults, and the documented cross-cut relationships, this structural block might hold a longer activity at least until Late Pleistocene, as suggested by observations at the Pellado fumarole area (Fig. 4-9). The overall fault geometry and kinematics within the CB, fairly describe a ca. 8 km wide and ca. 12 km long graben-like architecture beneath the main volcanic edifices. 147 The north-eastern Block (NEB), in turn, is defined along the Maule River valley where stations 1 to 6 and station 30 where mapped. This block, consistently depicts strike-slip to oblique-slip faulting. Here major fault zones occur at El Medano village, La Plata-Maule rivers junction and along the Maule River-Los Cóndores Gully NW morphostructural lineament (Figs. 4- 2 and 4-4). The northernmost fault zone is placed surrounding the El Medano village (ST. 1-4). Here, pervasive hydrothermal alteration and diffuse geothermal activity are localized within a wide damage zone controlled by NS-, NW- and NE- striking faults, displaying locally right lateral strike-slip splay faults. In turn, the La Plata Fault (ST.5) was defined as a NE-striking dextral strike slip at La Plata-Maule rivers junction (see stress tensor solution in Fig.4-4 or 6-5). Tens of measured fault planes are consistent with a paleostress tensor orientation outlined by s1 trending N218˚- plunging 3.5˚ and s3 trending N309˚, plunging 3.0˚. A similar paleostress tensor (s1: N71.4˚/109˚ and s3: N336.2˚/25.6˚) was obtained for the multiple-core Los Cóndores Fault at Maule-Los Cóndores NW lineament. Measured fault planes in this station (ST.6) strike dominantly NW to E-W. Los Cóndores Fault and corresponding damage zone is characterized by pervasive hydrothermal alteration demarking the margins of the damage zone. Lastly, this block considers kinematic information collected at Lo Aguirre structural station (ST.30). Here, N~60˚E striking right-lateral strike-slip to oblique-slip faults where measured. A strike-slip regime with NW-orientated s1 was estimated for this site. Paleostress fields from dike-orientation data inversion Subvertical dikes and hydrothermal veins have been mapped within the defined structural blocks, following an almost constant E-W to ENE-strike (Figs. 4-6, 8 and 9). Dikes range 148 in age and composition from post-Late Miocene to pre-Late-Pleistocene, and from basaltic to rhyolitic, respectively. Singer et al. (1997) obtained an 40Ar/39Ar isochrone age of 818±16 ka on plagioclase samples of a rhyolitic dike. In this project we unsuccessfully tried to date younger dikes cross-cutting lavas of the Pllv unit (i.e. <103 ka), so a relative age of >93 ka (i.e. pre-Late Pleistocene) is estimated. Older dikes, reported to be emplaced only in basement units are typically mafic with porphyritic texture and have up to 25 m of thickness. Examples of these dikes are presented in Figure 4-9 (a-c). In the Castillo Gully (i.e. western margin of TSPPC, Fig. 4-9b) for instance, subvertical E-W striking dikes are emplaced within hybrid fractures as observed by shearing textures and slickensides on their walls. Similar characteristics have been observed for up to 20 m wide mafic dikes in the La Plata-Maule valleys striking N50˚-60˚E (Fig. 4-9c) and in the Los Cóndores valley (i.e. eastern edge of the TSPPC) striking ~N70˚E. Younger dike families usually strike between N60˚ and N90˚, which geometries are characterized by along-strike and along-dip sinuosity and metric-scale off-sets, as examples presented in field pictures of Fig. 4-9 (d-f). Individual measured thickness of Pleistocene dikes range between 0.3 m up to 12 m. In mechanical terms dikes are considered to be dilatational fractures filled by magmas (Gudmundsson, 2011b).Their common tabular shape is supposed to represent the geometry of the dilatational feature triggered by the governing tensional stress field on the time of fracturing and magmatic filling. Dike geometry inversion for paleostress analysis have been conducted separately for older and younger dike families. A total of 96 basement dike orientations were collected and analysed in the TSPPC basement. This data set showed a minimum BIC at K=1 (Fig. 4-9), meaning that a single stress condition explains all basement dike orientations. This stress state corresponds to an tensional regime with NNW-trending σ3 (342,6˚/5,6˚) and stress ratio F=0,6320 (Fig. 4-9). In turn, a total of 149 54 intra-volcanic dike orientation data were measured and analysed within the TSPPC. This data set showed a minimum BIC at K=2 (Fig. 4-9), meaning that two stress conditions explain the dike orientations. These stress states correspond to slightly different tensional regimes. Solutions A is defined by a stress ratio of F=0,063 and principal stress axes trending and plunging as follows: σ3= 346,2˚/0,5˚; σ2= 76,2˚/4,3˚; and σ1: 249,1˚/85,7˚. Solution B is defined by F=0,3982 and stress axes oriented: σ3= 129,8˚/8,4˚; σ2=36,4˚/21,8˚; and σ1= 239,6˚/66,4˚ (refer to stereoplots in Fig. 4-9). Figure 4-7Glaciated Plmv rocks hosting the Pellado fumarole (structural station 25, see Fig. 4-4 for geographic reference). Set of faults and fractures spatially associated with fumarole activity spots and steaming soils. Left panel of pair picture-representation, shows normal-to-the-strike view of normal-right lateral fault set. Right panel shows an oblique view of the outcrop. 150 4.5 Discussion Integration of fault geometries and paleo-stress analyses plus active faulting (taken from Cardona et al., 2018) are used to define the nature, kinematics and volcano-tectonic significance of the ENE-oriented Tatara transtensional interaction damage zone (TDZ). This feature is developed between the Melado ~NS-striking dextral strike-slip fault (western master fault) and a set of NE- to NW- striking strike-slip Andean Transverse Faults (towards the eastern edge). Both flanks of the interaction damage zone coincide in the orientation of calculated stress fields, which are in line with the expected far-field (regional) subhorizontal and NE-trending σ1. Classic interaction damage zones between two oblique-to-each-other master faults are expected for isotropic media (Peacock et al., 2016). The TDZ however, is constructed within an extreme heterogeneous lithological setting and overprinted on older ductile structures (east-vergence folds). Bulk intra-arc transpressional state of stress in these latitudes might be internally partitioned in the volcanic front of south-central Andes, between 35.5 and 36.5˚S (TSVZ; e.g. Lavenu and Cembrano, 1999; Cembrano and Lara, 2009). We postulate that in intra-arc transtensional domains stress and strain might accommodate within juxtaposed structural blocks, where their frictional 151 Figure 4-8 Pleistocene dikes. a) View to the W at Colorado headwaters. The ridge separates the Colorado river to the east, from the Huemul river to the west, not observable in the picture. Tatara volcano summit is observable to the south. b) Schematic N-S oriented cross-section of picture in ‘a’: 6.4 Ma granitoid body on the north (Mg), basement strata (OMvc) and early to middle Pleistocene volcanic rocks from Tatara volcano. Observed mafic dikes are indicated in red (dashed when inferred). Mapped normal faults are in blue lines (dashed when inferred). Alteration zones in light brown are related to 818 ka rhyolitic dikes (Singer et al.,1997). c) Mafic dikes with ENE orientation and fault damage zone along the ridge of “a” and “b”; d) Slickensides in andesitic lava of Plmv unit; e) Fault-slip data of faults represented in b. interaction partially consumes the overall margin-parallel transpressional budget, transferred from interseismic plate coupling. Such transtensional elements within margin-parallel intra-arc strike-slip settings strain might accommodate within juxtaposed structural blocks, where their 152 frictional interaction partially consumes the overall margin-parallel transpressional budget, transferred from interseismic plate coupling. Such transtensional elements within margin- parallel intra-arc strike-slip settings should play a fundamental role in subvertical diking and fracture organization for volcanism and geothermal enhancement (Fig. 4-10). Therefore, at the TSPPC scale (i.e. ca. 100 km2), the here reported TDZ, provided a long-lived permeable pathway for ENE-oriented diking. The later, consequently, has driven the elongated shape of the volcanic complex. The TDZ has an internal graben-like architecture (Fig. 4-10), as depicted by the fault geometries governing the Central Block (Fig. 4-4). Such geometry, might have dissimilar fault off- sets in coherence with differential exhumation patterns observed on exposed basement rocks to the north and south of TDZ: for instance, the presence of Mesozoic units to the north, and its absence to the south of the TSPPC. In accordance with volcano-tectonic observations at historical scales (e.g. L Lara et al., 2004) and long-term/short-term numerical modelling (e.g. Stanton-Yonge et al., 2016), tectonic controlled volcanism might be triggered by any of the following large-scale processes: (1) co- to post-seismic relaxation during the subduction earthquake cycle, enhancing instantaneous extension in the direction of the regional convergence vector; such as observed from open fracturing and fissure eruptions (e.g. 2010 Maule Earthquake related co-seismic fractures in the main Cordillera and volcano subsidence, Arriagada et al., 2011; Hickson et al., 2011; Pritchard et al., 2013; Cordón Caulle 1960 NW-oriented fissure eruption, Lara et al., 2004); or (2) by steady state ENE- oriented inter-seismic crustal shortening in intra-arc regions enhancing diking and volcanism subparallel to the shortening axis. Evidence of such long-term processes can be observed in elongated volcanic complexes such as the Tatara-San Pedro-Pellado (e.g. 153 Singer et al., 1997), the eroded Sierra Nevada Volcano and the Callaqui–Copahue–Mandolehue volcanic alingment (Daniel Melnick et al., 2006b; Rosenau et al., 2006c; Sielfeld et al., 2016), and on the distribution of flank vents on stratovolcanoes (e.g. Llaima, Villarrica and Osorno composite volcanoes; Cembrano and Lara, 2009b; Lara et al., 2008; López-Escobar et al., 1995). The relatively long-lived (Pliocene-Recent) TDZ architecture might have been used as pathways for interseismic eruptive activity and sequential building of the TSPPC. The latter, might be associated with the heat source and permeability structure of the ENE- oriented Mariposa high enthalpy geothermal reservoir (MGS) identified from magnetotelluric modelling (Hickson et al., 2011; Reyes-Wagner et al., 2017). Specifically, the undeveloped MGS, discovered by low resistivity anomaly and three >600 m depth drillholes (Hickson et al., 2011), locates ca. ~0.7 to 1km beneath the Pellado composite volcano surface. Beside the Colorado Fault (Fig. 4-8), the Pellado-Fumarole Fault (ST. 25, Figs. 4-4 and 7) cross-cuts Middle Pleistocene volcanic products of the Pellado volcano. This fault-data set was collected in the Pellado Fumarole area, where several steaming spots are aligned following an ENE trend, parallel and just 8 to 10 m to the south of the Pellado-Fumarole Fault (Fig. 4-7). Its geometric and kinematic similarity with nearby stations cross-cutting the 6.2 Ma old Huemul leucogranite (i.e. ST.26-28), suggests that the governing transtensional stress field has been at least acting from ~Pliocene to Mid-Pleistocene. In line, the lava flows on top of the leucogranite rocks, exhibit marked ENE- morphologic scarps, which might be the neotectonic expression of faults mapped in the underlying basement (e.g. ST.28; Little Graben). In general, these features indicate that the transtensional regime driving pre-TSPPC deformation was still active during the time of Pleisto-Holocene volcanism and hydrothermal activity. Furthermore, the multiple stages 154 of erosion reported by Singer et al. (1997) might be associated with low volcanic output rate, which in turn, could be restricted to decreasing strain-rates of bulk transient deformation. Figure 4-9 Mafic dikes with ENE- predominant orientation. a-c) Field pictures of basement dikes. d-f) Field pictures of intra-volcanic dikes (<103 ka). Each picture shows a single family of dikes and where necessary are indicated by green arrows. g-h) results of fracture orientation data inversion for basement and intra-volcanic dikes, respectively. BIC-K graphs indicate the more plausible amount of paleostress clusters gathering conditions for fracture dilatation. Lower hemisphere equal area projections include pole to dike-planes (circles) and paleostress tensor orientation. Panel ‘h’ shows two solutions (K=2; clusters A and B), where color-coded circles indicate membership. As presented by Hickson et al. (2011) the Mariposa Geothermal reservoir might have an ENE-elongation, characterized by two elliptical lobes defined by a low resistivity anomaly. This shape, might be strongly controlled by the geometry of the fluids pathways, which might have 155 similar geometry and kinematics as the faults defining the transtensional stress-tensor for the TSPPC. We postulate that the fluid pathways are a mesh of WNW- to ENE-oriented normal- sinistral and normal-dextral faults and associated fractures, respectively. However, the reservoir level, supposed to be tapped by a clay cap between 400 and 1000 m below the surface (high resistivity; Hickson et al., 2011), might be constrained by the geometry and position of the TSSPC/Basement unconformity. It is also plausible that much of the water budget feeding the geothermal system has been accumulated within Pleistocene permeable domains in the damage zone, when the system apparently was more active. This led us to suspect that Mariposa-type geothermal systems within Quaternary transtensional zones in the Andes, gain significant structural and fluid conditions before and during the associated volcanic complex construction. Similar scenarios can be inferred for geothermal systems at Tinguiririca, Calabozo, Tolhuaca, Yates among other volcanic complexes. Figure 4-10 Conceptual model for the Tatara Interaction Damage Zone (TDZ). Light-blue surfaces represent major interacting intra-arc strike slip faults. Purple surfaces are a schematic interpretation of internal architecture of the TDZ. Red-pink volumes represent dikes. Greenish polygon indicates the Mariposa geothermal reservoir. 156 The Melado Fault described by Cardona et al., 2018 occur parallel to the glacio-fluvial incision along the Melado valley. This, as other surrounding valleys, might have a relatively long- lived morphologic evolution, and its NS-elongated shape might have been carved along the NS- oriented Melado Fault damage zone. Similar structural controlled Pleisto-Holocene glacio-fluvial landscapes have been documented along the Liquiñe-Ofqui Fault System in Southern and Patagonian Andes (e.g. Thomson, 2002) at lower altitudes as expected for higher latitude settings. This observation suggests that the Melado Fault could have been active during the Pleistocene and closely-coeval to the fault activity in the northern-east edge of the TSPPC (i.e Los Condores Fault, La Plata Fault; Fig. 4-10). Such structural controlled volcanic and geothermal systems are not only restricted to the Andes, but also might be expected to occur in other active convergent margins such as Central America, Sumatra and Japan (Valerio Acocella et al., 2018; La Femina et al., 2002; Mazzini et al., 2009; Suzuki et al., 2014). In particular, strain partitioning in Central America results in oblique- to-the-arc bookshelf faulting (La Femina et al., 2002), which creates permeable pathways for fluids cross-cutting the magmatic arc. 4.6 Conclusions 1. The Pleisto-Holocene Tatara-San Pedro-Pellado volcanic complex is made up by tens of eruptions and sub-parallel Pleistocene diking organized along an ENE-trend. 2. The volcanic complex is emplaced over a heterogeneous basement architecture prompted primarily by Late Miocene east-verging folding and the intrusion of ~6 Ma old granitoids; secondary, by frictionally deformation characterized by strike-slip to oblique-slip faulting contouring a transtensional interaction damage zone; the former are depicted by active intra-arc 157 margin-parallel dextral strike-slip faulting (the Melado Fault) and by a complex array of long-lived Andean transverse strike- to oblique-slip faults (e.g. La Plata and Los Cóndores faults). In turn, the interaction damage zone made up of dense normal faulting active at least since the Pliocene, configuring a graben-like structure. 3. Relative long-lived and staged volcanic history of the TSPPC (Singer et al., 1997) is compatible in time and position with the relative long-lived transtentional faulting evolution of the here- described Tatara interaction damage zone (TDZ). Stages of erosion and lack of volcanism might be associated with decreasing strain-rate of bulk transient deformation. 4. Paleo-stress states are compatible with a stress field rotation from uniaxial compression in a strike slip regime at the flanks of the TDZ to a radial extensional regime dominated by ENE- striking normal to oblique-slip faults in the central portion of the system. 5. Results of fault-slip data inversion on the edges of the TDZ are compatible and coupled with modern stress field orientation deduced from the relative strong Mw 6.2 Melado earthquake recorded the 6th of June, 2012. 6. Results of dynamic dike-geometry inversion are compatible with the solutions obtained from fault-slip data inversion on structural sites in the Central Block of the TDZ. This suggests a causal relationship between faulting and diking. 7. Stress field inversion for suits of late Tertiary to Quaternary dike-geometry and fault-slip data in south-central Andes is a consistent technic to evaluate modern stress regimes, as the positive correlation obtained between major strike-slip faults and seismic faulting in the study area. 8. Transtentional intra-arc settings generate effective geometries for magma circulation and emplacement in the upper schizosphere. The graben-like architecture of the TDZ has driven subvertical dike pathways where σ1 is subvertical and σ2 and σ3 subhorizontal, generating an 158 effective shape for the accumulation of eruptive products. Furthermore, transtensional tectonics in the TSPPC might have controlled the emplacement of a significant amount of geothermal resources, as estimated for the Mariposa Geothermal reservoir (~320MWe-inferred). 9. The transtensional interaction damage zone here describe, is not a classical step-over structure, because of the lack of clearly defined subparallel master faults. However, calculated state of stress in the eastern side of TDZ, as depicted by the La Plata NE- and Los Condores NW- striking faults, is totally compatible with the one associated with the Melado Fault (western master strike-slip fault). This particular evidence confirms the importance of ATF in the accommodation of bulk transpressional deformation within the magmatic arc. 4.7 Acknowledgments This research project has been supported by FONDAP project 15090013 CEGA “Centro de Excelencia en Geotermia de los Andes” and Fondecyt project 1141139 to J.C. Gerd Sielfeld acknowledges the support by CONICYT’s national Ph.D. scholarship 21140749. We thank the Energy Development Company (previously Magma Energy Chile) and landowners for allowing us to ingress to their concessions and properties for doing the field work. Special thanks to Carolina Rodriguez by helping us to get permissions to access along the La Plata Valley private road. We are grateful for the help provided during the field campaigns by P. Guzman, F. Aron, A. Griffith, R. St-Julien, A. Mercado, R. Torres, R. Torres Jr., M. González, Cochela, Boris and Juano. This work is dedicated to Patricio Mercado-Mercado (the duck that did not swam; R.I.P). 159 5 Conclusions The main conclusions of this thesis are summarized in the following paragraphs and are organized in chronological consistency as the three body (2, 3 and 4) chapters are presented. 5.1 Conclusions chapter 2. In this study, we present moderate to small magnitude seismicity ranging from Mw 3.6 to Mw 0.6, of a 16 month-long natural seismicity experiment. It is the first network to cover the northern termination of the Liquiñe-Ofqui Fault System at a regional scale, and provides high resolution for local earthquake detection and localization. The seismicity catalog consists of 356 crustal events, coherent with the bulk long-term tectonic construe from structural geology analysis for this segment of the Andean orogeny. Combined interpretations from geometric and kinematic analysis can be summarized into three major outcomes. 1. Frictional strain in the intra-arc region is compartmentalized into geometrically and kinematically distinctive latitudinal subdomains. In the northernmost portion of the LOFS (37.8˚-38.8˚S), seismic faulting is consistent with the long-term record of brittle deformation. Main NNE-oriented branches of the LOFS give way to a splay fault array, along which minor eruptive centers, stratovolcanoes and geothermal surface expressions are placed. A central zone (38.8˚-39.6˚S) is dominated by Andean transverse active faults, alternating NW- (e.g. Caburgua sequence, VL-Vi-La seismic cluster alignment) with NE-oriented faulting. In the southernmost sector of the study area (39.6˚-40.1˚S), margin-parallel faults accommodate most of the strike-slip component of the partitioned bulk strain. 160 2. The latitudinal convexity (as seen in WNW-ESE cross sections) and 30˚ west dip of the SLB in Southern Andes between latitudes 38.5˚S and 39.5˚S, is in-line with a geothermal gradient exerting a first-order longitudinal control on the depth of the frictional-plastic transition. From a geometric and linear extrapolation, we suggest that the geothermal gradient is at least three times higher in the arc-axis than in the forearc. Consequently, a higher geothermal gradient and thinner fractured crust in the arc-axis may enhance a feedback loop among frictional strain, permeability increase and convection conditions for shallow fluids to flow and/or emplace. 3. The Villarrica volcano strombolian eruption on March 3rd, 2015 did not have a significant pre-, syn- or post- eruption seismicity. However, an intense sequence, took place 45 days after the eruption at ~9 km depth in a NW-elongated cluster from the summit towards the southeast. We argue that this activity may have been induced by feeder dike cooling and contraction. 5.2 Conclusions chapter 3. 1. Dikes represent the position and spatial distribution of magmas intruding the crust, whose orientation is the result of complex mechanisms predominantly driven by the regional and local stress field and the orientation of pre-existent weakness zones (inherited faults). 2. Favourably oriented inherited fault zones and the operating stress field during the time of magmatism control transverse-to-the-orogen volcanic systems. 3. Significant differential stress fields, over favourable oriented transverse-to-the-orogen magma- pathways, are key factors to define transtensional tectonics dictating the overall feeder dikes geometry and distribution, and consequent volcano edifice up-building anatomy. Parallel andesitic–basaltic dike swarms constitute the fundamental architecture of elongated stratovolcanoes. 161 4. Tail crack geometry of Callaqui's parallel dike swarm and the en echelon array of individual Pleistocene dikes in distal domains are evidence of a right lateral component accompanying the predominant extension. 5. Progressive-forming elongated volcanoes may increasingly lead to normal-to-the-volcano- elongation steeper flanks, triggering avalanches and flank erosion. Consequently, this may reinforce along-strike eruptions. 6. Local scale gravitational effects along misoriented glacial or fluvial incisions, control successive oblique-to-the-incision dike surface-interceptions and vent topography-controlled breaching along unexpected directions, not necessarily following their underlying feeder dikes orientations. 7. A geologically stable differential stress field facilitates parallel dike swarm propagation and extrusion through long-lived juxtaposed parallel pathways, contributing to long-term morpho- structural stability of elongated volcano-architectures. 8. A positive feedback loop results among long-term inter-seismic shortening, transtension within favourable oriented inherited faults, parallel dike propagation, diachronic volcanic out-put, elongated volcano anatomy construction, and lateral steeper flank collapse (avalanches). The latter may result in local de-confinement, perturbing the balance between volcanic output and erosion. This may be reflected in a local increment of the extensional component and further dike propagation and volcanic output. 5.3 Conclusions chapter 4 1. The Pleisto-Holocene Tatara-San Pedro-Pellado volcanic complex is made up by tens of eruptions and sub-parallel Pleistocene diking organized along an ENE-trend. 2. The volcanic complex is emplaced over a heterogeneous basement architecture prompted primarily by Late Miocene east-verging folding and the intrusion of ~6 Ma old granitoids; 162 secondary, by frictionally deformation characterized by strike-slip to oblique-slip faulting contouring a transtensional interaction damage zone; the former are depicted by active intra-arc margin-parallel dextral strike-slip faulting (the Melado Fault) and by a complex array of long-lived Andean transverse strike- to oblique-slip faults (e.g. La Plata and Los Cóndores faults). In turn, the interaction damage zone made up of dense normal faulting active at least since the Pliocene, configuring a graben-like structure. 3. Relative long-lived and staged volcanic history of the TSPPC (Singer et al., 1997) is compatible in time and position with the relative long-lived transtentional faulting evolution of the here- described Tatara interaction damage zone (TDZ). Stages of erosion and lack of volcanism might be associated with decreasing strain-rate of bulk transient deformation. 4. Paleo-stress states are compatible with a stress field rotation from uniaxial compression in a strike slip regime at the flanks of the TDZ to a radial extensional regime dominated by ENE-striking normal to oblique-slip faults in the central portion of the system. 5. Results of fault-slip data inversion on the edges of the TDZ are compatible and coupled with modern stress field orientation deduced from the relative strong Mw 6.2 Melado earthquake recorded the 6th of June, 2012. 6. Results of dynamic dike-geometry inversion are compatible with the solutions obtained from fault- slip data inversion on structural sites in the Central Block of the TDZ. This suggests a causal relationship between faulting and diking. 7. Stress field inversion for suits of late Tertiary to Quaternary dike-geometry and fault-slip data in south-central Andes is a consistent technic to evaluate modern stress regimes, as the positive correlation obtained between major strike-slip faults and seismic faulting in the study area. 8. Transtentional intra-arc settings generate effective geometries for magma circulation and emplacement in the upper schizosphere. The graben-like architecture of the TDZ has driven 163 subvertical dike pathways where σ1 is subvertical and σ2 and σ3 subhorizontal, generating an effective shape for the accumulation of eruptive products. Furthermore, transtensional tectonics in the TSPPC might have controlled the emplacement of a significant amount of geothermal resources, as estimated for the Mariposa Geothermal reservoir (~320MWe-inferred). 9. The transtensional interaction damage zone here describe, is not a classical step-over structure, because of the lack of clearly defined subparallel master faults. However, calculated state of stress in the eastern side of TDZ, as depicted by the La Plata NE- and Los Condores NW-striking faults, is totally compatible with the one associated with the Melado Fault (western master strike-slip fault). This particular evidence confirms the importance of ATF in the accommodation of bulk transpressional deformation within the magmatic arc. 164 6 References Acocella, V., & Funiciello, F. (2010). Kinematic setting and structural control of arc volcanism. Earth and Planetary Science Letters, 289(1–2), 43–53. https://doi.org/10.1016/j.epsl.2009.10.027 Acocella, V., & Neri, M. (2009). Dike propagation in volcanic edifices: Overview and possible developments. Tectonophysics, 471(1–2), 67–77. https://doi.org/10.1016/j.tecto.2008.10.002 Acocella, V., Bellier, O., Sandri, L., Sébrier, M., & Pramumijoyo, S. (2018). Weak Tectono-Magmatic Relationships along an Obliquely Convergent Plate Boundary: Sumatra, Indonesia. Frontiers in Earth Science, 6(February). https://doi.org/10.3389/feart.2018.00003 Agurto, H., Rietbrock, A., Barrientos, S., Bataille, K., & Legrand, D. (2012). Seismo-tectonic structure of the Aysén Region, Southern Chile, inferred from the 2007 M w= 6.2 Aysén earthquake sequence. Geophysical Journal International, 190(1), 116–130. https://doi.org/10.1111/j.1365- 246X.2012.05507.x Aki, K. (1981). A probabilistic synthesis of precursory phenomena. In Simpson D.W. and P.G. Richards, Earthquake Prediction: An intenational review, 4, (pp. 566–574). Maurice Ewing Ser. Allmendinger, R. W., Cardozo, N. C., & Fisher, D. (2012). Structural Geology Algorithms: Vectors & Tensors. Cambridge, England: Cambridge University Press. Angelier, J. (1984). Tectonic analysis of Fault Slip Data Sets. Journal of Geophysical Research, 89, 5835–5848. Angermann, D., Klotz, J., & Reigber, C. (1999). Space-geodetic estimation of the Nazca-South America Euler vector. Earth and Planetary Science Letters, 171(3), 329–334. https://doi.org/10.1016/S0012-821X(99)00173-9 Annen, C., Blundy, J. D., & Sparks, R. S. J. (2006). The genesis of intermediate and silicic magmas in deep crustal hot zones. Journal of Petrology, 47(3), 505–539. https://doi.org/10.1093/petrology/egi084 Aoki, Y. (1999). Imaging Magma Transport During the 1997&nbsp;Seismic Swarm off the Izu Peninsula, Japan. Science, 286(5441), 927–930. https://doi.org/10.1126/science.286.5441.927 Arancibia, G., Cembrano, J., & Lavenu, A. (1999). Transpresión dextral y partición de la deformación en la Zona de Falla Liquiñe-Ofqui, Aisén, Chile (44-45°S). Revista Geológica de Chile, 26(1). https://doi.org/10.4067/S0716-02081999000100001 Aron, F., Cembrano, J., Astudillo, F., Allmendinger, R. W., & Arancibia, G. (2014). Constructing forearc architecture over megathrust seismic cycles: Geological snapshots from the Maule 165 earthquake region, Chile. Geological Society of America Bulletin, 127(3–4), 464–479. https://doi.org/10.1130/B31125.1 Aron, F., Allmendinger, R. W., Cembrano, J., González, G., & Yáñez, G. (2013). Permanent fore-arc extension and seismic segmentation: Insights from the 2010 Maule earthquake, Chile. Journal of Geophysical Research: Solid Earth, 118(2), 724–739. https://doi.org/10.1029/2012JB009339 Arriagada, C., Arancibia, G., Cembrano, J., Martínez, F., Carrizo, D., Van Sint Jan, M., … Yañez, G. (2011). Nature and tectonic significance of co-seismic structures associated with the Mw 8.8 Maule earthquake, central-southern Chile forearc. Journal of Structural Geology, 33(5), 891–897. https://doi.org/10.1016/j.jsg.2011.03.004 Astudillo, L., Cortés-aranda, J., Melnick, D., & Tassara, A. (2018). Holocene deformation along the Liquiñe – Ofqui Fault Zone , southern Chile : Field observations and geomorphic analysis. In INQUA Focus Group Earthquake Geology and Seismic Hazards (pp. 25–27). Barrientos, S. E., & Acevedo-Aránguiz, P. S. (1992). Seismological aspects of the 1988–1989 Lonquimay (Chile) volcanic eruption. Journal of Volcanology and Geothermal Research, 53(1–4), 73– 87. https://doi.org/http://dx.doi.org/10.1016/0377-0273(92)90075-O Barrientos, S., Vera, E., Alvarado, P., & Monfret, T. (2004). Crustal seismicity in central Chile, 16, 759–768. https://doi.org/10.1016/j.jsames.2003.12.001 Beck, M. E. (1983). On the mechanism of tectonic transport in zones of oblique subduction. Tectonophysics, 93(1–2), 1–11. https://doi.org/10.1016/0040-1951(83)90230-5 Beck, M. E. (1991). Coastwise transport reconsidered: lateral displacements in oblique subduction zones, and tectonic consequences. Physics of the Earth and Planetary Interiors, 68(1–2), 1–8. https://doi.org/10.1016/0031-9201(91)90002-Y Betka, P., Klepeis, K., & Mosher, S. (2016). Fault kinematics of the Magallanes-Fagnano fault system, southern Chile an example of diffuse strain and sinistral transtension along a continental transform margin. Journal of Structural Geology, 85, 130–153. https://doi.org/10.1016/j.jsg.2016.02.001 Blanquat, M. D. Saint, Tikoff, B., Teyssier, C., & Vigneresse, J. L. (1998). Transpressional kinematics and magmatic arcs. Geological Society , London , Special Publications, 135, 327–340. https://doi.org/10.1144/GSL.SP.1998.135.01.21 Bloch, W., Kummerow, J., Salazar, P., Wigger, P., & Shapiro, S. A. (2014). High-resolution image of the North Chilean subduction zone: Seismicity, reflectivity and fluids. Geophysical Journal International, 197(3), 1744–1749. https://doi.org/10.1093/gji/ggu084 Bohm, M., Lüth, S., Echtler, H., Asch, G., Bataille, K., Bruhn, C., … Wigger, P. (2002). The Southern Andes between 36° and 40°S latitude: Seismicity and average seismic velocities. Tectonophysics, 166 356(4), 275–289. https://doi.org/10.1016/S0040-1951(02)00399-2 Bommer, J. J., Bénito, B., Ciudad-Real, M., López-Menjívar, M., Mankelow, J. M., Méndez de Hasbun, P., Rodríguez Pineda, C. (2002). The El Salvador Earthquakes of 2001: Implication for seismic risk from crustal and subduction seismicity. Twelfth European Conference on Earthquake Engineering, 693, Paper No. 693. Bott, M. H. P. (1959). The Mechanics of Oblique Slip Faulting. Geological Magazine, 96(02), 109– 117. https://doi.org/10.1017/S0016756800059987 Bouvet de Maisonneuve, C., Dungan, M. a., Bachmann, O., & Burgisser, A. (2012). Insights into shallow magma storage and crystallization at Volcán Llaima (Andean Southern Volcanic Zone, Chile). Journal of Volcanology and Geothermal Research, 211–212, 76–91. https://doi.org/10.1016/j.jvolgeores.2011.09.010 Brasse, H., Kapinos, G., Li, Y., Mütschard, L., Soyer, W., & Eydam, D. (2009). Structural electrical anisotropy in the crust at the South-Central Chilean continental margin as inferred from geomagnetic transfer functions. Physics of the Earth and Planetary Interiors, 173(1–2), 7–16. https://doi.org/10.1016/j.pepi.2008.10.017 Breitkreuz, C. (1991). Permo-Carboniferous magmatism, basin development, and tectonics in the north Chilean Andes. In XII International Congress on Carboniferous and Permian Geology and Stratigraphy (pp. 17–18). Brogi, A., Liotta, D., Meccheri, M., & Fabbrini, L. (2010). Transtensional shear zones controlling volcanic eruptions: the Middle Pleistocene Mt Amiata volcano (inner Northern Apennines, Italy). Terra Nova, 22(2), 137–146. Burns, W. M., Jordan, T. E., Copeland, P., & Kelley, S. a. (2006). The case for extensional tectonics in the Oligocene-Miocene Southern Andes as recorded in the Cura Mallín basin (36° -38°S). Geological Society of America Special Papers, 407(08), 163–184. https://doi.org/10.1130/2006.2407(08). Byerlee, J. D., & Brace, W. F. (1968). Stick slip, stable sliding, and earthquakes-Effect of rock type, pressure, strain rate, and stiffness. Journal of Geophysical Research, 73(18), 6031–6037. https://doi.org/10.1029/JB073i018p06031 Cardona, C., Tassara, A., Gil-Cruz, F., Lara, L., Morales, S., Kohler, P., & Franco, L. (2018). Crustal seismicity associated to rapid surface uplift at Laguna del Maule Volcanic Complex, Southern Volcanic Zone of the Andes. Journal of Volcanology and Geothermal Research, 353(February), 83– 94. https://doi.org/10.1016/j.jvolgeores.2018.01.009 Carpinelli, A. (2000). Analisis estratigrafico, paleoambiental, estructural y modelo tectono- estratigrafico de la Cuenca Cura-Mallin VIII y IX Region. Universidad de Concepción. 167 Cembrano, J., Lavenu, A., Yañez, G., Riquelme, R., García, M., González, G., & Hérail, G. (2005). Neotectonics. In Geology of Chile (pp. 231–261). Cembrano, J., & Herve, F. (1993). The Liquine Ofqui Fault Zone : a major Cenozoic strike slip duplex in the Southern Andes. Secoud ISAG, 175–178. Cembrano, J., & Lara, L. (2009). The link between volcanism and tectonics in the southern volcanic zone of the Chilean Andes: A review. Tectonophysics, 471(1–2), 96–113. https://doi.org/10.1016/j.tecto.2009.02.038 Cembrano, J., Herve, F., & Lavenu, A. (1996). The Liquifie Ofqui fault zone: a long-lived intra-arc fault system in southern Chile. Tectonophysics, 256, 55–56. Cembrano, J., Schermer, E., Lavenu, A., & Sanhueza, A. (2000). Contrasting nature of deformation along an intra-arc shear zone , the Liquiñe – Ofqui fault zone , southern Chilean Andes. Tectonophysics, 319, 129–149. https://doi.org/10.1016/S0040-1951(99)00321-2 Cembrano, J., Lavenu, A., Reynolds, P., Arancibia, G., López, G., & Sanhueza, A. (2002). Late Cenozoic transpressional ductile deformation north of the Nazca – South America – Antarctica triple junction. Tectonophysics, 354, 289–314. Charrier, R. (2007). Cenozoic magmatism and deformation in the northern and central Chilean Andes : differing paths in the construction of the modern orogen. Charrier, R., Wyss, A. R., & Flynn, J. J. (1996). New evidences for Late Mesozoic-Early Cenozoic evolution of the Chilean Andes in the Upper Tinguiririca Valley (35oS), Central Chile. Journal of South American Earth Sciences, 9, 393–422. Charrier, R., Pinto, L., & Rodriguez, M. P. (2007). Tectonostratigraphic evolution of the Andean Orogen in Chile. In Geology of Chile (pp. 21–114). Chesley, C., LaFemina, P. C., Puskas, C., & Kobayashi, D. (2012). The 1707 Mw8.7 Hoei earthquake triggered the largest historical eruption of Mt. Fuji. Geophysical Research Letters, 39(24), 1–8. https://doi.org/10.1029/2012GL053868 Chinn, D. S., & Isacks, B. L. (1983). Accurate source depths and focal mechanisms of shallow earthquakes in western South America and in the New Hebrides island arc. Tectonics, 2, 529–563. Chouet, B. (2003). Volcano Seismology. Pure and Applied Geophysics, 160, 739–788. Clemens, J. D. (1998). Observations on the origins and ascent mechanisms of granitic magmas, 155, 843–851. https://doi.org/10.1144/gsjgs.155.5.0843 Clemens, J. D., & Petford, N. (1999). Granitic melt viscosity and silicic magma dynamics in contrasting tectonic settings, 156(1972), 1057–1060. https://doi.org/10.1144/gsjgs.156.6.1057 168 Corazzato, C., & Tibaldi, A. (2006). Fracture control on type , morphology and distribution of parasitic volcanic cones : An example from Mt . Etna , Italy. https://doi.org/10.1016/j.jvolgeores.2006.04.018 Corbett, G. J., & Leach, T. M. (1998). Southwest Pacific rim gold–copper systems: structure, alteration and mineralization. Society of Economic Geologists, Special Publication, 6., (May 1997), 236. Corti, G., Moratti, G., & Sani, F. (2005). Relations between surface faulting and granite intrusions in analogue models of strike-slip deformation, 27, 1547–1562. https://doi.org/10.1016/j.jsg.2005.05.011 Costa, C. H., Audemard, F. a M., Bezerra, F. H. R., Lavenu, A., Machette, M. N., & París, G. (2006). An overview of the main Quaternary deformation of South America. Revista de La Asociacion Geologica Argentina, 61(4), 461–479. Cowie, P. A., & Scholz, C. H. (1992). Growth of faults by accumulation of seismic slip. Journal of Geophysical Research, 97, 11085–11095. Cox, S. F. (1999). Deformational controls on the dynamics of fluid flow in mesothermal gold systems. Geological Society, London, Special Publications, 155(1), 123–140. https://doi.org/10.1144/GSL.SP.1999.155.01.10 Cox, S. F. (2010). The application of failure mode diagrams for exploring the roles of fluid pressure and stress states in controlling styles of fracture-controlled permeability enhancement in faults and shear zones. Geofluids, 217–233. https://doi.org/10.1111/j.1468-8123.2010.00281.x Davidson, J. P., Ferguson, K. M., Colucci, M. T., & Dungan, M. A. (1988). The origin and evolution of magmas from the San Pedro–Pellado volcanic complex, south Chile: Multiple sources and open system evolution. Contributions to Mineralogy and Petrology, 100, 429–445. Delaney, P. T., Pollard, D. D., Ziony, I., & Mckee, E. H. (1986). Field Relations Between Dikes and Joints: Emplacement Processes and Paleostress Analysis. Journal of Geophysical Research, 91(5), 4920–4938. Déruelle, B., Nini, J., & Kambou, R. (1987). Mount Cameroon e an active volcano of the Cameroon line. Journal of African Earth Sciences, 6(2), 197–214. Dewey, J. F., & Lamb, S. H. (1992). Active tectonics of the Andes. Tectonophysics, 205, 79–95. Drake, R. E. (1976). The chronology of Cenozoic igneous and tectonic events in the central Chilean Andes. In: Proceedings of the sym- posium on Andean and Antarctic volcanology problems, Santiago, Chile, September 1974. Int Assoc Volcanol and Chem Earth’s Interior, Santiago,. In Symposium of Andean and Antartic volcanology problems - IAVCEI - Chile (pp. 670–697). 169 Dumond, G., Yoshinobu, A. S., & Barnes, C. G. (2005). Midcrustal emplacement of the Sausfjellet pluton, central Norway: Ductile flow, stoping, and in situ assimilation. Bulletin of the Geological Society of America, 117(3–4), 383–395. https://doi.org/10.1130/B25464.1 Dungan, M., Wulff, a, & Thompson, R. (2001). Eruptive Stratigraphy of the Tatara-San Pedro Complex, 36°S, Southern Volcanic Zone, Chilean Andes: reconstruction method and implications for magma evolution at long-lived Arc Volcanic Centers. Journal of Petrology, Vol. 42, No., 3p(3), 555–626. https://doi.org/10.1093/petrology/42.3.555 Dzierma, Y., Rabbel, W., Thorwart, M., Koulakov, I., Wehrmann, H., Hoernle, K., & Comte, D. (2012a). Seismic velocity structure of the slab and continental plate in the region of the 1960 Valdivia (Chile) slip maximum - Insights into fluid release and plate coupling. Earth and Planetary Science Letters, 331–332, 164–176. https://doi.org/10.1016/j.epsl.2012.02.006 Dzierma, Y., Thorwart, M., Rabbel, W., Siegmund, C., Comte, D., Bataille, K., Prezzi, C. (2012b). Seismicity near the slip maximum of the 1960 Mw 9.5 Valdivia earthquake (Chile): Plate interface lock and reactivation of the subducted Valdivia Fracture Zone. Journal of Geophysical Research: Solid Earth, 117(6), 1–13. https://doi.org/10.1029/2011JB008914 Dziewonski, a. M., Ekström, G., Woodhouse, J. H., & Zwart, G. (1990). Centroid-moment tensor solutions for January–March 1989. Physics of the Earth and Planetary Interiors, 59(4), 233–242. https://doi.org/10.1016/0031-9201(90)90232-M Emerman, S. H., & Marrett, R. (1990). Why dikes? Geology, 18(3), 231. https://doi.org/10.1130/0091-7613(1990)018<0231 Engdhal, E., & Villaseñor, A. (2002). Global seismicity: 1900-1999. In W. H. K. et al. lee (Ed.), International Handbook of Earthquake and Engineering Seismology (p. vol. 81, 665-690). Int. Geophys. Ser. Farías, M., Comte, D., Charrier, R., Martinod, J., David, C., Tassara, A., Fock, A. (2010). Crustal-scale structural architecture in central Chile based on seismicity and surface geology: Implications for Andean mountain building. Tectonics, 29(3). https://doi.org/10.1029/2009TC002480 Farías, M., Comte, D., Roecker, S., Carrizo, D., & Pardo, M. (2011). Crustal extensional faulting triggered by the 2010 Chilean earthquake: The Pichilemu Seismic Sequence. Tectonics, 30(6), 1– 11. https://doi.org/10.1029/2011TC002888 Faye, G. D., Yamaji, A., Yonezu, K., Tindell, T., & Watanabe, K. (2018). Paleostress and fluid- pressure regimes inferred from the orientations of Hishikari low sulfidation epithermal gold veins in southern Japan. Journal of Structural Geology, 110(March), 131–141. https://doi.org/10.1016/j.jsg.2018.03.002 Feigl, K. L., Le Mével, H., Ali, S. T., Córdova, L., Andersen, N. L., Demets, C., & Singer, B. S. (2014). 170 Rapid uplift in Laguna del Maule volcanic field of the Andean Southern Volcanic zone (Chile) 2007– 2012. Geophysical Journal International, 196(2), 885–901. La Femina, P. C. La, Dixon, T. H., & Strauch, W. (2002). Bookshelf faulting in Nicaragua, Geology, 30(8), 751–754. Fiske, R. S., & Jackson, E. D. (1972). Orientation and growth of Hawaiian volcanic rifts: the effect of regional structure and gravitational stresses. Philosophical Transactions of the Royal Society of London, 329, 299–326. Fitch, T. J. (1972). Plate Convergence, Transcurrent Faults, and Internal Deformation Adjacent to Southeast Asia and the Western Pacific. Journal of Geophysical Research, 77(23), 4432–4460. Flynn, J. J., Wyss, A. R., Croft, D. A., & Charrier, R. (2003). The Tinguiririca Fauna, Chile: biochronology, paleoecology, biogeography, and a new earliest Oligocene South American Land Mammal “Age.” Paleogeography, Paleoclimatology, Paleoecology, 195(3–4), 229–259. Folguera, A., Ramos, V. a., Hermanns, R. L., & Naranjo, J. (2004). Neotectonics in the foothills of the southernmost central Andes (37°-38°S): Evidence of strike-slip displacement along the Antiñir-Copahue fault zone. Tectonics, 23(5), 1–23. https://doi.org/10.1029/2003TC001533 Foulger, G. R., & Long, R. E. (1992). Non-double couple earthquake focal mechanism. In Gasparini, P., R. Scarpa, and K. Aki (Eds.), Volcanic Seismology. Springer-Verlag. https://doi.org/10.1007/978-3- 642-77008-1 Fourmier, T. J., Pritchard, M. E., & Riddick, S. N. (2010). Duration, magnitude, and frequency of subaerial volcano deformation events: new results from Latin America using InSAR a global synthesis. Geochemistry, Geophysics, Geosystems2, 11(1), 1–29. Frey, F. A., Gerlach, D. C., Hickey, R. L., & López, L. (1984). Petrogenesis of the Laguna del Maule volcanic complex, Chile (36°S). Contributions to Mineralogy and Petrology, 88, 133–149. Glodny, J., Echtler, H., Collao, S., Ardiles, M., Burón, P., & Figueroa, O. (2008). Differential Late Paleozoic active margin evolution in South-Central Chile (37°S-40°S) - the Lanalhue Fault Zone. Journal of South American Earth Sciences, 26(4), 397–411. https://doi.org/10.1016/j.jsames.2008.06.001 Gomila, R., Arancibia, G., Mitchell, T. M., Cembrano, J. M., & Faulkner, D. R. (2016). Palaeopermeability structure within fault-damage zones: A snap-shot from microfracture analyses in a strike-slip system. Journal of Structural Geology, 83, 103–120. https://doi.org/10.1016/j.jsg.2015.12.002 González, O., & Vergara, M. (1962). Reconocimiento geológico de la Cordillera de Los Andes entre los paralelos 35o y 38o Lat. Sur. 171 Gratier, J.-P. (2011). Fault Permeability and Strength Evolution Related to Fracturing and Healing Episodic Processes (Years to Millennia): the Role of Pressure Solution. Oil & Gas Science and Technology – Revue d’IFP Energies Nouvelles, 66(3), 491–506. https://doi.org/10.2516/ogst/2010014 Griffin, W. L., & O’Reilly, S. Y. (1987). Is the continental Moho the crust-mantle boundary? Geology, 15(3), 241–244. https://doi.org/10.1130/0091-7613(1987) Groeber, P. (1921). La region de Copahue y su glaciacion diluvial. Revista de La Asociacion Geologica Argentina, 1, 92–110. Gudmundsson, A. (2011). Rock Fractures in Geological Processes. https://doi.org/10.1017/CBO9780511975684 Haberland, C., Rietbrock, A., Lange, D., Bataille, K., & Dahm, T. (2009). Structure of the seismogenic zone of the southcentral Chilean margin revealed by local earthquake traveltime tomography, 114, 1–17. https://doi.org/10.1029/2008JB005802 Haberland, C., Rietbrock, A., Lange, D., Bataille, K., & Hofmann, S. (2006). Interaction between forearc and oceanic plate at the south-central Chilean margin as seen in local seismic data, 33(June 2005), 1–5. https://doi.org/10.1029/2006GL028189 Hauser, A. (1997). Catastro aguas termales Chile.pdf (Boletín No). Servicio Nacional de Geología y Minería. Hickson, C. J., Ferraris, F., Rodriquez, C., Sielfeld, G., Henriquez, R., Gislason, T., … Yehia, R. (2011). The Mariposa geothermal system, Chile. In Transactions - Geothermal Resources Council (Vol. 35 1, pp. 817–825). Hildreth, W., & Moorbath, S. (1988). Crustal contributions to arc magmatism in the Andes of Central Chile. Contributions to Mineralogy and Petrology, 98(4), 455–489. https://doi.org/10.1007/BF00372365 Hildreth, W., Fierstein, J., Godoy, E., Drake, R. E., & Singer, B. (1999). The Puelche Volcanic Field: extensive Pleistocene rhyolite lava flows in the Andes of central Chile. Revista Geologica De Chile, 26(July 2016), 275–309. https://doi.org/10.4067/S0716-02081999000200008 Hildreth, W., Godoy, E., Fierstein, J., & Singer, B. S. (2010). Laguna del Maule volcanic field: Eruptive history of a Quaternary basalt to rhyolite distributed volcanic field on the Andean range crest in central Chile. Servicio Nacional de Geologia y Mineria, Bolletin, 63(63), 145. Hill, D. P. (1977). A model for earthquake swarms. Journal of Geophysical Research, 82(8), 1347– 1352. https://doi.org/10.1029/JB082i008p01347 Hou, G. (2012). Mechanism for three types of mafic dike swarms. Geoscience Frontiers, 3(2), 217– 223. https://doi.org/10.1016/j.gsf.2011.10.003 172 Husen, S., Kissling, E., Flueh, E., & Asch, G. (1999). Accurate hypocentre determination in the seismogenic zone of the subducting Nazca Plate in northern Chile using a combined on- / offshore network, 687–701. Iturrieta, P., Hurtado, D. E., Cembrano, J., & Stanton-yonge, A. (2017). States of stress and slip partitioning in a continental scale strike-slip duplex : Tectonic and magmatic implications by ... Earth and Planetary Science Letters, 473(May), 71–82. https://doi.org/10.1016/j.epsl.2017.05.041 Jarrard, R. D. (1986). Terrane motion by strike-slip faulting of forearc slivers. Geology, 14(9), 780– 783. https://doi.org/10.1130/0091-7613(1986)14<780:TMBSFO>2.0.CO Jay, J. A., Pritchard, M. E., West, M. E., Christensen, D., Haney, M., Minaya, E., … Zabala, M. (2011). Shallow seismicity , triggered seismicity , and ambient noise tomography at the long-dormant Uturuncu Volcano , Bolivia, 817–837. https://doi.org/10.1007/s00445-011-0568-7 Jensen, E., Cembrano, J., Faulkner, D., Veloso, E., & Arancibia, G. (2011). Development of a self- similar strike-slip duplex system in the Atacama Fault system, Chile. Journal of Structural Geology, 33(11), 1611–1626. https://doi.org/10.1016/j.jsg.2011.09.002 Jolly, R. J. H., & Sanderson, D. J. (1997). A Mohr circle construction for the opening of a pre-existing fracture. Journal of Structural Geology, 19(6), 887–892. https://doi.org/10.1016/S0191- 8141(97)00014-X Jordan, T. E., Matthew Burns, W., Veiga, R., Pángaro, F., Copeland, P., Kelley, S., & Mpodozis, C. (2001). Extension and basin formation in the southern Andes caused by increased convergence rate: A mid-Cenozoic trigger for the Andes. Tectonics, 20(3), 308–324. https://doi.org/10.1029/1999TC001181 de Joussineau, G., Aydin, A., Geophysics, A., de Joussineau, G., & Aydin, A. (2009). Segmentation along Strike-Slip Faults Revisited. Pure and Applied Geophysics, 166(10–11), 1575–1594. https://doi.org/10.1007/s00024-009-0511-4 Kanamori, H., & Stewart, G. S. (1979). A SLOW EARTHQUAKE HIROO KANAMORI and GORDON S. STEWART. Readings, 18, 167–175. https://doi.org/10.1016/0031-9201(79)90112-2 Katz, H. R. (1971). Continental margin in Chile; is tectonic style compressional or extensional? AAPG Bulletin , 55(10), 1753–1758. Kay, S. M., Mpodozis, C., & Ramos, V. A. (2005). Andes. Encyclopedia of Geology, 118. https://doi.org/10.1016/B978-0-7020-2863-2.50024-7 Kervyn, M., Vries, B. W. De, Walter, T. R., Njome, M. S., Suh, C. E., & Ernst, G. G. J. (2014). Directional flank spreading at Mount Cameroon volcano: Evidence from analogue modeling. Journal of Geophysical Research : Solid Earth, 119. https://doi.org/10.1002/2014JB011330.Received 173 Kim, Y. S., Peacock, D. C. P., & Sanderson, D. J. (2004). Fault damage zones. Journal of Structural Geology, 26(3), 503–517. https://doi.org/10.1016/j.jsg.2003.08.002 Kimura, G. (1986). Oblique subduction and collision : Forearc tectonics of the Kuril arc. Geology, 14, 404–407. Kissling, E. (1988). Geotomography with local earthquake data. Reviews of Geophysics, 26(4), 659– 698. Kissling, E., Ellsworth, W. L., Eberhardt-Phillips, D., & Kradolfer, U. (1994). Initial reference model in local earthquake tomography. Journal of Geophysical Research, 99, 19,635-19,646. Kisslinger, C. (1980). Evaluation of S to P amplitude ratios for determining focal mechanisms from regional network observations. Bulletin of the Seismological Society of Americ, 70, 999–1014. Koulakov, I., Sobolev, S. V., & Asch, G. (2006). P - And S-velocity images of the lithosphere- asthenosphere system in the Central Andes from local-source tomographic inversion. Geophysical Journal International, 167(1), 106–126. https://doi.org/10.1111/j.1365- 246X.2006.02949.x Lagmay, a. M. F., & Valdivia, W. (2006). Regional stress influence on the opening direction of crater amphitheaters in Southeast Asian volcanoes. Journal of Volcanology and Geothermal Research, 158(1–2), 139–150. https://doi.org/10.1016/j.jvolgeores.2006.04.020 Lahsen, A., Muñoz, N., & Parada, M. A. (2010). Geothermal Development in Chile. World Geothermal Congress, (April), 25–29. Lange, D., Cembrano, J., Rietbrock, a., Haberland, C., Dahm, T., & Bataille, K. (2008). First seismic record for intra-arc strike-slip tectonics along the Liquiñe-Ofqui fault zone at the obliquely convergent plate margin of the southern Andes. Tectonophysics, 455(1–4), 14–24. https://doi.org/10.1016/j.tecto.2008.04.014 Lange, D., Tilmann, F., Barrientos, S. E., Contreras-reyes, E., Methe, P., Moreno, M., … Beck, S. (2012). Aftershock seismicity of the 27 February 2010 Mw 8 . 8 Maule earthquake rupture zone. Earth and Planetary Science Letters, 317–318(February 2010), 413–425. https://doi.org/10.1016/j.epsl.2011.11.034 Lange, D., Rietbrock, A., Haberland, C., Bataille, K., Dahm, T., Tilmann, F., & Flu, E. R. (2007). Seismicity and geometry of the south Chilean subduction zone ( 41 . 5 ° S – 43 . 5 ° S ): Implications for controlling parameters, 34, 1–5. https://doi.org/10.1029/2006GL029190 Lara, L. E., Cembrano, J., & Lavenu, a. (2008). Quaternary Vertical Displacement along the Liquiñe- Ofqui Fault Zone: Differential Uplift and Coeval Volcanism in the Southern Andes? International Geology Review, 50(11), 975–993. https://doi.org/10.2747/0020-6814.50.11.975 174 Lara, L. E., Orozco, G., & Piña-Gauthier, M. (2012). The 1835 AD fissure eruption at Osorno volcano, Southern Andes: tectonic control by the intraarc stress field instead of remote megathrust-related dynamic strain. Tectonophysics, 102–110. https://doi.org/10.1016/j.tecto.2011.11.28 Lara, L., Naranjo, J., & Moreno, H. (2004). Rhyodacitic fissure eruption in Southern Andes (Cordón Caulle; 40.5°S) after the 1960 (Mw:9.5) Chilean earthquake: a structural interpretation. Journal of Volcanology and Geothermal Research, 138(1–2), 127–138. https://doi.org/10.1016/j.jvolgeores.2004.06.009 Lara, L., Lavenu, A., Cembrano, J., & Rodriguez, C. (2006). Structural controls of volcanism in transversal chains: Resheared faults and neotectonics in the Cordón Caulle–Puyehue area (40.5°S), Southern Andes. Journal of Volcanology and Geothermal Research, 158(1–2), 70–86. https://doi.org/10.1016/j.jvolgeores.2006.04.017 Lavenu, a, & Cembrano, J. (1999). Compressional and transpressional stress pattern for the Pliocene and Quaternary (Andes of central and southern Chile). Journal of Structural Geology, 21, 1669–1691. https://doi.org/10.1016/S0191-8141(99)00111-X Lavenu, A., & Cembrano, J. (1999). Compressional- and transpressional-stress pattern for Pliocene and Quaternary brittle deformation in fore arc and intra-arc zones (Andes of Central and Southern Chile). Journal of Structural Geology, 21(12), 1669–1691. https://doi.org/10.1016/S0191-8141(99)00111-X Legrand, D., Barrientos, S., Bataille, K., Cembrano, J., & Pavez, a. (2011). The fluid-driven tectonic swarm of Aysen Fjord, Chile (2007) associated with two earthquakes (Mw=6.1 and Mw=6.2) within the Liquiñe-Ofqui Fault Zone. Continental Shelf Research, 31(3–4), 154–161. https://doi.org/10.1016/j.csr.2010.05.008 Lienert, B. R. E., Berg, E., & Frazer, L. N. (1986). Hypocenter: An earthquake location method using centered, scaled, and adaptively least squares. Bulletin of the Seismological Society of America, 76, 771–783. Lienert, B. R. E., & Havskov, J. (1995). A computer program for locating earthquakes both locally and globally. Seismological Research Letters, 66, 26–36. Lira, E., Sielfeld, G. G., Bosch, A., Castellón, R., Cumming, D., Figueroa, R., … Yañez, G. (2015). Reactivación de fallas de larga vida transversales al orógeno andino. Perspectivas desde la Falla Pichilemu. In Congreso Geológico Chileno. Lister, J. R., & Kerr, R. C. (1991). Fluid-mechanical models of crack propagation and their application to magma transport in dikes. Journal of Geophysical Research, 96(B6), 10049–10077. https://doi.org/10.1029/91JB00600 Lomax, A., Virieux, J., Volant, P., & Berge-Thierry, C. (2000). Probabilistic earthquake location in 3D 175 and layered models. In C. H. Thurber & N. Rabinowitz (Eds.), Advances is seismic event location (pp. 101–134). Springer-Science+Buiness Media B.V. López-Escobar, L., Cembrano, J., & Moreno, H. (1995). Geochemistry and tectonics of the Chilean Southern Andes basaltic Quaternary volcanism. Andean Geology, 22(2), 219–234. Retrieved from file:///C:/Users/User/Dropbox/Publicaciones/pdf L-Z/López_Escobar et al. 1995.pdf Lupi, M., & Miller, S. a. (2014). Short-lived tectonic switch mechanism for long-term pulses of volcanic activity after mega-thrust earthquakes. Solid Earth, 5(1), 13–24. https://doi.org/10.5194/se-5-13-2014 Manga, M., & Brodsky, E. (2006). Seismic Triggering of Eruptions in the Far Field : Volcanoes and Geysers. https://doi.org/10.1146/annurev.earth.34.031405.125125 Marrett, R., & Allmendinger, R. W. (1990). Kinematic analysis of fault-slip data. Journal of Structural Geology, 12(8), 973–986. https://doi.org/10.1016/0191-8141(90)90093-E Martinod, J., Husson, L., Roperch, P., Guillaume, B., & Espurt, N. (2010). Horizontal subduction zones , convergence velocity and the building of the Andes. Earth and Planetary Science Letters, 299(3–4), 299–309. https://doi.org/10.1016/j.epsl.2010.09.010 Mathieu, L., van Wyk de Vries, B., Pilato, M., & Troll, V. R. (2011). The interaction between volcanoes and strike-slip, transtensional and transpressional fault zones: Analogue models and natural examples. Journal of Structural Geology, 33(5), 898–906. https://doi.org/10.1016/j.jsg.2011.03.003 Mazzini, A., Nermoen, A., Krotkiewski, M., Podladchikov, Y., Planke, S., & Svensen, H. (2009). Strike- slip faulting as a trigger mechanism for overpressure release through piercement structures . Implications for the Lusi mud volcano , Indonesia. Marine and Petroleum Geology, 26(9), 1751– 1765. https://doi.org/10.1016/j.marpetgeo.2009.03.001 Mccaffrey, K. J. W. (1992). Igneous emplacement in a transpressive shear zone : Ox Mountains igneous complex, 149, 221–235. McCaffrey, R. (1996). Estimates of modern arc-parallel strain rates in fore arcs. Geology, 24(1), 27– 30. https://doi.org/10.1130/0091-7613(1996)024<0027:EOMAPS>2.3.CO;2 McCloskey, J., Nalbant, S. S., & Steacy, S. (2005). Earthquake risk from co-seismic stress. Nature, 434(7031), 291. https://doi.org/10.1038/434291a Melnick, D., & Echtler, H. P. (2006). Morphotectonic and geologic digital map compilations of the south-central Andes (36°-42°S). In O. Oncken & et al. (Eds.), The Andes - Active Subduction Orogeny (pp. 565–568). Springer-Verlag, Berlin Heidelberg - New York. Melnick, D., Bookhagen, B., Strecker, M. R., & Echtler, H. P. (2009). Segmentation of megathrust 176 rupture zones from fore-arc deformation patterns over hundreds to millions of years, Arauco peninsula, Chile. Journal of Geophysical Research: Solid Earth, 114(1). https://doi.org/10.1029/2008JB005788 Melnick, D., Folguera, A., & Ramos, V. (2006). Structural control on arc volcanism: The Caviahue– Copahue complex, Central to Patagonian Andes transition (38°S). Journal of South American Earth Sciences, 22(1–2), 66–88. https://doi.org/10.1016/j.jsames.2006.08.008 Mescua, J. F., Giambiagi, L. B., & Ramos, V. A. (2013). Late Cretaceous Uplift in the Malargüe fold- and-thrust belt ( 35oS ), southern Central Andes of Argentina and Chile. Andean Geology, 40(1), 102–116. https://doi.org/10.5027/andgeoV40n1-a05 Mitchell, T. M., & Faulkner, D. R. (2009). The nature and origin of off-fault damage surrounding strike-slip fault zones with a wide range of displacements: A field study from the Atacama fault system, northern Chile. Journal of Structural Geology, 31(8), 802–816. https://doi.org/10.1016/j.jsg.2009.05.002 Mora-Stock, C., Thorwart, M., Wunderlich, T., Bredemeyer, S., Hansteen, T. H., & Rabbel, W. (2012). Comparison of seismic activity for Llaima and Villarrica volcanoes prior to and after the Maule 2010 earthquake. International Journal of Earth Sciences, 103(7), 2015–2028. https://doi.org/10.1007/s00531-012-0840-x Moreno, H., Thiele, R., Lahsen, A., Varela, J., & Lopez-Escobar, L. (1984). Estudio del Volcan Callaqui: Geología y Riesgo Volcanico. Moreno, M., Melnick, D., Rosenau, M., Baez, J., Klotz, J., Oncken, O., Hase, H. (2012). Toward understanding tectonic control on the Mw8.8 2010 Maule Chile earthquake. Earth and Planetary Science Letters, 321–322, 152–165. https://doi.org/10.1016/j.epsl.2012.01.006 Moreno, M., Melnick, D., Rosenau, M., Bolte, J., Klotz, J., Echtler, H., … Oncken, O. (2011). Heterogeneous plate locking in the South-Central Chile subduction zone: Building up the next great earthquake. Earth and Planetary Science Letters, 305(3–4), 413–424. https://doi.org/10.1016/j.epsl.2011.03.025 Moreno, M. S., Klotz, J., Melnick, D., Echtler, H., & Bataille, K. (2008). Active faulting and heterogeneous deformation across a megathrust segment boundary from GPS data, south central Chile (36–39 ° S), (November). https://doi.org/10.1029/2008GC002198 Moreno, M. S., Bolte, J., Klotz, J., & Melnick, D. (2009). Impact of megathrust geometry on inversion of coseismic slip from geodetic data : Application to the 1960 Chile earthquake, 36(May), 1–5. https://doi.org/10.1029/2009GL039276 Mori, J., White, R. A., Harlow, D. H., Okubo, P., Power, J. A., Hoblitt, R. P., … Bautista, B. C. (1991). Volcanic Earthquakes following the 1991 Climatic Eruption of Mount Pinatubo: Strong Seismicity during a Waning Eruption. U.S. Geological Survay Publications, 177 Https://Pubs.Usgs.Gov/Pinatubo/Mori1/. Moser, T. J., van Eck, T., & Nolet, G. (1992). Hypocenter determination in strongly heterogeneous earth models using the shortest path method. Journal of Geophysical Research, 97, 6563–6572. Muñoz, B. J., & Niemeyer, R. H. (1984). Hoja Laguna del Maule. Nacif, S., Lupari, M., Triep, E. G., Nacif, A., Álvarez, O., Folguera, A., & Gímenez, M. (2017). Tectonophysics Change in the pattern of crustal seismicity at the Southern Central Andes from a local seismic network, 708, 56–69. https://doi.org/10.1016/j.tecto.2017.04.012 Nakamura, K. (1977). Volcanoes as possible indicators of tectonic stress orientation - principle and proposal. Journal of Volcanology and Geothermal Research, 2, 1–16. https://doi.org/10.1007/BF01637099 Naranjo, J. A. (2015). Nuevo estilo eruptivo del volcán Villarrica : 3 de marzo 2015. In XIV Congreso Geológico Chileno (pp. 2–4). Naranjo, J. A., & Moreno, H. (2005). Geología del volcán Llaima, Región de la Araucanía. Servicio Nacional de Geología y Minería, Carta Geol_ogica de Chile. Nelson, S. T., Davidson, J. P., Heizler, M. T., & Kowallis, B. J. (1999). Tertiary tectonic history of the southern Andes: The subvolcanic sequence to the Tatara-San Pedro volcanic complex, lat 36??S. Bulletin of the Geological Society of America, 111(9), 1387–1404. https://doi.org/10.1130/0016- 7606(1999)111<1387:TTHOTS>2.3.CO;2 Ollier, C. (1988). Volcanoes. Basil Blackwell, New York. Otsuki, K., Minagawa, J., Aono, M., & Ohtake, M. (1997). On the Curved at the Striations of Nojima Seismic Fault Engraved Hyogoken-Nambu. Zisin (Journal of the Seismological Society of Japan. 2nd Ser.), 49(4), 451–460. Ottemöller, L., & Havskov, J. (2003). Moment magnitude determination for local and regional earthquakes based on source spectra. Bulletin of the Seismological Society of America, 93(1), 203– 214. https://doi.org/10.1785/0120010220 Ottemöller, L., Voss, P., & Havskov, J. (2014). SEISAN EARTHQUAKE ANALYSIS SOFTWARE FOR WINDOWS, SOLARIS, LINUX and MACOSX. Pankhurst, R. J., Rapela, C. W., Fanning, C. M., & Márquez, M. (2006). Gondwanide continental collision and the origin of Patagonia, 76, 235–257. https://doi.org/10.1016/j.earscirev.2006.02.001 Pardo-Casas, F., & Molnar, P. (1987). Relative Motion of The Nazca (Farellon) and South American Plates Since late Cretaceous Time, 6(3), 233–248. Patanè, D., Nazionale, I., & Osservatorio, C. (2011). Interplay between Tectonics and Mount Etna’ 178 s Volcanism : Insights into the Geometry of the Plumbing System. Payne, S. J., Hackett, W. R., & Smith, R. P. (1996). Chapter 4 Paleoseismology of volcanic environments. International Geophysics (2nd ed., Vol. 62). Elsevier Inc. https://doi.org/10.1016/S0074- 6142(96)80071-4 Peacock, D. C. P., Nixon, C. W., Rotevatn, A., Sanderson, D. J., & Zuluaga, L. F. (2016). Glossary of fault and other fracture networks. Journal of Structural Geology, 92, 12–29. https://doi.org/10.1016/j.jsg.2016.09.008 Pérez-Flores, P., Veloso, E., Cembrano, J., Sánchez-Alfaro, P., Lizama, M., & Arancibia, G. (2017). Fracture network, fluid pathways and paleostress at the Tolhuaca geothermal field. Journal of Structural Geology, 96. https://doi.org/10.1016/j.jsg.2017.01.009 Pérez-Flores, P., Cembrano, J., Sanchez, P., Veloso, E., Arancibia, G., & Roquer, T. (2016). Tectonics, magmatism and paleo-fluid distribution in a strike-slip setting: Insights from the northern termination of the Liquiñe-Ofqui fault System, Chile. Tectonophysics, 680, 192–210. https://doi.org/10.1016/j.tecto.2016.05.016 Petford, N., Cruden, a R., McCaffrey, K. J., & Vigneresse, J. L. (2000). Granite magma formation, transport and emplacement in the Earth’s crust. Nature, 408(6813), 669–673. https://doi.org/10.1038/35047000 Petit, J. P. (1987). Criteria for the sense o f movement on fault surfaces in brittle rocks. Journal of Structural Geology, 9(5), 597–608. https://doi.org/10.1016/0191-8141(87)90145-3 Pinel, V. C., & Jaupart. (2003). Magma chamber behavior beneath a volcanic edifice. Journal of Geophysical Research, 108(B2), 1–17. https://doi.org/10.1029/2002JB001751 Piquer, J., Berry, R. F., & Cooke, D. R. (2016). Arc-oblique fault systems : Their role in the Cenozoic structural evolution and metallogenesis of the Andes of central Chile. Journal of Structural Geology, 89(June), 101–117. https://doi.org/10.1016/j.jsg.2016.05.008 Polanco, E., & Naranjo, A. (2008). Colapso Holoceno en el Volcán Callaqui ( 37o55 ’ S ), Andes del Sur SUR. In Congreso Geológico Argentino (pp. 1157–1158). Pollitz, F. F., Banerjee, P., Bürgmann, R., Hashimoto, M., & Choosakul, N. (2006). Stress changes along the Sunda trench following the 26 December 2004 Sumatra-Andaman and 28 March 2005 Nias earthquakes. Geophysical Research Letters, 33(6), 26–29. https://doi.org/10.1029/2005GL024558 Potent, S. (2003). Kinematik und Dynamik neogener Deformationsprozesse des südzentralchilenischen Subduktionssystems, nördlichste Patagonische Anden Pritchard, M. E., Jay, J. a., Aron, F., Henderson, S. T., & Lara, L. E. (2013). Subsidence at southern 179 Andes volcanoes induced by the 2010 Maule, Chile earthquake. Nature Geoscience, 6(8), 632–636. https://doi.org/10.1038/ngeo1855 Radic, J. P. (2010a). Las cuencas cenozoicas y su control en el volcanismo de los Complejos Nevados de Chillán y Copahue-Callaqui (Andes del sur, 36-39??S). Andean Geology, 37(1), 220–246. https://doi.org/10.4067/S0718-71062010000100009 Rapela, C. W., Pankhurst, R. J., Fanning, C. M., & Grecco, L. E. (2003). Basement evolution of the Sierra de la Ventana Fold Belt: new evidence for Cambrian continental rifting along the southern margin of Gondwana. Journal of the Geological Society, London, 160, 613–628. Reasenberg, P., & Oppenheimer, D. (1985). FPFIT, FPLOT and FPPAGE: fortran computer programs for calculating and displaying earthquake fault-plane solutions. U.S. Geological Survey. Reid, H. F. (1913). Sudden erath-movements in Sumatra in 1892. Bulletin of the Seismological Society of America, 3, 72–79. Renne, P. R., Deino, A. L., Walter, R. C., Turrin, B. D., Swisher, C. C., Becker, T. A., … Jaouni, A. R. (1994). Intercalibrartion of astonomical and radioisotopic time. Geology, 22, 738–786. Reutter, K.-J., Scheuber, E., & Wigger, P. J. (1994). Tectonics of the Southern Central Andes. https://doi.org/10.1007/978-3-642-77353-2 Reyes-Wagner, V., Díaz, D., Cordell, D., & Unsworth, M. (2017). Regional electrical structure of the Andean subduction zone in central Chile ( 35 ° – 36 ° S ) using magnetotellurics. Earth, Planets and Space, 69, 142–151. https://doi.org/10.1186/s40623-017-0726-z Risacher, F., Fritz, B., & Hauser, A. (2011). Origin of components in Chilean thermal waters. Journal of South American Earth Sciences, 31(1), 153–170. https://doi.org/10.1016/j.jsames.2010.07.002 Rivera, O., & Cembrano, J. M. (2000). Modelo de formación de cuencas volcano-tectónicas en zonas de transferencia oblicuas a la cadena Andina: el caso de las cuencas Oligo-Miocenas de Chile central y su relación con estructuras WNW-NW (33-34˚30’). In IX Congreso Geológico Chileno (pp. 649–654). Roman, D. C. (2005). Numerical models of volcanotectonic earthquake triggering on non-ideally oriented faults. Geophysical Research Letters, 32(2), 1–4. https://doi.org/10.1029/2004GL021549 Roman, D. C., & Cashman, K. V. (2006). The origin of volcano-tectonic earthquake swarms. Geology, 34(6), 457–460. https://doi.org/10.1130/G22269.1 Roquer, T., Arancibia, G., Rowland, J., Iturrieta, P., Morata, D., & Cembrano, J. (2017). Fault- controlled development of shallow hydrothermal systems: Structural and mineralogical insights from the Southern Andes. Geothermics, 66, 156–173. https://doi.org/10.1016/j.geothermics.2016.12.003 180 Rosenau, M., Melnick, D., & Echtler, H. (2006). Kinematic constraints on intra-arc shear and strain partitioning in the southern Andes between 38°S and 42°S latitude. Tectonics, 25(4), 1–16. https://doi.org/10.1029/2005TC001943 Rosendahl, B. R. (1987). Architecture of continental rifts with special reference to East Africa. Annual Review of Earth and Planetary Sciences, 15, 445–503. Rowland, J. V., & Sibson, R. H. (2004). Structural controls on hydrothermal flow in a segmented rift system, Taupo Volcanic Zone, New Zealand. Geofluids, 4(4), 259–283. https://doi.org/10.1111/j.1468-8123.2004.00091.x Rowland, J. V, Simmons, S. F., Zealand, N., & Zealand, N. (2012). Hydrologic, Magmatic, and Tectonic Controls on Hydrothermal Flow, Taupo Volcanic Zone, New Zealand: Implications for the Formation of Epithermal Vein Deposits. Economic Geology, 107(3), 427–457. https://doi.org/10.2113/econgeo.107.3.427 Sagripanti, L., Rojas Vera, E. a., Gianni, G. M., Folguera, A., Harvey, J. E., Farías, M., & Ramos, V. a. (2015). Neotectonic reactivation of the western section of the Malargüe fold and thrust belt (Tromen volcanic plateau, Southern Central Andes). Geomorphology, 232, 164–181. https://doi.org/10.1016/j.geomorph.2014.12.022 Salazar, P., Kummerow, J., Wigger, P., Shapiro, S., & Asch, G. (2017). State of stress and crustal fluid migration related to west-dipping structures in the slab-forearc system in the northern Chilean subduction zone. Geophysical Journal International, 208(3), 1403–1413. https://doi.org/10.1093/gji/ggw463 Sanchez-Alfaro, P., Sielfeld, G., Campen, B. Van, Dobson, P., Fuentes, V., Reed, A., … Morata, D. (2015). Geothermal barriers, policies and economics in Chile – Lessons for the Andes. Renewable and Sustainable Energy Reviews, 51(August), 1390–1401. https://doi.org/10.1016/j.rser.2015.07.001 Sánchez, L., & Drewes, H. (2016). Crustal deformation and surface kinematics after the 2010 earthquakes in Latin America. Journal of Geodynamics. https://doi.org/10.1016/j.jog.2016.06.005 Sánchez, P., Pérez-Flores, P., Reich, M., Arancibia, G., & Cembrano, J. (2013). Crustal deformation effects on the chemical evolution of geothermal systems : the intra-arc Liquiñe–Ofqui Fault System , Southern Andes. International Geology Review, 55(11), 37–41. https://doi.org/10.1080/00206814.2013.775731 Savage, H. M., & Brodsky, E. E. (2011). Collateral damage: Evolution with displacement of fracture distribution and secondary fault strands in fault damage zones. Journal of Geophysical Research: Solid Earth, 116(3). https://doi.org/10.1029/2010JB007665 Scheuber, E., & Gonzalez, G. (1999). Tectonics of the Jurassic-Early Cretaceous magmatic arc of the north Chilean Coastal Cordillera (22°-26°S): A story of crustal deformation along a convergent 181 plate boundary. Tectonics, 18(5), 895–910. https://doi.org/10.1029/1999TC900024 Scholz, C. H. (1987). Wear and gouge formation in brittle faulting. Geology1, 15, 493–495. Scholz, C. H. (1988). The brittle-plastic transition and the depth of seismic faulting. Geologische Rundschau, 77(1), 319–328. https://doi.org/10.1007/BF01848693 Scholz, C. H. (2002). The Mechanics of Earthquakes and Faulting (2nd Editio). Cambridge University Press. Sernageomin. (2003). Mapa geologico de chile: version digital. Publicacion Geologia Digital, 4, 25. Shaw, H. R. (1980). Fracture mechanism of magma transport from the mantle to the surface. In Hardgraves, R.B. (Ed.), Physics of Magmatic Processes (pp. 201–264). Princeton Uni. Press. Shearer, P. M. (2009). Introduction to Seismology (2nd Ed.). Cambridge University Press. Sibson, R. (1984). Roughness at the Base of the Seismogenic Zone: Contributing Factors. Journal of Geophysical Research, 89, 5791–5799. Sibson, R. (2002). Geology of the Crustal Earthquake Source. In Lee, W.H.K.; Kanamori, H.; Jennings, P.; Kisslinger, C. (eds.), International Handbook of Earthquake and Engineering Seismology, Part A (pp. 455–473). Sibson, R. H. (1989). Earthquake faulting as a structural process. Journal of Structural Geology, 11(1–2), 1–14. https://doi.org/10.1016/0191-8141(89)90032-1 Sibson, R. H. (1996). Structural permeability of fluid-driven fault-fracture meshes. Journal of Structural Geology, 18(8), 1031–1042. https://doi.org/10.1016/0191-8141(96)00032-6 Sibson, R. H., Moore, J. M. M., & Rankin, a. H. (1975). Seismic pumping--a hydrothermal fluid transport mechanism. Journal of the Geological Society, 131(6), 653–659. https://doi.org/10.1144/gsjgs.131.6.0653 Siebert, L., Simkin, T., & Kimberly, P. (2010). VOLCANOES of the World - 3rd ed. Sieh, K., & Natawidjaja, D. (2000). Neotectonics of the Sumatran fault, Indonesia. Journal of Geophysical Research: Solid Earth, 105(B12), 28295–28326. https://doi.org/10.1029/2000JB900120 Sielfeld, G., Lange, D., & Cembrano, J. (2019). Intra - arc Crustal Seismicity : Seismo - tectonic Implications for the Southern Andes Volcanic Zone , Chile. Tectonics. https://doi.org/10.1029/2018TC004985 Sielfeld, G., Cembrano, J., & Lara, L. (2016). Transtension driving volcano-edifice anatomy: Insights from Andean transverse-to-the-orogen tectonic domains. Quaternary International, 1–17. 182 https://doi.org/10.1016/j.quaint.2016.01.002 Singer, B. S., Thompson, R. a., Dungan, M. a., Feeley, T. C., Nelson, S. T., Pickens, J. C., Metzger, J. (1997). Volcanism and erosion during the past 930 k.y. at the Tatara-San Pedro complex, Chilean Andes. Bulletin of the Geological Society of America, 109(2), 127–142. https://doi.org/10.1130/0016- 7606(1997)109<0127:VAEDTP>2.3.CO;2 Singer, B. S., Anderson, N., Le Mevel, H., Feigl, K. L., Demets, C., Tikoff, B., Vazquez, J. (2014). Dynamics of a large , restless , rhyolitic magma system at. GSA Today, 24(12), 4–10. https://doi.org/10.1130/GSATG216A.1. Snoke, J. A., Munsey, J. W., Teague, A. G., & Bollinger, G. A. (1984). A program for focal mechanism determination by combined use of polarity and SV-P amplitude ratio data. Earthquake Notes, 55. Soffia, J. M., & Clavero, J. (2010). Doing geothermal exploration business in Chile. Energía Andina experience. In Geothermal Resources Council Transaccions (p. 34:637-641). Somoza, R. (1998). Updated Nazca (Farallon)-South America relative motions during the last 40 My: Implications for mountain building in the central Andean region. Journal of South American Earth Sciences, 11(3), 211–215. https://doi.org/10.1016/S0895-9811(98)00012-1 Springer, M. (1999). Interpretation of heat-flow density in the Central Andes. Tectonophysics, 306(3–4), 377–395. https://doi.org/10.1016/S0040-1951(99)00067-0 Stanton-Yonge, A., Griffith, W. A., Cembrano, J., St. Julien, R., & Iturrieta, P. (2016). Tectonic role of margin-parallel and margin-transverse faults during oblique subduction in the Southern Volcanic Zone of the Andes: Insights from Boundary Element Modeling. Tectonics, 35(9), 1990–2013. https://doi.org/10.1002/2016TC004226 Stern, C. R. (2004). Active Andean volcanism: its geologic and tectonic setting. Revista Geológica de Chile, 31(2), 161–206. Stern, C., Moreno, H., Lopez-Escobar, L., Clavero, J., Naranjo, J. A., Parada, M. Á., & Skewes, M. A. (2007). Chilean volcanoes. In Geology of Chile (pp. 147–178). Stesky, R. M. (1975). The mechanical behavior of faulted rocks at high temperature and pressure, Ph. D. thesis at Massachusetts Institute of technologu. Suzuki, Y., Ioka, S., & Muraoka, H. (2014). Determining the maximum depth of hydrothermal circulation using geothermal mapping and seismicity to delineate the depth to brittle-plastic transition in northern Honshu, Japan. Energies, 7(5), 3503–3511. https://doi.org/10.3390/en7053503 Tankard, A. ., Uliana, M. A., Welsink, H. J., Ramos, V. A., Tunik, M., Franca, A. B., … Santa Ana et al., H. (1995). Structural and tectonic controls of basin evolution in southwestern Gondwana during 183 the Phanerozoic. Tectonic Controls of Basin Evolution in Southwestern Gondwana, in A.J. Tankard, R. Suarez S., and H.J. Weslsink, Petroleum Basins of South America: AAPG Memoir 62, 5–52. Tarantola, A., & Valette, B. (1982). Generalized Nonlinear Inverse Problems Solved Using the Least Square Criterion. Reviews of Geophysics, 20(2), 219–232. Tardani, D., Reich, M., Roulleau, E., Takahata, N., Sano, Y., Pérez-Flores, P., … Arancibia, G. (2016). Exploring the structural controls on helium, nitrogen and carbon isotope signatures in hydrothermal fluids along an intra-arc fault system. Geochimica et Cosmochimica Acta, 184(July), 193–211. https://doi.org/10.1016/j.gca.2016.04.031 Tassara, A., & Echaurren, A. (2012). Anatomy of the Andean subduction zone: three-dimensional density model upgraded and compared against global-scale models. Geophys. J. Int, 189, 161–168. https://doi.org/10.1111/j.1365-246X.2012.05397.x Thomson, S. N. (2002). Late Cenozoic geomorphic and tectonic evolution of the Patagonian Andes between latitudes 42??S and 46??S: An appraisal based on fission-track results from the transpressional intra-arc Liqui??e-Ofqui fault zone. Bulletin of the Geological Society of America, 114(9), 1159–1173. https://doi.org/10.1130/0016-7606(2002)114<1159:LCGATE>2.0.CO;2 Tibaldi, a. (2001). Multiple sector collapses at Stromboli volcano, Italy: How they work. Bulletin of Volcanology, 63(2–3), 112–125. https://doi.org/10.1007/s004450100129 Tibaldi, a., & Lagmay, a. M. F. (2006). Interaction between volcanoes and their basement. Journal of Volcanology and Geothermal Research, 158(1–2), 1–5. https://doi.org/10.1016/j.jvolgeores.2006.04.011 Tibaldi, A. (1995). Morphology of pyroclastic cones and tectonics. Journal of Geophysical Research, 100(B12), 24521–24535. https://doi.org/10.1029/95JB02250 Tikoff, B., & De Saint Blanquat, M. (1997). Transpressional shearing and strike-slip partitioning in the Late Cretaceous Sierra Nevada magmatic arc, California. Tectonics, 16(3), 442–459. Retrieved from file:///C:/Users/User/Dropbox/Publicaciones/pdf L-Z/Tikoff and Saint Blanquat.pdf Tikoff, B., & Teyssier, C. (1994). Strain modeling of displacement-field partitioning in transpressional orogens. Journal of Structural Geology, 16(11), 1575–1588. https://doi.org/10.1016/0191-8141(94)90034-5 Turcotte, D. L. (1997). Fractals and chaos in geology and geophysics (2nd Ed.). Cambridge University Press. Utsu, T. (2002). Statistical Features of Seismicity. In Lee, W., H. Kanamori, P.C. Jennings and C. Kisslinger (eds.), International Handbook of Earthquake & Engineering Seismology, Part A (pp. 719– 732). 184 Veloso, E., Gomila, R., González, R., Cembrano, J., Jensen, E., & Arancibia, G. (2015). Stress fields recorded on large-scale strike-slip fault systems: Effects on the tectonic evolution of crustal slivers during oblique subduction. Tectonophysics, 664, 244–255. https://doi.org/10.1016/j.tecto.2015.09.022 Völker, D., Kutterolf, S., & Wehrmann, H. (2011). Comparative mass balance of volcanic edifices at the southern volcanic zone of the Andes between 33°S and 46°S. Journal of Volcanology and Geothermal Research, 205(3–4), 114–129. https://doi.org/10.1016/j.jvolgeores.2011.03.011 Wallace, R. E. (1951). Geometry of Shearing Stress and Relation to Faulting. The Journal of Geology. Wang, K., Hu, Y., Bevis, M., Kendrick, E., Robert, S., Vargas, R. B., Lauría, E. (2007). Crustal motion in the zone of the 1960 Chile earthquake: Detangling earthquake-cycle deformation and forearc- silver translation. Geochemistry, Geophysics, Geosystems, 8(10), 1–14. https://doi.org/10.1029/2007GC001721 Weijermars, R. (1991). The role of stress in ductile deformation. Journal of Structural Geology, 13(9), 1061–1078. https://doi.org/doi:10.1016/0191-8141(91)90057-P Weller, O., Lange, D., Tilmann, F., Natawidjaja, D., Rietbrock, A., Collings, R., & Gregory, L. (2012). The structure of the Sumatran Fault revealed by local seismicity, 39(November 2011), 1–7. https://doi.org/10.1029/2011GL050440 White, Randall A, Harlow, D. H. (1993). America Since 1900. Bulletin of the Seismological Society of Americ, 83(4), 1115–1142. White, R. (1991). Tectonic implications of upper-crustal seismicity in Central America. In D. B. Slemmon, E. R. Engdahl, M. D. Zoback, & D. Blackwell (Eds.), Neotectonics of North America: Slemmon, D.B., Engdahl, E.R., Zoback, M.D., and Blackwell, D. (Editors) (pp. 323–338). Geological Society of America. Wibberley, C. a. J., Yielding, G., Toro, G. D. I., Umr, C., Antipolis, D. N. S., Einstein, A., … Di Toro, G. (2008). Recent advances in the understanding of fault zone internal structure: a review. Geological Society, London, Special Publications, 299(1), 5–33. https://doi.org/10.1144/SP299.2 Wigger, P., & et al. (1994). Variation of the crustal structure of the Southern Central Andes deduced from seismic refraction investigations. In Reutter, K.-J., E. Scheuber, and P. Wigger (eds.), Tectonics of the Sourthern Central Andes (pp. 23–48). Springer Verlag. https://doi.org/10.1007/978- 3-642-77353-2_2 Wittlinger, G., Herquel, G., & Nakache, T. (1993). Earthquake location in strongly heterogeneous media. Geophysical Journal International, 115, 759–777. Wrage, J., Tardani, D., Reich, M., Daniele, L., Arancibia, G., Cembrano, J., … Pérez-Moreno, R. (2017). Geochemistry of thermal waters in the Southern Volcanic Zone, Chile – Implications for structural 185 controls on geothermal fluid composition. Chemical Geology, 466(July), 545–561. https://doi.org/10.1016/j.chemgeo.2017.07.004 Van Wyk De Vries, B., & Merle, O. (1998). Extension induced by volcanic loading in regional strike- slip zones. Geology, 26(11), 983–986. https://doi.org/10.1130/0091- 7613(1998)026<0983:EIBVLI>2.3.CO;2 Yamaguchi, A., Cox, S. F., & Kimura, G. (2011). Dynamic changes in fl uid redox state associated with episodic fault rupture along a megasplay fault in a subduction zone. Earth and Planetary Science Letters, 302(3–4), 369–377. https://doi.org/10.1016/j.epsl.2010.12.029 Yamaji, A. (2000). The multiple inverse method: a new technique to separate stresses from heterogeneous fault-slip data. Journal of Structural Geology, 22(4), 441–452. https://doi.org/10.1016/S0191-8141(99)00163-7 Yamaji, A. (2016). Genetic algorithm for fitting a mixed Bingham distribution to 3D orientations: A tool for the statistical and paleostress analyses of fracture orientations. Island Arc, 25(1), 72–83. https://doi.org/10.1111/iar.12135 Yamaji, A., Sato, K., & Otsubo, M. (2011). Multiple Inverse Method Software Package. Kyoto University, Japan. Retrieved from file:///C:/Users/User/AppData/Local/Mendeley Ltd./Mendeley Desktop/Downloaded/Yamaji, Sato - 2011 - Multiple Inverse Method Software Package-User’s Guide, Version 6.2.pdf Yamaji, A., & Sato, K. (2011). Clustering of fracture orientations using a mixed Bingham distribution and its application to paleostress analysis from dike or vein orientations. Journal of Structural Geology, 33(7), 1148–1157. https://doi.org/10.1016/j.jsg.2011.05.006 Yamaji, A., Otsubo, M., & Sato, K. (2006). Paleostress analysis using the Hough transform for separating stresses from heterogeneous fault-slip data. Journal of Structural Geology, 28(6), 980– 990. https://doi.org/10.1016/j.jsg.2006.03.016 Yañez, G. a, Gana, P., & Fernandez, R. (1998). Origin and geological significance of the Melipilla Anomaly, Central Chile. Revista Geologica De Chile. Yáñez, G., & Cembrano, J. (2004). Role of viscous plate coupling in the late Tertiary Andean tectonics. Journal of Geophysical Research: Solid Earth, 109(B2), B02407–B02407. https://doi.org/10.1029/2003JB002494 Yoon, M., Buske, S., Lüth, S., Schulze, A., Shapiro, S. A., Stiller, M., & Wigger, P. (2003). Along-strike variations of crustal reflectivity related to the Andean subduction process. Geophysical Research Letters, 30(4), 1–4. https://doi.org/10.1029/2002GL015848 Yoon, M., Buske, S., Shapiro, S. A., & Wigger, P. (2009). Reflection Image Spectroscopy across the Andean subduction zone. Tectonophysics, 472(1–4), 51–61. https://doi.org/10.1016/j.tecto.2008.03.014 186 Yoshida, S., Koketsu, K., Shibasaki, T., Kato, T., & Yoshida, Y. (1996). Joint Inversion of Near- and Far-field Waveforms and Geodetic Data for the Rupture Process of the 1995 Kobe Earthquake. Journal of Physics of the Earth, 44, 437–454. https://doi.org/10.4294/jpe1952.44.437 Yuan, X., Sobolev, S. V., Kind, R., Oncken, O., & Group, A. S. (2000). New constraints on subduction and collision processes in the central Andes from P-to-S converted seismic phases. Nature, 408, 958–961. Zellmer, G. F., Annen, C., Charlier, B. L. a, George, R. M. M., Turner, S. P., & Hawkesworth, C. J. (2005). Magma evolution and ascent at volcanic arcs: Constraining petrogenetic processes through rates and chronologies. Journal of Volcanology and Geothermal Research, 140(1–3), 171– 191. https://doi.org/10.1016/j.jvolgeores.2004.07.020 Zobin, V. M. (2003). Introduction to volcanic seismology. Elsevier Science B.V. 187 7 Supplementary Material 7.1 Supplementary material chapter 2 Introduction Here we present complementary information for the interested reader. A brief description is here given in the order as the Supplementary Information is organized and/or mentioned in the main text (chapter 2). First, we include a map with main geologic units, strato volcanoes, and the temporary network of this project (Fig. 6-1). Additionally, the map includes OVDAS permanent stations used in this work. Section 6.1.2 is a summary of station installation set up, which includes Figure 6-2 and 6-3. Section 6.1.3 explains the NonLinLoc algorithm and location procedure. Table 6-1 lists the geographical position of the temporary SAIAS-stations and permanent OVDAS-stations. Table 6-2 contains the origin time, strike, dip and rake of calculated focal mechanism in the Aki-Richards convention. Figure 6-4 graphics the temporal distribution of seismicity (number of events) and moment magnitudes constrained to the Villarrica volcano, two months before and two months after its strombolian eruption on the 3rd of March, 2015. Finally, in section 6.1.4 we present a table caption for the complete SAIAS catalog (link to dataset [ds01] is indicated), which contains the 356 events reported and discussed in the main text. 188 Figure 7-1 Geological map extracted from the 1:1.000.000 map from Sernageomin (2002). Seismic network installed in by this study in inverted blue and red triangles. Light green inverted triangles are OVDAS permanent stations included (z-component only) this study. 189 Station installation and set-up. The seismic catalog (ds01) was built upon the continuous record of 34 stations running from march 2014 to June 2015. Each station consists of a seismometer (3 component sensor), an EarthDataLogger (EDL) set at a 100 Hz sampling rate, an energy supply (photovoltaic solar panel) and a 12V deep cycle battery with 65Ah. The installation set up is illustrated with field pictures and a schematic section on Fig. 6-2. We installed 8 Trillium Compact force balance broadband seismometers, made by Nanometrics (Fig. 6-3a). The remaining 27 stations, were equipped with Mark L4-3D 3- component geophones (instrument response 1Hz - 100 Hz), built by MARK PRODUCTS (Fig. 6-3b). Data was logged with digital recording system PR6-24, made by EARTH DATA (Fig. 6- 3c). Detailed information of the sensors and EDL can be found in the following e-link: https://www.gfz-potsdam.de/en/section/geophysical-deep-sounding/infrastructure/geophysical-instrument-pool-potsdam- gipp/pool-components/seismic-pool/ Visual inspection picking and data processing was done between September 2015 and August 2016. In turn, focal mechanism calculations were carried out in June 2017. 190 Figure 7-2 a-d) Field pictures showing the EarthDataLogger (a) and installation and servicing of stations running with solar panels. e) Scheme of main components of a station. Sensor was commonly set and buried at ~1 m ±0.2 depth. 191 Figure 7-3 Instruments used in the seismic deployment a) Trillium compact broadband. b) Mark L4-3D 3-components short-period sensor. c) EarthDataLogger. Non-linear location technic and procedure (NonLinLoc; Lomax et al., 2000) In contrast to iterative-linearized location programs that produce a single-point solution (i.e. HYPOCENTER; Lienert, et al., 1986; Lienert and Havskov, 1995), we applied a nonlinear localization open-code software package named NonLinLoc (Lomax et al.,2000). The program consists of a probabilistic earthquake location methodology that provides reasonable uncertainty and resolution information. The program searches for the best- probabilistic hypocentre coordinates using a probability density function over the trial spatial coordinates (x,y,z) and origin time (t) unknown parameters (Lomax et al., 2000). The location is then represented as a hyper-volume that would include multiple “optimal” solutions and may have an irregular ellipsoidal shape (Lomax et al., 2000). The non–linear earthquake location algorithms described and developed by Lomax et al. (2000) are based on the probabilistic formulation of inversion presented in Tarantola and Valette (1982); Tarantola (1987), and the equivalent methodology for earthquake location (i.e., Tarantola and Valette,1982; Moser et al.,1992; Wittlinger et al.,1993). The inversion process tries to constrain the values of unknown parameters p, by relating a given vector of observed data d and a theoretical relationship q(d,p). According to Tarantola and 192 Valette (1982) and to Tarantola (1987), if the density functions for the model parameters rp(p) and for the observed data rd(d) are independent, and the theoretical relationship is supposed to be a conditional density function q(d|p)µp(p), a complete probabilistic solution can be expressed as a posterior probability density function (PDF) sp(p) 𝜎𝑝(𝑝) = 𝜌𝑝(𝑝)∫ H𝑑(J)K(J|:) 𝑑𝑝 (1) M𝑑(J) where µp(p) and µd(d) are null information density functions specifying the state of total ignorance. For the propose of seismic events determination, unknown parameters (p) are the hypocentre coordinates x=(x,y,z) and the origin time (T). The observed data are the arrival times t, obtained by visual inspection and picking of the waveforms. So, the theoretical relation gives predicted travel times h. Tarantola and Valette (1982) demonstrated that, if the theoretical relationship and observed arrival times are assumed to have Gaussian uncertainties with covariance matrices CT and Ct, respectively, and if the prior information on T is considered as uniform, then it is possible to integrate the PDF over T to obtain the marginal PDF for the spatial location parameters (x). This marginal PDF reduces to (Tarantola & Valette, 1982; Moser et al., 1992) 𝜎(𝑥) = 𝐾 𝜌(𝑥) ∙ exp [− T 𝑔(𝑥)] (2) U 𝑔(𝑥) = V?̂?0 − ℎ[(𝑥) ] \ [𝐶𝑡 + 𝐶𝑇]1T[?̂?0 − ℎ[(𝑥)] (3) 193 where 𝐾 is a normalization factor, 𝜌(𝑥) is a density function of prior information on the model parameters, and 𝑔(𝑥) is the likelihood function. The variable ?̂?0 is the vector of observed arrival times t minus their weighted mean, and ℎ[ is the vector of theoretical travel times h minus their weighted mean, where the weights wi are given by 𝑤𝑖 = ∑d 𝑤𝑖𝑗 ; 𝑤𝑖𝑗 = [(𝐶𝑡 + 𝐶𝑇)1T]𝑖𝑗 . (3) Furthermore, the maximum likelihood origin time for a hypocentre at x, y, z is given by (Moser et al., 1992) ( ) ∑ ∑𝑇 𝒙 = l ki𝑖𝑗[;𝑖1j𝑖(𝒙)]𝑚𝑙 ∑ . (4) l ∑ki𝑖𝑗 So, the posterior density function (PDF) of Equation (2) constitutes a complete, probabilistic solution to the location problem, including information on uncertainty and resolution (Lomax et al., 2000). The latter includes location uncertainties related to (a) spatial relation between the network and the event, (b) measurement uncertainty in the observed arrival times, and (c) errors in the calculation of theoretical travel times. 194 The PDF can be computed using any of three different algorithms (Lomax et al., 2000), specifically the grid-search, Metropolis-Gibbs sampling and Oct-Tree methods. In this study we used an Oct-Tree sampling algorithm. This method gives accurate, efficient and complete mapping of earthquake location PDFs in the 3d space (x,y,z), resulting in a cloud of potential locations for individual events. The Oct-Tree method uses recursive subdivision and sampling of cells in 3d space to generate a cascade of sampled cells (i.e. from 1 original sample it gets 8 new samples by subdividing the original cell into 8 new ones), where the density of sampled cells follows the PDF values of the cell centre. For each event, we estimated a 68% spatial confidence for the final PDF sample from 3D error ellipsoids fitted to the 3D PDF scatter samples. The NonLinLoc algorithm estimates several errors, including the “picking” errors, an error for the velocity model and travel-time error as function of travel time for each ray. These, are set in the configuration file. Therefore, NonLinLoc algorithm propagates all parameter uncertainties into the thousands/hundreds of locations estimated for each event and then assumed a Gaussian distribution for calculating the error ellipsoid in 3D, which then is projected onto lat, lon, z direction during the conversion to Nordic files. We used the centre of the ellipsoid for event visualization, including vertical and horizontal major axes of the error ellipsoid. The only parameter that NonLinLoc does not take into account are slightly wrong GPS times. If the error is significant NonLinLoc will not consider the arrival since it cannot associate the pick to the event and this should clearly come out as an anomaly in the 195 station corrections. Therefore, GPS timing have been carefully checked during and after the network deployment. Detailed user tutorial for running the program can be found in http://alomax.free.fr/nlloc/. 196 Table 7-1 Geographical coordinates of stations used for the SAIAS catalogue construction. Temporary stations installed for this project are enlisted in the left panel. Last letter of ID indicates short period ‘S’ or broad band ‘B’ sensor-type. In the right panel are the OVDAS stations, which are located in main stratovolcanoes (only vertical component). SAIAS OVDAS Station Latitude Longitude Elevation Station Latitude Longitude Elevation ID (˚) (˚) (m) ID (˚) (˚) (m) LX6S -38.474 -71.775 635 COPZ -37.837 -71.160 2431 LX2S -38.234 -71.728 1167 PAIZ -38.873 -71.643 1075 LX4S -38.463 -71.638 930 MOTZ -38.675 -71.748 1952 LB3S -38.212 -71.174 1123 CONZ -38.740 -71.770 1763 LB1S -38.156 -71.320 764 LAVZ -38.702 -71.646 1077 LB5S -38.694 -71.330 1172 CRUZ -38.709 -71.800 1695 LB2B -38.167 -71.422 771 ROCZ -38.678 -71.683 1700 LB4S -38.541 -71.257 1118 AGUZ -38.723 -71.715 2150 LX5S -38.470 -71.509 1131 MRNZ -39.409 -71.936 2301 LX3S -38.434 -71.389 1052 TRAZ -39.451 -71.885 1893 LL3S -38.833 -71.638 522 VN2Z -39.399 -71.963 1474 LL4S -38.873 -71.822 414 KIKZ -39.419 -71.864 1707 LV3S -39.343 -71.971 750 PUCZ -39.275 -71.984 239 LV3R -39.343 -71.971 750 VOIZ -39.437 -71.962 1794 LS4S -39.197 -71.854 655 VTDZ -39.435 -71.961 1838 LS5S -39.215 -71.613 693 CVVZ -39.309 -72.161 308 LS5R -39.215 -71.613 693 LV4B -39.420 -71.784 734 LS3S -39.117 -71.475 828 LZ1B -39.631 -71.928 234 LV6B -39.477 -71.550 456 LV2S -39.333 -71.838 435 LS6S -39.363 -71.575 422 LV7S -39.489 -72.089 336 LV1S -39.301 -72.113 242 LZ3S -39.775 -71.754 579 LZ2S -39.665 -71.845 423 LV8B -39.507 -71.855 895 LS2S -39.065 -71.720 495 LL1B -38.590 -71.801 743 LB6S -38.809 -71.282 1141 LL2B -38.672 -71.624 1054 LX1S -38.076 -71.701 739 LV5B -39.437 -71.792 781 LS1B -38.873 -71.493 1196 LV9S -39.525 -71.926 715 197 Table 7-2 Focal mechanism in the Aki and Richards convention. Number of observations (Obs.), GAP and moment magnitudes are indicated. Date Origin time Latitude Longitude Depth Strike Dip Rake Obs. GAP Mw 22.04.14 02:52:43.80 -38.374 -71.621 9 310.95 47.99 23.22 15 138 2 08.03.15 18:34:11.80 -38.415 -71.597 9 87 78 -24 15 160 2.2 06.04.14 11:02:36.70 -38.427 -71.611 9 55.57 48.44 -30.79 12 163 2 06.09.14 08:35:49.50 -38.547 -71.659 9 68.81 64.34 -16.1 25 166 2.4 09.12.14 07:28:14.90 -38.585 -71.624 9 265.93 40.26 -82.25 27 161 2.1 08.12.14 22:03:46.30 -38.596 -71.612 9 75.6 62.12 -28.45 36 176 2.6 31.07.14 09:51:12.20 -38.653 -71.607 9.4 256.76 35.53 -30.64 34 143 3.4 15.04.15 08:54:30.00 -38.823 -71.731 12.3 283.97 38.29 36.2 28 111 2.3 11.07.14 17:36:03.80 -38.825 -71.719 15 95.1 75.52 26.57 7 132 1.5 14.04.15 12:40:52.00 -38.842 -71.713 13.8 174.99 65.41 -4.63 21 99 1.7 10.09.14 06:39:41.40 -38.844 -71.736 11.8 125 90 10 8 101 2 04.04.15 09:00:11.70 -38.94 -71.781 9 96 78 -2 11 142 1.8 14.11.14 01:13:24.50 -39.126 -71.707 9 307.02 67.48 -20.36 17 102 2.3 13.11.14 06:55:37.40 -39.127 -71.71 9 310.43 82.36 6.47 11 102 2 14.11.14 00:53:27.70 -39.128 -71.699 9 177.69 60 90 31 97 3 13.11.14 07:59:23.00 -39.128 -71.709 8.8 260.95 43.96 22.18 17 102 2.4 13.11.14 07:00:30.60 -39.128 -71.708 9 105.89 41.41 -40.89 14 102 1.7 14.11.14 04:15:47.60 -39.129 -71.705 8.6 48 69 -180 15 100 2.2 14.11.14 04:21:59.60 -39.132 -71.707 8.9 280.68 35.53 53.95 15 100 2 19.07.14 11:16:29.50 -39.136 -71.754 9.9 328.32 61.98 -11.17 12 116 1.6 17.04.15 20:05:53.20 -39.154 -71.645 9 74.91 73.73 -11.79 15 164 2.2 19.04.15 15:16:04.80 -39.419 -71.886 9 164.21 83.59 -39.57 18 130 2.1 22.04.15 11:27:40.20 -39.423 -71.896 9 80.57 72.61 42.19 12 77 2.2 25.04.15 04:03:26.20 -39.423 -71.892 8.9 331.32 85.02 29.62 38 75 2.2 30.04.15 18:05:03.30 -39.424 -71.889 7.6 109.3 85.17 -74.95 21 85 2.2 27.03.14 17:02:22.30 -39.424 -71.892 7.5 317.61 50 0 11 162 2.3 10.05.15 18:30:47.10 -39.426 -71.886 9 319.93 88.29 -4.7 30 61 2.5 23.04.15 22:22:23.50 -39.426 -71.888 9 324.79 71.11 16.69 23 73 2.5 23.04.15 00:27:59.70 -39.426 -71.892 8.8 141.45 67.48 -45.9 29 135 2.4 23.04.15 22:20:24.50 -39.427 -71.889 8.9 22.99 53.32 -10.7 15 134 2.4 10.02.15 02:16:11.70 -39.427 -71.883 8.8 140.11 86.79 3.83 11 131 2.3 11.04.15 22:43:25.90 -39.429 -71.884 9 321.7 63.05 -13.71 15 132 2.4 19.04.15 22:50:14.80 -39.43 -71.89 9 320.41 48.36 -18.88 15 77 2.2 16.03.15 08:22:35.20 -39.43 -71.892 9 192.52 84.18 75.93 29 75 2.3 19.02.15 15:12:17.20 -39.443 -71.845 9 115.67 62.82 -62.79 23 129 2.4 23.06.14 00:39:40.10 -39.51 -71.903 4 48.33 60 0 13 90 1.7 08.04.14 00:06:16.20 -39.697 -71.84 8.6 14.6 71.25 -68.83 12 174 2.1 198 Figure 7-4 Number of events per day (grey bars) and daily cumulative moment magnitude (blue bars with a dot in the tip) observed at the Villarrica Volcano edifice, from the 1st of January to the 13th of June 2015. The dashed green line indicates the Strombolian eruption on the 3rd of march 2015. Data Set S1. Southern Andes Intra-Arc Seismicity (SAIAS) catalog. Each row contains information of a single seismic event, chronologically organized. Date, origin time, geographical coordinates (lon. lat. in degrees), depth (km), uncertainties in km. for x, y and z; number of observations (N. Obs), travel times residuals for final earthquake location (RMS), GAP (greatest azimuthal angle without observation), local magnitude (Ml) and moment magnitude (Mw) are enlisted. Data set file is named ds01 and is uploaded to both, dropbox (click link below) and to AGU’s journal submission site. 199 https://www.dropbox.com/sh/58knmvggbjab3vb/AABxKIpNQ_feAsK0UE5zaxDka?dl=0 Video 6-1. Time-lapse of the whole catalog made with ParaView open-source application and program in python. *(find video in the following link) https://www.dropbox.com/s/84mjo470pf73gpw/seism_anim_osx.mp4?dl=0 https://www.dropbox.com/s/si9ge5gridnow20/seism_anim.mp4?dl=0 Video 6-2. Time-lapse of the Caburgua sequence made with ParaView open-source application and program in python. *(find video in the following link) https://www.dropbox.com/s/5inh2t7f7byvlgg/caburgua_anim2_osx.mp4?dl=0 https://www.dropbox.com/s/x1wvkcnp0ts0k0r/caburgua_anim2.mp4?dl=0 200 7.2 Supplementary material chapter 4 Table 7-3 List and location of structural stations where FSD was collected ST_TFZ UTM WGS84 Station_ID Station_N x y Elevation LaMina 1 337789 6034046 982 McClure 2 339512 6034501 992 LaIsla 3 342733 6033543 1037 Sepultura 4 342733 6033543 1037 La Plata 5 348967 6028241 1371 Los Cóndores 6 350102 6027722 1273 El Indio 7 347559 6024268 1661 MUR_N 8 346922 6021255 1767 MUR_S 9 346909 6020689 1885 Q_AGUDA 10 344483 6019929 1919 VALSG_2 11 341737 6020243 2273 VALSG_1 12 341192 6020517 2101 LUNA 13 340542 6020186 2249 PEJE_RIDGE 14 341134 6022187 2181 Qcas_N 15 325942 6019567 2005 LaPiscina 16 326039 6019494 2035 Pesadilla 17 326987 6018571 1928 QCas1_HW 18 327154 6018333 2001 QCas2_HW 19 327024 6018215 2027 QCAS_S1S2 20 327056 6017985 2195 Cipres_1 21 329005 6012593 2499 Cipres_2 22 328300 6011423 2009 SanPedro_S1 23 330447 6011746 2073 SanPedro_S2 24 330438 6011701 2087 PelladoFum 25 337612 6015439 2373 Pell_Up_1 26 337834 6014673 2096 Pell_Up2 27 337197 6014579 2018 LitGrab_Pell 28 337634 6014342 2014 Huemul 29 337857 6012732 1805 Lo Aguirre 30 358962 6017935 2020 Saso 31 349474 6006169 1983 201 Data base The structural Data Base is available on the following dropbox link. Data is organized under different directories, where FSData_STs/ has all fault slip data sets per ST in fdt format.; phi_STs contains the calculated stress tensor and phi value after running the MIM (phi-histogram are included for each station); and mi4_STs includes the .mi4 files with the state of stress solutions after the inversion process with k=5 for each ST. Dikes are separated enlisted in the Basement\Dikes.xlsx and Intravolcanic\Dikes.xlsx files. https://www.dropbox.com/sh/rxst1l1vmawdn33/AACNpn4UuY2bhZvM1-StR_jva?dl=0 202 203 204 Figure 7-5 Compilation of stress-state paired stereograms and fault paleo-stress tensor solutions for each single structural station. 205 Figure 7-6 Field pictures of folds.