Filtern
Erscheinungsjahr
- 2019 (24) (entfernen)
Dokumenttyp
- Dissertation (24) (entfernen)
Gehört zur Bibliographie
- ja (24)
Schlagworte
- Anden (2)
- Andes (2)
- numerische Modellierung (2)
- AMOC (1)
- Abschiebungshorizonte (1)
- Alterationsgeochemie (1)
- Ambient noise (1)
- Angewandte Geophysik (1)
- Applied Geophysics (1)
- Arctic tundra (1)
Institut
- Institut für Geowissenschaften (24) (entfernen)
The Atlantic Meridional Overturning Circulation (AMOC) is likely the most well-known system of ocean currents on Earth, redistributing heat, nutrients and carbon over a large part of the Earth’s surface and affecting global climate as a result. Due to enhanced freshwater fluxes into the subpolar North Atlantic as a response to global warming, the AMOC is expected, and may have already started, to weaken and these changes will likely have global impacts. It is therefore of considerable relevance to improve our understanding of past and future AMOC changes. My thesis tries to answer some of the open questions in this field by giving strong evidence that the AMOC has already weakened over the last century, by narrowing future projections of this slowdown and
by studying the impacts on global surface warming.
While there have been various studies trying to reconstruct the strength of the overturning circulation in the past, often based on model simulations in combination with observations (Jackson et al., 2016, Kanzow et al., 2010) or proxies (Frajka-Williams, 2015, Latif et al., 2006), the results so far, due to lack of direct measurements, have been inconclusive. In the first paper I build on previous work that links the anomalously low sea surface temperatures (SSTs) in the North Atlantic with the reduced meridional heat transport due to a weaker AMOC. Using the output of a high-resolution global climate model, I derive a characteristic spatial and seasonal SST fingerprint of an AMOC slowdown and an improved SST-based AMOC index. The same fingerprint is seen in
the observational SSTs since the late 19th Century, giving strong evidence that since then the AMOC has slowed down. In addition, the reconstruction of the historical overturning strength with the new AMOC index agrees well with and extends the results of earlier studies as well as the direct measurements from the RAPID project and shows a strong decline of the AMOC by about 15% (3±1 Sv) since the mid-20th Century (Caesar et al., 2018).
The reconstruction of the historical overturning strength with the AMOC index enables us to weight future AMOC projections based on their skill in modeling the historical AMOC as described in the second paper of this thesis (Olson et al., 2018). Using Bayesian model averaging we considerably narrow the projections of the CMIP5 ensemble to a decrease of -4.0 Sv and -6.8 Sv between the years 1960-1999 and 2060-2099 for the RCP4.5 and RCP8.5 emission scenarios, respectively. These values fit to, yet are at the lower end of, previously published estimates.
In the third paper I examine how the AMOC slowdown affects the global mean surface temperature (GMST) with a focus on how it will change the ocean heat uptake (OHC). Accounting for the effect of changes in the radiative forcing on the GMST, I test how AMOC variations correlate with the residual part of surface temperature changes in the past. I find that the correlation is positive which fits the understanding that the deep-water formation that is important in driving the AMOC cools the deep ocean and therefore warms the surface (Caesar et al., 2019). The future weakening of the overturning circulation could therefore delay global surface warming.
Due to nonlinear behavior and scale specific changes it can be difficult to study the dominant processes and modes that drive climate variability. In the fourth paper we develop and test a new technique based on the wavelet multiscale correlation (WMC) similarity measure to study climate variability on different temporal and spatial scales (Agarwal et al., 2018). In a fifth contribution to my thesis this method is applied to the observed sea surface temperatures. The results reconfirm well-known relations between SST anomalies such as the El Niño-Southern Oscillation (ENSO) and the Pacific Decadal Oscillation (PDO) on inter-annual and decadal timescales, respectively. They
furthermore give new insights into the characteristics and origins of long-range teleconnections, for example, that the teleconnection between ENSO and Indian Ocean dipole exist mainly between the northern part of the ENSO tongue and the equatorial Indian Ocean, and provides therefore valuable knowledge about the regions that are necessary to include when modeling regional climate variability at a certain scale (Agarwal et al., 2019).
In summary, my PhD thesis investigates past and future AMOC variability and its effects on global mean surface temperature by utilizing a combination of observational sea surface data and the output of historical and future climate model simulations from both the high-resolution CM2.6 model as well as the CMIP5 ensemble. It further includes the development and validation of a new method to study climate variability, that, applied to the observed sea surface temperatures, gives new insight about teleconnections in the Earth System. My findings provide evidence that the AMOC has already slowed down, will continue to do so in the future, and will impact the global mean temperature. Further impacts of an AMOC slowdown may include increased sea-level rise at the U.S. east coast (Ezer, 2015), heat extremes in Europe (Duchez et al., 2016) and increased storm activity in the North Atlantic region (Jackson et al., 2015), all of which have significant socio-economic implications.
Der Porenraum eines Karbonatgesteins ist zumeist aus einer spezifischen Vergesellschaftung verschiedenster Porentypen aufgebaut, die eine unterschiedliche Herkunft aufweisen und zusätzlich in ihrer Form und Größe stark variieren können (e.g., Melim et al., 2001; Lee et al., 2009; He et al., 2014; Dernaika & Sinclair, 2017; Zhang et al., 2017). Diese für Karbonate typischen multimodalen Porensysteme entstehen sowohl durch primäre Ablagerungsprozesse, als auch durch mehrmalige Modifikation des Porenraumes nach Ablagerung des Sediments. Dies führt zu einer ungleichen Verteilung der Porenraumeigenschaften auf engstem Raum und das zeitgleiche Auftreten von effektiven und ineffektiven Poren. Diese immanenten Unterschiede in der Effektivität einzelner Porentypen sind der Hauptgrund für die häufig sehr niedrige Korellation zwischen Porosität und Permeabilität in Karbonaten (e.g., Mazzullo 2004; Ehrenberg & Nadeau, 2005; Hollis et al., 2010; He et al., 2014; Rashid et al., 2015; Dernaika & Sinclair, 2017). Durch die Extraktion von miteinander verbundenen und somit effektiven Porentypen jedoch kann das Verständnis und die Vorhersage der Permeabilität für einen gegeben Porositätswert stark verbessert werden (e.g., Melim et al., 2001; Zhang et al., 2017). Dazu wird in dieser Arbeit eine auf der digitalen Bildanalyse (DIA) beruhende Methode vorgestellt, mit der schrittweise die Effektivität von Poren aus den analysierten mittelmiozänen lakustrinen Karbonaten des Nördlinger Ries Kratersees (Süddeutschland) berechnet werden kann. Mithilfe des Porenformfaktors (sensu Anselmetti et al., 1998), der als Parameter zur Quantifizierung der Interkonnektivität zwischen Poren dient, wird der potentiellen Beitrag an Permeabilität jedes Porentyps zur Gesamtpermeabilität bestimmt. Somit können die effektivsten Porentypen innerhalb der analysierten Karbonate identifiziert werden. Desweiteren wird die digitale Bildanalyse dazu benutzt, zementierte Porenräume zu extrahieren, um den Einfluss der Zementation auf die Porenraumeigenschaften zu quantifizieren. Durch eine unabhängige Methode (Fluid-Flow-Simulation), deren Ergebnisse wiederum mit der digitalen Bildanalyse ausgewertet werden, können die vorherigen Erkentnisse bestätigt werden: Interpeloidale Poren und Lösungsporen sind die beiden effektivsten Porentypen im Porenraum der Riesseekarbonate. Die Extraktion des miteinander verbundenen (d.h. effektiven) Porennetzwerkes führt schließlich zu einer erheblich verbesserten Korrelation zwischen Porosität und Permeabilität in den analysierten Karbonaten. Die in dieser Arbeit beschriebene Methode bietet ein quantitatives petrographisches Werkzeug, mit dessen Hilfe die effektive Porosität eines Porenraumes extrahiert werden kann. Dies führt zu einem besseren Verständnis darüber, wie Porensysteme von Karbonaten Permeabilität erzeugen. Diese Dissertation zeigt auch, dass die Formkomplexität von Poren einer der wichtigsten Parameter ist, der die Interkonnektivität zwischen einzelnen Poren und somit die Entstehung von effektiver Porosität steuert. Außerdem erweist sich die digitale Bildanalyse als ausgezeichnetes Werkzeug um die Porosität und Permeabilität direkt an ihren gemeinsamen Ursprung zu knüpfen: die Gesteinstextur und die damit assoziierte Porenstruktur.
Interactions and feedbacks between tectonics, climate, and upper plate architecture control basin geometry, relief, and depositional systems. The Andes is part of a longlived continental margin characterized by multiple tectonic cycles which have strongly modified the Andean upper plate architecture. In the Andean retroarc, spatiotemporal variations in the structure of the upper plate and tectonic regimes have resulted in marked along-strike variations in basin geometry, stratigraphy, deformational style, and mountain belt morphology. These along-strike variations include high-elevation plateaus (Altiplano and Puna) associated with a thin-skin fold-and-thrust-belt and thick-skin deformation in broken foreland basins such as the Santa Barbara system and the Sierras Pampeanas. At the confluence of the Puna Plateau, the Santa Barbara system and the Sierras Pampeanas, major along-strike changes in upper plate architecture, mountain belt morphology, basement exhumation, and deformation style can be recognized. I have used a source to sink approach to unravel the spatiotemporal tectonic evolution of the Andean retroarc between 26 and 28°S. I obtained a large low-temperature thermochronology data set from basement units which includes apatite fission track, apatite U-Th-Sm/He, and zircon U-Th/He (ZHe) cooling ages. Stratigraphic descriptions of Miocene units were temporally constrained by U-Pb LA-ICP-MS zircon ages from interbedded pyroclastic material.
Modeled ZHe ages suggest that the basement of the study area was exhumed during the Famatinian orogeny (550-450 Ma), followed by a period of relative tectonic quiescence during the Paleozoic and the Triassic. The basement experienced horst exhumation during the Cretaceous development of the Salta rift. After initial exhumation, deposition of thick Cretaceous syn-rift strata caused reheating of several basement blocks within the Santa Barbara system. During the Eocene-Oligocene, the Andean compressional setting was responsible for the exhumation of several disconnected basement blocks. These exhumed blocks were separated by areas of low relief, in which humid climate and low erosion rates facilitated the development of etchplains on the crystalline basement. The exhumed basement blocks formed an Eocene to Oligocene broken foreland basin in the back-bulge depozone of the Andean foreland. During the Early Miocene, foreland basin strata filled up the preexisting Paleogene topography. The basement blocks in lower relief positions were reheated; associated geothermal gradients were higher than 25°C/km. Miocene volcanism was responsible for lateral variations on the amount of reheating along the Campo-Arenal basin. Around 12 Ma, a new deformational phase modified the drainage network and fragmented the lacustrine system. As deformation and rock uplift continued, the easily eroded sedimentary cover was efficiently removed and reworked by an ephemeral fluvial system, preventing the development of significant relief. After ~6 Ma, the low erodibility of the basement blocks which began to be exposed caused relief increase, leading to the development of stable fluvial systems. Progressive relief development modified atmospheric circulation, creating a rainfall gradient. After 3 Ma, orographic rainfall and high relief lead to the development of proximal fluvial-gravitational depositional systems in the surrounding basins.
Fold and thrust belts are characteristic features of collisional orogen that grow laterally through time by deforming the upper crust in response to stresses caused by convergence. The deformation propagation in the upper crust is accommodated by shortening along major folds and thrusts. The formation of these structures is influenced by the mechanical strength of décollements, basement architecture, presence of preexisting structures and taper of the wedge. These factors control not only the sequence of deformation but also cause differences in the structural style.
The Himalayan fold and thrust belt exhibits significant differences in the structural style from east to west. The external zone of the Himalayan fold and thrust belt, also called the Subhimalaya, has been extensively studied to understand the temporal development and differences in the structural style in Bhutan, Nepal and India; however, the Subhimalaya in Pakistan remains poorly studied. The Kohat and Potwar fold and thrust belts (herein called Kohat and Potwar) represent the Subhimalaya in Pakistan. The Main Boundary Thrust (MBT) marks the northern boundary of both Kohat and Potwar, showing that these belts are genetically linked to foreland-vergent deformation within the Himalayan orogen, despite the pronounced contrast in structural style. This contrast becomes more pronounced toward south, where the active strike-slip Kalabagh Fault Zone links with the Kohat and Potwar range fronts, known as the Surghar Range and the Salt Range, respectively. The Surghar and Salt Ranges developed above the Surghar Thrust (SGT) and Main Frontal Thrust (MFT). In order to understand the structural style and spatiotemporal development of the major structures in Kohat and Potwar, I have used structural modeling and low temperature thermochronolgy methods in this study. The structural modeling is based on construction of balanced cross-sections by integrating surface geology, seismic reflection profiles and well data. In order to constrain the timing and magnitude of exhumation, I used apatite (U-Th-Sm)/He (AHe) and apatite fission track (AFT) dating. The results obtained from both methods are combined to document the Paleozoic to Recent history of Kohat and Potwar.
The results of this research suggest two major events in the deformation history. The first major deformation event is related to Late Paleozoic rifting associated with the development of the Neo-Tethys Ocean. The second major deformation event is related to the Late Miocene to Pliocene development of the Himalayan fold and thrust belt in the Kohat and Potwar. The Late Paleozoic rifting is deciphered by inverse thermal modelling of detrital AFT and AHe ages from the Salt Range. The process of rifting in this area created normal faulting that resulted in the exhumation/erosion of Early to Middle Paleozoic strata, forming a major unconformity between Cambrian and Permian strata that is exposed today in the Salt Range. The normal faults formed in Late Paleozoic time played an important role in localizing the Miocene-Pliocene deformation in this area. The combination of structural reconstructions and thermochronologic data suggest that deformation initiated at 15±2 Ma on the SGT ramp in the southern part of Kohat. The early movement on the SGT accreted the foreland into the Kohat deforming wedge, forming the range front. The development of the MBT at 12±2 Ma formed the northern boundary of Kohat and Potwar. Deformation propagated south of the MBT in the Kohat on double décollements and in the Potwar on a single basal décollement. The double décollement in the Kohat adopted an active roof-thrust deformation style that resulted in the disharmonic structural style in the upper and lower parts of the stratigraphic section. Incremental shortening resulted in the development of duplexes in the subsurface between two décollements and imbrication above the roof thrust. Tectonic thickening caused by duplexes resulted in cooling and exhumation above the roof thrust by removal of a thick sequence of molasse strata. The structural modelling shows that the ramps on which duplexes formed in Kohat continue as tip lines of fault propagation folds in the Potwar. The absence of a double décollement in the Potwar resulted in the preservation of a thick sequence of molasse strata there. The temporal data suggest that deformation propagated in-sequence from ~ 8 to 3 Ma in the northern part of Kohat and Potwar; however, internal deformation in the Kohat was more intense, probably required for maintaining a critical taper after a significant load was removed above the upper décollement. In the southern part of Potwar, a steeper basement slope (β≥3°) and the presence of salt at the base of the stratigraphic section allowed for the complete preservation of the stratigraphic wedge, showcased by very little internal deformation. Activation of the MFT at ~4 Ma allowed the Salt Range to become the range front of the Potwar. The removal of a large amount of molasse strata above the MFT ramp enhanced the role of salt in shaping the structural style of the Salt Range and Kalabagh Fault Zone. Salt accumulation and migration resulted in the formation of normal faults in both areas. Salt migration in the Kalabagh fault zone has triggered out-of-sequence movement on ramps in the Kohat.
The amount of shortening calculated between the MBT and the SGT in Kohat is 75±5 km and between the MBT and the MFT in Potwar is 65±5 km. A comparable amount of shortening is accommodated in the Kohat and Potwar despite their different widths: 70 km Kohat and 150 km Potwar. In summary, this research suggests that deformation switched between different structures during the last ~15 Ma through different modes of fault propagation, resulting in different structural styles and the out-of-sequence development of Kohat and Potwar.
The foreland of the Andes in South America is characterised by distinct along strike changes in surface deformational styles. These styles are classified into two end-members, the thin-skinned and the thick-skinned style. The superficial expression of thin-skinned deformation is a succession of narrowly spaced hills and valleys, that form laterally continuous ranges on the foreland facing side of the orogen. Each of the hills is defined by a reverse fault that roots in a basal décollement surface within the sedimentary cover, and acted as thrusting ramp to stack the sedimentary pile. Thick-skinned deformation is morphologically characterised by spatially disparate, basement-cored mountain ranges. These mountain ranges are uplifted along reactivated high-angle crustal-scale discontinuities, such as suture zones between different tectonic terranes.
Amongst proposed causes for the observed variation are variations in the dip angle of the Nazca plate, variation in sediment thickness, lithospheric thickening, volcanism or compositional differences. The proposed mechanisms are predominantly based on geological observations or numerical thermomechanical modelling, but there has been no attempt to understand the mechanisms from a point of data-integrative 3D modelling. The aim of this dissertation is therefore to understand how lithospheric structure controls the deformational behaviour. The integration of independent data into a consistent model of the lithosphere allows to obtain additional evidence that helps to understand the causes for the different deformational styles. Northern Argentina encompasses the transition from the thin-skinned fold-and-thrust belt in Bolivia, to the thick-skinned Sierras Pampeanas province, which makes this area a well suited location for such a study. The general workflow followed in this study first involves data-constrained structural- and density-modelling in order to obtain a model of the study area. This model was then used to predict the steady-state thermal field, which was then used to assess the present-day rheological state in northern Argentina.
The structural configuration of the lithosphere in northern Argentina was determined by means of data-integrative, 3D density modelling verified by Bouguer gravity. The model delineates the first-order density contrasts in the lithosphere in the uppermost 200 km, and discriminates bodies for the sediments, the crystalline crust, the lithospheric mantle and the subducting Nazca plate. To obtain the intra-crustal density structure, an automated inversion approach was developed and applied to a starting structural model that assumed a homogeneously dense crust. The resulting final structural model indicates that the crustal structure can be represented by an upper crust with a density of 2800 kg/m³, and a lower crust of 3100 kg/m³. The Transbrazilian Lineament, which separates the Pampia terrane from the Río de la Plata craton, is expressed as a zone of low average crustal densities.
In an excursion, we demonstrate in another study, that the gravity inversion method developed to obtain intra-crustal density structures, is also applicable to obtain density variations in the uppermost lithospheric mantle. Densities in such sub-crustal depths are difficult to constrain from seismic tomographic models due to smearing of crustal velocities. With the application to the uppermost lithospheric mantle in the north Atlantic, we demonstrate in Tan et al. (2018) that lateral density trends of at least 125\,km width are robustly recovered by the inversion method, thereby providing an important tool for the delineation of subcrustal density trends.
Due to the genetic link between subduction, orogenesis and retroarc foreland basins the question rises whether the steady-state assumption is valid in such a dynamic setting. To answer this question, I analysed (i) the impact of subduction on the conductive thermal field of the overlying continental plate, (ii) the differences between the transient and steady-state thermal fields of a geodynamic coupled model. Both studies indicate that the assumption of a thermal steady-state is applicable in most parts of the study area. Within the orogenic wedge, where the assumption cannot be applied, I estimated the transient thermal field based on the results of the conducted analyses.
Accordingly, the structural model that had been obtained in the first step, could be used to obtain a 3D conductive steady-state thermal field. The rheological assessment based on this thermal field indicates that the lithosphere of the thin-skinned Subandean ranges is characterised by a relatively strong crust and a weak mantle. Contrarily, the adjacent foreland basin consists of a fully coupled, very strong lithosphere. Thus, shortening in northern Argentina can only be accommodated within the weak lithosphere of the orogen and the Subandean ranges. The analysis suggests that the décollements of the fold-and-thrust belt are the shallow continuation of shear zones that reside in the ductile sections of the orogenic crust. Furthermore, the localisation of the faults that provide strain transfer between the deeper ductile crust and the shallower décollement is strongly influenced by crustal weak zones such as foliation. In contrast to the northern foreland, the lithosphere of the thick-skinned Sierras Pampeanas is fully coupled and characterised by a strong crust and mantle. The high overall strength prevents the generation of crustal-scale faults by tectonic stresses. Even inherited crustal-scale discontinuities, such as sutures, cannot sufficiently reduce the strength of the lithosphere in order to be reactivated. Therefore, magmatism that had been identified to be a precursor of basement uplift in the Sierras Pampeanas, is the key factor that leads to the broken foreland of this province. Due to thermal weakening, and potentially lubrication of the inherited discontinuities, the lithosphere is locally weakened such that tectonic stresses can uplift the basement blocks. This hypothesis explains both the spatially disparate character of the broken foreland, as well as the observed temporal delay between volcanism and basement block uplift.
This dissertation provides for the first time a data-driven 3D model that is consistent with geophysical data and geological observations, and that is able to causally link the thermo-rheological structure of the lithosphere to the observed variation of surface deformation styles in the retroarc foreland of northern Argentina.
This dissertation combines field and geochemical observations and analyses with numerical modeling to understand the formation of vein-hosted Sn-W ore in the Panasqueira deposit of Portugal, which is among the ten largest worldwide. The deposit is located above a granite body that is altered by magmatic-hydrothermal fluids in its upper part (greisen). These fluids are thought to be the source of metals, but that was still under debate. The goal of this study is to determine the composition and temperature of hydrothermal fluids at Panasqueira, and with that information to construct a numerical model of the hydrothermal system. The focus is on analysis of the minerals tourmaline and white mica, which formed during mineralization and are widespread throughout the deposit. Tourmaline occurs mainly in alteration zones around mineralized veins and is less abundant in the vein margins. White mica is more widespread. It is abundant in vein margins as well as alteration zones, and also occurs in the granite greisen. The laboratory work involved in-situ microanalysis of major- and trace elements in tourmaline and white mica, and boron-isotope analysis in both minerals by secondary ion mass spectrometry (SIMS).
The boron-isotope composition of tourmaline and white mica suggests a magmatic source. Comparison of hydrothermally-altered and unaltered rocks from drill cores shows that the ore metals (W, Sn, Cu, and Zn) and As, F, Li, Rb, and Cs were introduced during the alteration. Most of these elements are also enriched in tourmaline and mica, which confirms their potential value as exploration guides to Sn-W ores elsewhere.
The thermal evolution of the hydrothermal system was estimated by B-isotope exchange thermometry and the Ti-in-quartz method. Both methods yielded similar temperatures for the early hydrothermal phase: 430° to 460°C for B-isotopes and 503° ± 24°C for Ti-in-quartz. Mineral pairs from a late fault zone yield significantly lower median temperatures of 250°C. The combined results of thermometry with variations in chemical and B-isotope composition of tourmaline and mica suggest that a similar magmatic-hydrothermal fluid was active at all stages of mineralization. Mineralization in the late stage shows the same B-isotope composition as in the main stage despite a ca. 250°C cooling, which supports a multiple injection model of magmatic-hydrothermal fluids.
Two-dimensional numerical simulations of convection in a multiphase NaCl hydrothermal system were conducted: (a) in order to test a new approach (lower dimensional elements) for flow through fractures and faults and (b) in order to identify conditions for horizontal fluid flow as observed in the flat-lying veins at Panasqueira. The results show that fluid flow over an intrusion (heat and fluid source) develops a horizontal component if there is sufficient fracture connectivity. Late, steep fault zones have been identified in the deposit area, which locally contain low-temperature Zn-Pb mineralization. The model results confirm that the presence of subvertical faults with enhanced permeability play a crucial role in the ascent of magmatic fluids to the surface and the recharge of meteoric waters. Finally, our model results suggest that recharge of meteoric fluids and mixing processes may be important at later stages, while flow of magmatic fluids dominate the early stages of the hydrothermal fluid circulation.
Partial melting is a first order process for the chemical differentiation of the crust (Vielzeuf et al., 1990). Redistribution of chemical elements during melt generation crucially influences the composition of the lower and upper crust and provides a mechanism to concentrate and transport chemical elements that may also be of economic interest. Understanding of the diverse processes and their controlling factors is therefore not only of scientific interest but also of high economic importance to cover the demand for rare metals.
The redistribution of major and trace elements during partial melting represents a central step for the understanding how granite-bound mineralization develops (Hedenquist and Lowenstern, 1994). The partial melt generation and mobilization of ore elements (e.g. Sn, W, Nb, Ta) into the melt depends on the composition of the sedimentary source and melting conditions. Distinct source rocks have different compositions reflecting their deposition and alteration histories. This specific chemical “memory” results in different mineral assemblages and melting reactions for different protolith compositions during prograde metamorphism (Brown and Fyfe, 1970; Thompson, 1982; Vielzeuf and Holloway, 1988). These factors do not only exert an important influence on the distribution of chemical elements during melt generation, they also influence the volume of melt that is produced, extraction of the melt from its source, and its ascent through the crust (Le Breton and Thompson, 1988). On a larger scale, protolith distribution and chemical alteration (weathering), prograde metamorphism with partial melting, melt extraction, and granite emplacement are ultimately depending on a (plate-)tectonic control (Romer and Kroner, 2016). Comprehension of the individual stages and their interaction is crucial in understanding how granite-related mineralization forms, thereby allowing estimation of the mineralization potential of certain areas. Partial melting also influences the isotope systematics of melt and restite. Radiogenic and stable isotopes of magmatic rocks are commonly used to trace back the source of intrusions or to quantify mixing of magmas from different sources with distinct isotopic signatures (DePaolo and Wasserburg, 1979; Lesher, 1990; Chappell, 1996). These applications are based on the fundamental requirement that the isotopic signature in the melt reflects that of the bulk source from which it is derived. Different minerals in a protolith may have isotopic compositions of radiogenic isotopes that deviate from their whole rock signature (Ayres and Harris, 1997; Knesel and Davidson, 2002). In particular, old minerals with a distinct parent-to-daughter (P/D) ratio are expected to have a specific radiogenic isotope signature. As the partial melting reaction only involves selective phases in a protolith, the isotopic signature of the melt reflects that of the minerals involved in the melting reaction and, therefore, should be different from the bulk source signature. Similar considerations hold true for stable isotopes.
The thesis comprises three experimental studies, which were carried out to unravel the short- as well as the long-term mechanical properties of shale rocks. Short-term mechanical properties such as compressive strength and Young’s modulus were taken from recorded stress-strain curves of constant strain rate tests. Long-term mechanical properties are represented by the time– dependent creep behavior of shales. This was obtained from constant stress experiments, where the test duration ranged from a couple minutes up to two weeks. A profound knowledge of the mechanical behavior of shales is crucial to reliably estimate the potential of a shale reservoir for an economical and sustainable extraction of hydrocarbons (HC). In addition, healing of clay-rich forming cap rocks involving creep and compaction is important for underground storage of carbon dioxide and nuclear waste.
Chapter 1 introduces general aspects of the research topic at hand and highlights the motivation for conducting this study. At present, a shift from energy recovered from conventional resources e.g., coal towards energy provided by renewable resources such as wind or water is a big challenge. Gas recovered from unconventional reservoirs (shale plays) is considered a potential bridge technology.
In Chapter 2, short-term mechanical properties of two European mature shale rocks are presented, which were determined from constant strain rate experiments performed at ambient and in situ deformation conditions (confining pressure, pc ≤ 100 MPa, temperature, T ≤ 125 °C, representing pc, T - conditions at < 4 km depth) using a Paterson– type gas deformation apparatus. The investigated shales were mainly from drill core material of Posidonia (Germany) shale and weathered material of Bowland (United Kingdom) shale. The results are compared with mechanical properties of North American shales. Triaxial compression tests performed perpendicular to bedding revealed semibrittle deformation behavior of Posidonia shale with pronounced inelastic deformation. This is in contrast to Bowland shale samples that deformed brittle and displayed predominantly elastic deformation. The static Young’s modulus, E, and triaxial compressive strength, σTCS, determined from recorded stress-strain curves strongly depended on the applied confining pressure and sample composition, whereas the influence of temperature and strain rate on E and σTCS was minor. Shales with larger amounts of weak minerals (clay, mica, total organic carbon) yielded decreasing E and σTCS. This may be related to a shift from deformation supported by a load-bearing framework of hard phases (e.g., quartz) towards deformation of interconnected weak minerals, particularly for higher fractions of about 25 – 30 vol% weak phases. Comparing mechanical properties determined at reservoir conditions with mechanical data applying effective medium theories revealed that E and σTCS of Posidonia and Bowland shale are close to the lower (Reuss) bound. Brittleness B is often quoted as a measure indicating the response of a shale formation to stimulation and economic production. The brittleness, B, of Posidonia and Bowland shale, estimated from E, is in good agreement with the experimental results. This correlation may be useful to predict B from sonic logs, from which the (dynamic) Young’s modulus can be retrieved.
Chapter 3 presents a study of the long-term creep properties of an immature Posidonia shale. Constant stress experiments (σ = const.) were performed at elevated confining pressures (pc = 50 – 200 MPa) and temperatures (T = 50 – 200 °C) to simulate reservoir pc, T - conditions. The Posidonia shale samples were acquired from a quarry in South Germany. At stresses below ≈ 84 % compressive strength of Posidonia shale, at high temperature and low confining pressure, samples showed pronounced transient (primary) creep with high deformation rates in the semibrittle regime. Sample deformation was mainly accommodated by creep of weak sample constituents and pore space reduction. An empirical power law relation between strain and time, which also accounts for the influence of pc, T and σ on creep strain was formulated to describe the primary creep phase. Extrapolation of the results to a creep period of several years, which is the typical time interval for a large production decline, suggest that fracture closure is unlikely at low stresses. At high stresses as expected for example at the contact between the fracture surfaces and proppants added during stimulation measures, subcritical crack growth may lead to secondary and tertiary creep. An empirical power law is suggested to describe secondary creep of shale rocks as a function of stress, pressure and temperature. The predicted closure rates agree with typical production decline curves recorded during the extraction of hydrocarbons. At the investigated conditions, the creep behavior of Posidonia shale was found to correlate with brittleness, calculated from sample composition.
In Chapter 4 the creep properties of mature Posidonia and Bowland shales are presented. The observed long-term creep behavior is compared to the short-term behavior determined in Chapter 2. Creep experiments were performed at simulated reservoir conditions of pc = 50 – 115 MPa and T = 75 – 150 °C. Similar to the mechanical response of immature Posidonia shale samples investigated in Chapter 3, creep strain rates of mature Bowland and Posidonia shales were enhanced with increasing stress and temperature and decreasing confining pressures. Depending on applied deformation conditions, samples displayed either only a primary (decelerating) or in addition also a secondary (quasi-steady state) and subsequently a tertiary (accelerating) creep phase before failure. At the same deformation conditions, creep strain of Posidonia shale, which is rich in weak constituents, is tremendously higher than of quartz-rich Bowland shale. Typically, primary creep strain is again mostly accommodated by deformation of weak minerals and local pore space reduction. At the onset of tertiary creep most of the deformation was accommodated by micro crack growth. A power law was used to characterize the primary creep phase of Posidonia and Bowland shale. Primary creep strain of shale rocks is inversely correlated to triaxial compressive strength and brittleness, as described in Chapter 2.
Chapter 5 provides a synthesis of the experimental findings and summarizes the major results of the studies presented in Chapters 2 – 4 and potential applications in the Exploration & Production industry.
Chapter 6 gives a brief outlook on potential future experimental research that would help to further improve our understanding of processes leading to fracture closure involving proppant embedment in unconventional shale gas reservoirs. Such insights may allow to improve stimulation techniques aimed at maintaining economical extraction of hydrocarbons over several years.
Earthquake swarms are characterized by large numbers of events occurring in a short period of time within a confined source volume and without significant mainshock aftershock pattern as opposed to tectonic sequences. Intraplate swarms in the absence of active volcanism usually occur in continental rifts as for example in the Eger Rift zone in North West Bohemia, Czech Republic. A common hypothesis links event triggering to pressurized fluids. However, the exact causal chain is often poorly understood since the underlying geotectonic processes are slow compared to tectonic sequences. The high event rate during active periods challenges standard seismological routines as these are often designed for single events and therefore costly in terms of human resources when working with phase picks or computationally costly when exploiting full waveforms.
This methodological thesis develops new approaches to analyze earthquake swarm seismicity as well as the underlying seismogenic volume. It focuses on the region of North West (NW) Bohemia, a well studied, well monitored earthquake swarm region.
In this work I develop and test an innovative approach to detect and locate earthquakes using deep convolutional neural networks. This technology offers great potential as it allows to efficiently process large amounts of data which becomes increasingly important given that seismological data storage grows at increasing pace. The proposed deep neural network trained on NW Bohemian earthquake swarm records is able to locate 1000 events in less than 1 second using full waveforms while approaching precision of double difference relocated catalogs. A further technological novelty is that the trained filters of the deep neural network’s first layer can be repurposed to function as a pattern matching event detector without additional training on noise datasets. For further methodological development and benchmarking, I present a new toolbox to generate realistic earthquake cluster catalogs as well as synthetic full waveforms of those clusters in an automated fashion. The input is parameterized using constraints on source volume geometry, nucleation and frequency-magnitude relations. It harnesses recorded noise to produce highly realistic synthetic data for benchmarking and development. This tool is used to study and assess detection performance in terms of magnitude of completeness Mc of a full waveform detector applied to synthetic data of a hydrofracturing experiment at the Wysin site, Poland.
Finally, I present and demonstrate a novel approach to overcome the masking effects of wave propagation between earthquake and stations and to determine source volume attenuation directly in the source volume where clustered earthquakes occur. The new event couple spectral ratio approach exploits high frequency spectral slopes of two events sharing the greater part of their rays. Synthetic tests based on the toolbox mentioned before show that this method is able to infer seismic wave attenuation within the source volume at high spatial resolution. Furthermore, it is independent from the distance towards a station as well as the complexity of the attenuation and velocity structure outside of the source volume of swarms. The application to recordings of the NW Bohemian earthquake swarm shows increased P phase attenuation within the source volume (Qp < 100) based on results at a station located close to the village Luby (LBC). The recordings of a station located in epicentral proximity, close to Nový Kostel (NKC), show a relatively high complexity indicating that waves arriving at that station experience more scattering than signals recorded at other stations. The high level of complexity destabilizes the inversion. Therefore, the Q estimate at NKC is not reliable and an independent proof of the high attenuation finding given the geometrical and frequency constraints is still to be done. However, a high attenuation in the source volume of NW Bohemian swarms has been postulated before in relation to an expected, highly damaged zone bearing CO 2 at high pressure.
The methods developed in the course of this thesis yield the potential to improve our understanding regarding the role of fluids and gases in intraplate event clustering.