Refine
Has Fulltext
- yes (1746) (remove)
Year of publication
Document Type
- Doctoral Thesis (1746) (remove)
Language
- English (1746) (remove)
Keywords
- climate change (49)
- Klimawandel (48)
- Modellierung (26)
- Nanopartikel (22)
- machine learning (21)
- Fernerkundung (17)
- Blickbewegungen (16)
- Synchronisation (15)
- remote sensing (15)
- Arktis (14)
Institute
- Institut für Physik und Astronomie (351)
- Institut für Geowissenschaften (300)
- Institut für Biochemie und Biologie (263)
- Institut für Chemie (193)
- Extern (129)
- Hasso-Plattner-Institut für Digital Engineering GmbH (87)
- Institut für Umweltwissenschaften und Geographie (84)
- Institut für Informatik und Computational Science (65)
- Department Linguistik (64)
- Institut für Mathematik (57)
Deep learning has seen widespread application in many domains, mainly for its ability to learn data representations from raw input data. Nevertheless, its success has so far been coupled with the availability of large annotated (labelled) datasets. This is a requirement that is difficult to fulfil in several domains, such as in medical imaging. Annotation costs form a barrier in extending deep learning to clinically-relevant use cases. The labels associated with medical images are scarce, since the generation of expert annotations of multimodal patient data at scale is non-trivial, expensive, and time-consuming. This substantiates the need for algorithms that learn from the increasing amounts of unlabeled data. Self-supervised representation learning algorithms offer a pertinent solution, as they allow solving real-world (downstream) deep learning tasks with fewer annotations. Self-supervised approaches leverage unlabeled samples to acquire generic features about different concepts, enabling annotation-efficient downstream task solving subsequently.
Nevertheless, medical images present multiple unique and inherent challenges for existing self-supervised learning approaches, which we seek to address in this thesis: (i) medical images are multimodal, and their multiple modalities are heterogeneous in nature and imbalanced in quantities, e.g. MRI and CT; (ii) medical scans are multi-dimensional, often in 3D instead of 2D; (iii) disease patterns in medical scans are numerous and their incidence exhibits a long-tail distribution, so it is oftentimes essential to fuse knowledge from different data modalities, e.g. genomics or clinical data, to capture disease traits more comprehensively; (iv) Medical scans usually exhibit more uniform color density distributions, e.g. in dental X-Rays, than natural images. Our proposed self-supervised methods meet these challenges, besides significantly reducing the amounts of required annotations.
We evaluate our self-supervised methods on a wide array of medical imaging applications and tasks. Our experimental results demonstrate the obtained gains in both annotation-efficiency and performance; our proposed methods outperform many approaches from related literature. Additionally, in case of fusion with genetic modalities, our methods also allow for cross-modal interpretability. In this thesis, not only we show that self-supervised learning is capable of mitigating manual annotation costs, but also our proposed solutions demonstrate how to better utilize it in the medical imaging domain. Progress in self-supervised learning has the potential to extend deep learning algorithms application to clinical scenarios.
Arctic climate change is marked by intensified warming compared to global trends and a significant reduction in Arctic sea ice which can intricately influence mid-latitude atmospheric circulation through tropo- and stratospheric pathways. Achieving accurate simulations of current and future climate demands a realistic representation of Arctic climate processes in numerical climate models, which remains challenging.
Model deficiencies in replicating observed Arctic climate processes often arise due to inadequacies in representing turbulent boundary layer interactions that determine the interactions between the atmosphere, sea ice, and ocean. Many current climate models rely on parameterizations developed for mid-latitude conditions to handle Arctic turbulent boundary layer processes.
This thesis focuses on modified representation of the Arctic atmospheric processes and understanding their resulting impact on large-scale mid-latitude atmospheric circulation within climate models. The improved turbulence parameterizations, recently developed based on Arctic measurements, were implemented in the global atmospheric circulation model ECHAM6. This involved modifying the stability functions over sea ice and ocean for stable stratification and changing the roughness length over sea ice for all stratification conditions. Comprehensive analyses are conducted to assess the impacts of these modifications on ECHAM6's simulations of the Arctic boundary layer, overall atmospheric circulation, and the dynamical pathways between the Arctic and mid-latitudes.
Through a step-wise implementation of the mentioned parameterizations into ECHAM6, a series of sensitivity experiments revealed that the combined impacts of the reduced roughness length and the modified stability functions are non-linear. Nevertheless, it is evident that both modifications consistently lead to a general decrease in the heat transfer coefficient, being in close agreement with the observations.
Additionally, compared to the reference observations, the ECHAM6 model falls short in accurately representing unstable and strongly stable conditions.
The less frequent occurrence of strong stability restricts the influence of the modified stability functions by reducing the affected sample size. However, when focusing solely on the specific instances of a strongly stable atmosphere, the sensible heat flux approaches near-zero values, which is in line with the observations. Models employing commonly used surface turbulence parameterizations were shown to have difficulties replicating the near-zero sensible heat flux in strongly stable stratification.
I also found that these limited changes in surface layer turbulence parameterizations have a statistically significant impact on the temperature and wind patterns across multiple pressure levels, including the stratosphere, in both the Arctic and mid-latitudes. These significant signals vary in strength, extent, and direction depending on the specific month or year, indicating a strong reliance on the background state.
Furthermore, this research investigates how the modified surface turbulence parameterizations may influence the response of both stratospheric and tropospheric circulation to Arctic sea ice loss.
The most suitable parameterizations for accurately representing Arctic boundary layer turbulence were identified from the sensitivity experiments. Subsequently, the model's response to sea ice loss is evaluated through extended ECHAM6 simulations with different prescribed sea ice conditions.
The simulation with adjusted surface turbulence parameterizations better reproduced the observed Arctic tropospheric warming in vertical extent, demonstrating improved alignment with the reanalysis data. Additionally, unlike the control experiments, this simulation successfully reproduced specific circulation patterns linked to the stratospheric pathway for Arctic-mid-latitude linkages. Specifically, an increased occurrence of the Scandinavian-Ural blocking regime (negative phase of the North Atlantic Oscillation) in early (late) winter is observed. Overall, it can be inferred that improving turbulence parameterizations at the surface layer can improve the ECHAM6's response to sea ice loss.
The wide distribution of location-acquisition technologies means that large volumes of spatio-temporal data are continuously being accumulated. Positioning systems such as GPS enable the tracking of various moving objects' trajectories, which are usually represented by a chronologically ordered sequence of observed locations. The analysis of movement patterns based on detailed positional information creates opportunities for applications that can improve business decisions and processes in a broad spectrum of industries (e.g., transportation, traffic control, or medicine). Due to the large data volumes generated in these applications, the cost-efficient storage of spatio-temporal data is desirable, especially when in-memory database systems are used to achieve interactive performance requirements.
To efficiently utilize the available DRAM capacities, modern database systems support various tuning possibilities to reduce the memory footprint (e.g., data compression) or increase performance (e.g., additional indexes structures). By considering horizontal data partitioning, we can independently apply different tuning options on a fine-grained level. However, the selection of cost and performance-balancing configurations is challenging, due to the vast number of possible setups consisting of mutually dependent individual decisions.
In this thesis, we introduce multiple approaches to improve spatio-temporal data management by automatically optimizing diverse tuning options for the application-specific access patterns and data characteristics. Our contributions are as follows:
(1) We introduce a novel approach to determine fine-grained table configurations for spatio-temporal workloads. Our linear programming (LP) approach jointly optimizes the (i) data compression, (ii) ordering, (iii) indexing, and (iv) tiering. We propose different models which address cost dependencies at different levels of accuracy to compute optimized tuning configurations for a given workload, memory budgets, and data characteristics. To yield maintainable and robust configurations, we further extend our LP-based approach to incorporate reconfiguration costs as well as optimizations for multiple potential workload scenarios.
(2) To optimize the storage layout of timestamps in columnar databases, we present a heuristic approach for the workload-driven combined selection of a data layout and compression scheme. By considering attribute decomposition strategies, we are able to apply application-specific optimizations that reduce the memory footprint and improve performance.
(3) We introduce an approach that leverages past trajectory data to improve the dispatch processes of transportation network companies. Based on location probabilities, we developed risk-averse dispatch strategies that reduce critical delays.
(4) Finally, we used the use case of a transportation network company to evaluate our database optimizations on a real-world dataset. We demonstrate that workload-driven fine-grained optimizations allow us to reduce the memory footprint (up to 71% by equal performance) or increase the performance (up to 90% by equal memory size) compared to established rule-based heuristics.
Individually, our contributions provide novel approaches to the current challenges in spatio-temporal data mining and database research. Combining them allows in-memory databases to store and process spatio-temporal data more cost-efficiently.
This thesis presents an attempt to use source code synthesised from Coq formalisations of device drivers for existing (micro)kernel operating systems, with a particular focus on the Linux Kernel.
In the first part, the technical background and related work are described. The focus is here on the possible approaches to synthesising certified software with Coq, namely the extraction to functional languages using the Coq extraction plugin and the extraction to Clight code using the CertiCoq plugin. It is noted that the implementation of CertiCoq is verified, whereas this is not the case for the Coq extraction plugin. Consequently, there is a correctness guarantee for the generated Clight code which does not hold for the code being generated by the Coq extraction plugin. Furthermore, the differences between user space and kernel space software are discussed in relation to Linux device drivers. It is elaborated that it is not possible to generate working Linux kernel module components using the Coq extraction plugin without significant modifications. In contrast, it is possible to produce working user space drivers both with the Coq extraction plugin and CertiCoq. The subsequent parts describe the main contributions of the thesis.
In the second part, it is demonstrated how to extend the Coq extraction plugin to synthesise foreign function calls between the functional language OCaml and the imperative language C. This approach has the potential to improve the type-safety of user space drivers. Furthermore, it is shown that the code being synthesised by CertiCoq cannot be used in kernel space without modifications to the necessary runtime. Consequently, the necessary modifications to the runtimes of CertiCoq and VeriFFI are introduced, resulting in the runtimes becoming compatible components of a Linux kernel module. Furthermore, justifications for the transformations are provided and possible further extensions to both plugins and solutions to failing garbage collection calls in kernel space are discussed.
The third part presents a proof of concept device driver for the Linux Kernel. To achieve this, the event handler of the original PC Speaker driver is partially formalised in Coq. Furthermore, some relevant formal properties of the formalised functionality are discussed. Subsequently, a kernel module is defined, utilising the modified variants of CertiCoq and VeriFFI to compile a working device driver. It is furthermore shown that it is possible to compile the synthesised code with CompCert, thereby extending the guarantee of correctness to the assembly layer. This is followed by a performance evaluation that compares a naive formalisation of the PC speaker functionality with the original PC Speaker driver pointing out the weaknesses in the formalisation and possible improvements. The part closes with a summary of the results, their implications and open questions being raised.
The last part lists all used sources, separated into scientific literature, documentations or reference manuals and artifacts, i.e. source code.
This thesis focuses on the molecular evolution of Macroscelidea, commonly referred to as sengis. Sengis are a mammalian order belonging to the Afrotherians, one of the four major clades of placental mammals. Sengis currently consist of twenty extant species, all of which are endemic to the African continent. They can be separated in two families, the soft-furred sengis (Macroscelididae) and the giant sengis (Rhynchocyonidae). While giant sengis can be exclusively found in forest habitats, the different soft-furred sengi species dwell in a broad range of habitats, from tropical rain-forests to rocky deserts.
Our knowledge on the evolutionary history of sengis is largely incomplete. The high level of superficial morphological resemblance among different sengi species (especially the soft-furred sengis) has for example led to misinterpretations of phylogenetic relationships, based on morphological characters. With the rise of DNA based taxonomic inferences, multiple new genera were defined and new species described. Yet, no full taxon molecular phylogeny exists, hampering the answering of basic taxonomic questions. This lack of knowledge can be to some extent attributed to the limited availability of fresh-tissue samples for DNA extraction. The broad African distribution, partly in political unstable regions and low population densities complicate contemporary sampling approaches. Furthermore, the DNA information available usually covers only short stretches of the mitochondrial genome and thus a single genetic locus with limited informational content.
Developments in DNA extraction and library protocols nowadays offer the opportunity to access DNA from museum specimens, collected over the past centuries and stored in natural history museums throughout the world. Thus, the difficulties in fresh-sample acquisition for molecular biological studies can be overcome by the application of museomics, the research field which emerged from those laboratory developments.
This thesis uses fresh-tissue samples as well as a vast collection museum specimens to investigate multiple aspects about the macroscelidean evolutionary history. Chapter 4 of this thesis focuses on the phylogenetic relationships of all currently known sengi species. By accessing DNA information from museum specimens in combination of fresh tissue samples and publicly available genetic resources it produces the first full taxon molecular phylogeny of sengis. It confirms the monophyly of the genus Elephantulus and discovers multiple deeply divergent lineages within different species, highlighting the need for species specific approaches. The study furthermore focuses on the evolutionary time frame of sengis by evaluating the impact of commonly varied parameters on tree dating. The results of the study show, that the mitochondrial information used in previous studies to temporal calibrate the Macroscelidean phylogeny led to an overestimation of node ages within sengis. Especially soft-furred sengis are thus much younger than previously assumed. The refined knowledge of nodes ages within sengis offer the opportunity to link e.g. speciation events to environmental changes.
Chapter 5 focuses on the genus Petrodromus with its single representative Petrodromus tetradactylus. It again exploits the opportunities of museomics and gathers a comprehensive, multi-locus genetic dataset of P. tetradactylus individuals, distributed across most the known range of this species. It reveals multiple deeply divergent lineages within Petrodromus, whereby some could possibly be associated to previously described sub-species, at least one was formerly unknown. It underscores the necessity for a revision of the genus Petrodromus through the integration of both molecular and morphological evidence. The study, furthermore identifies changing forest distributions through climatic oscillations as main factor shaping the genetic structure of Petrodromus.
Chapter 6 uses fresh tissue samples to extent the genomic resources of sengis by thirteen new nuclear genomes, of which two were de-novo assembled. An extensive dataset of more than 8000 protein coding one-to-one orthologs allows to further refine and confirm the temporal time frame of sengi evolution found in Chapter 4. This study moreover investigates the role of gene-flow and incomplete lineage sorting (ILS) in sengi evolution. In addition it identifies clade specific genes of possible outstanding evolutionary importance and links them to potential phenotypic traits affected. A closer investigation of olfactory receptor proteins reveals clade specific differences. A comparison of the demographic past of sengis to other small African mammals does not reveal a sengi specific pattern.
Ecosystems play a pivotal role in addressing climate change but are also highly susceptible to drastic environmental changes. Investigating their historical dynamics can enhance our understanding of how they might respond to unprecedented future environmental shifts. With Arctic lakes currently under substantial pressure from climate change, lessons from the past can guide our understanding of potential disruptions to these lakes. However, individual lake systems are multifaceted and complex. Traditional isolated lake studies often fail to provide a global perspective because localized nuances—like individual lake parameters, catchment areas, and lake histories—can overshadow broader conclusions. In light of these complexities, a more nuanced approach is essential to analyze lake systems in a global context.
A key to addressing this challenge lies in the data-driven analysis of sedimentological records from various northern lake systems. This dissertation emphasizes lake systems in the northern Eurasian region, particularly in Russia (n=59). For this doctoral thesis, we collected sedimentological data from various sources, which required a standardized framework for further analysis. Therefore, we designed a conceptual model for integrating and standardizing heterogeneous multi-proxy data into a relational database management system (PostgreSQL). Creating a database from the collected data enabled comparative numerical analyses between spatially separated lakes as well as between different proxies.
When analyzing numerous lakes, establishing a common frame of reference was crucial. We achieved this by converting proxy values from depth dependency to age dependency. This required consistent age calculations across all lakes and proxies using one age-depth modeling software. Recognizing the broader implications and potential pitfalls of this, we developed the LANDO approach ("Linked Age and Depth Modelling"). LANDO is an innovative integration of multiple age-depth modeling software into a singular, cohesive platform (Jupyter Notebook). Beyond its ability to aggregate data from five renowned age-depth modeling software, LANDO uniquely empowers users to filter out implausible model outcomes using robust geoscientific data. Our method is not only novel but also significantly enhances the accuracy and reliability of lake analyses.
Considering the preceding steps, this doctoral thesis further examines the relationship between carbon in sediments and temperature over the last 21,000 years. Initially, we hypothesized a positive correlation between carbon accumulation in lakes and modelled paleotemperature. Our homogenized dataset from heterogeneous lakes confirmed this association, even if the highest temperatures throughout our observation period do not correlate with the highest carbon values. We assume that rapid warming events contribute more to high accumulation, while sustained warming leads to carbon outgassing. Considering the current high concentration of carbon in the atmosphere and rising temperatures, ongoing climate change could cause northern lake systems to contribute to a further increase in atmospheric carbon (positive feedback loop). While our findings underscore the reliability of both our standardized data and the LANDO method, expanding our dataset might offer even greater assurance in our conclusions.
Improving permafrost dynamics in land surface models: insights from dual sensitivity experiments
(2024)
The thawing of permafrost and the subsequent release of greenhouse gases constitute one of the most significant and uncertain positive feedback loops in the context of climate change, making predictions regarding changes in permafrost coverage of paramount importance. To address these critical questions, climate scientists have developed Land Surface Models (LSMs) that encompass a multitude of physical soil processes. This thesis is committed to advancing our understanding and refining precise representations of permafrost dynamics within LSMs, with a specific focus on the accurate modeling of heat fluxes, an essential component for simulating permafrost physics.
The first research question overviews fundamental model prerequisites for the representation of permafrost soils within land surface modeling. It includes a first-of-its-kind comparison between LSMs in CMIP6 to reveal their differences and shortcomings in key permafrost physics parameters. Overall, each of these LSMs represents a unique approach to simulating soil processes and their interactions with the climate system. Choosing the most appropriate model for a particular application depends on factors such as the spatial and temporal scale of the simulation, the specific research question, and available computational resources.
The second research question evaluates the performance of the state-of-the-art Community Land Model (CLM5) in simulating Arctic permafrost regions. Our approach overcomes traditional evaluation limitations by individually addressing depth, seasonality, and regional variations, providing a comprehensive assessment of permafrost and soil temperature dynamics. I compare CLM5's results with three extensive datasets: (1) soil temperatures from 295 borehole stations, (2) active layer thickness (ALT) data from the Circumpolar Active Layer Monitoring Network (CALM), and (3) soil temperatures, ALT, and permafrost extent from the ESA Climate Change Initiative (ESA-CCI). The results show that CLM5 aligns well with ESA-CCI and CALM for permafrost extent and ALT but reveals a significant global cold temperature bias, notably over Siberia. These results echo a persistent challenge identified in numerous studies: the existence of a systematic 'cold bias' in soil temperature over permafrost regions. To address this challenge, the following research questions propose dual sensitivity experiments.
The third research question represents the first study to apply a Plant Functional Type (PFT)-based approach to derive soil texture and soil organic matter (SOM), departing from the conventional use of coarse-resolution global data in LSMs. This novel method results in a more uniform distribution of soil organic matter density (OMD) across the domain, characterized by reduced OMD values in most regions. However, changes in soil texture exhibit a more intricate spatial pattern. Comparing the results to observations reveals a significant reduction in the cold bias observed in the control run. This method shows noticeable improvements in permafrost extent, but at the cost of an overestimation in ALT. These findings emphasize the model's high sensitivity to variations in soil texture and SOM content, highlighting the crucial role of soil composition in governing heat transfer processes and shaping the seasonal variation of soil temperatures in permafrost regions.
Expanding upon a site experiment conducted in Trail Valley Creek by \citet{dutch_impact_2022}, the fourth research question extends the application of the snow scheme proposed by \citet{sturm_thermal_1997} to cover the entire Arctic domain. By employing a snow scheme better suited to the snow density profile observed over permafrost regions, this thesis seeks to assess its influence on simulated soil temperatures. Comparing this method to observational datasets reveals a significant reduction in the cold bias that was present in the control run. In most regions, the Sturm run exhibits a substantial decrease in the cold bias. However, there is a distinctive overshoot with a warm bias observed in mountainous areas. The Sturm experiment effectively addressed the overestimation of permafrost extent in the control run, albeit resulting in a substantial reduction in permafrost extent over mountainous areas. ALT results remain relatively consistent compared to the control run. These outcomes align with our initial hypothesis, which anticipated that the reduced snow insulation in the Sturm run would lead to higher winter soil temperatures and a more accurate representation of permafrost physics.
In summary, this thesis demonstrates significant advancements in understanding permafrost dynamics and its integration into LSMs. It has meticulously unraveled the intricacies involved in the interplay between heat transfer, soil properties, and snow dynamics in permafrost regions. These insights offer novel perspectives on model representation and performance.
Today, near-surface investigations are frequently conducted using non-destructive or minimally invasive methods of applied geophysics, particularly in the fields of civil engineering, archaeology, geology, and hydrology. One field that plays an increasingly central role in research and engineering is the examination of sedimentary environments, for example, for characterizing near-surface groundwater systems. A commonly employed method in this context is ground-penetrating radar (GPR). In this technique, short electromagnetic pulses are emitted into the subsurface by an antenna, which are then reflected, refracted, or scattered at contrasts in electromagnetic properties (such as the water table). A receiving antenna records these signals in terms of their amplitudes and travel times. Analysis of the recorded signals allows for inferences about the subsurface, such as the depth of the groundwater table or the composition and characteristics of near-surface sediment layers. Due to the high resolution of the GPR method and continuous technological advancements, GPR data acquisition is increasingly performed in three-dimensional (3D) fashion today.
Despite the considerable temporal and technical efforts involved in data acquisition and processing, the resulting 3D data sets (providing high-resolution images of the subsurface) are typically interpreted manually. This is generally an extremely time-consuming analysis step. Therefore, representative 2D sections highlighting distinctive reflection structures are often selected from the 3D data set. Regions showing similar structures are then grouped into so-called radar facies. The results obtained from 2D sections are considered representative of the entire investigated area. Interpretations conducted in this manner are often incomplete and highly dependent on the expertise of the interpreters, making them generally non-reproducible.
A promising alternative or complement to manual interpretation is the use of GPR attributes. Instead of using the recorded data directly, derived quantities characterizing distinctive reflection structures in 3D are applied for interpretation. Using various field and synthetic data sets, this thesis investigates which attributes are particularly suitable for this purpose. Additionally, the study demonstrates how selected attributes can be utilized through specific processing and classification methods to create 3D facies models. The ability to generate attribute-based 3D GPR facies models allows for partially automated and more efficient interpretations in the future. Furthermore, the results obtained in this manner describe the subsurface in a reproducible and more comprehensive manner than what has typically been achievable through manual interpretation methods.
The dynamic landscape of digital transformation entails an impact on industrial-age manufacturing companies that goes beyond product offerings, changing operational paradigms, and requiring an organization-wide metamorphosis. An initiative to address the given challenges is the creation of Digital Innovation Units (DIUs) – departments or distinct legal entities that use new structures and practices to develop digital products, services, and business models and support or drive incumbents’ digital transformation. With more than 300 units in German-speaking countries alone and an increasing number of scientific publications, DIUs have become a widespread phenomenon in both research and practice.
This dissertation examines the evolution process of DIUs in the manufacturing
industry during their first three years of operation, through an extensive longitudinal single-case study and several cross-case syntheses of seven DIUs. Building on the lenses of organizational change and development, time, and socio-technical systems, this research provides insights into the fundamentals, temporal dynamics, socio-technical interactions, and relational dynamics of a DIU’s evolution process. Thus, the dissertation promotes a dynamic understanding of DIUs and adds a two-dimensional perspective to the often one-dimensional view of these units and their interactions with the main organization throughout the startup and growth phases of a DIU.
Furthermore, the dissertation constructs a phase model that depicts the early stages of DIU evolution based on these findings and by incorporating literature from information systems research. As a result, it illustrates the progressive intensification of collaboration between the DIU and the main organization. After being implemented, the DIU sparks initial collaboration and instigates change within (parts of) the main organization. Over time, it adapts to the corporate environment to some extent, responding to changing circumstances in order to contribute to long-term transformation. Temporally, the DIU drives the early phases of cooperation and adaptation in particular, while the main organization triggers the first major evolutionary step and realignment of the DIU.
Overall, the thesis identifies DIUs as malleable organizational structures that are crucial for digital transformation. Moreover, it provides guidance for practitioners on the process of building a new DIU from scratch or optimizing an existing one.
A comprehensive study on seismic hazard and earthquake triggering is crucial for effective mitigation of earthquake risks. The destructive nature of earthquakes motivates researchers to work on forecasting despite the apparent randomness of the earthquake occurrences. Understanding their underlying mechanisms and patterns is vital, given their potential for widespread devastation and loss of life. This thesis combines methodologies, including Coulomb stress calculations and aftershock analysis, to shed light on earthquake complexities, ultimately enhancing seismic hazard assessment.
The Coulomb failure stress (CFS) criterion is widely used to predict the spatial distributions of aftershocks following large earthquakes. However, uncertainties associated with CFS calculations arise from non-unique slip inversions and unknown fault networks, particularly due to the choice of the assumed aftershocks (receiver) mechanisms. Recent studies have proposed alternative stress quantities and deep neural network approaches as superior to CFS with predefined receiver mechanisms. To challenge these propositions, I utilized 289 slip inversions from the SRCMOD database to calculate more realistic CFS values for a layered-half space and variable receiver mechanisms. The analysis also investigates the impact of magnitude cutoff, grid size variation, and aftershock duration on the ranking of stress metrics using receiver operating characteristic (ROC) analysis. Results reveal the performance of stress metrics significantly improves after accounting for receiver variability and for larger aftershocks and shorter time periods, without altering the relative ranking of the different stress metrics.
To corroborate Coulomb stress calculations with the findings of earthquake source studies in more detail, I studied the source properties of the 2005 Kashmir earthquake and its aftershocks, aiming to unravel the seismotectonics of the NW Himalayan syntaxis. I simultaneously relocated the mainshock and its largest aftershocks using phase data, followed by a comprehensive analysis of Coulomb stress changes on the aftershock planes. By computing the Coulomb failure stress changes on the aftershock faults, I found that all large aftershocks lie in regions of positive stress change, indicating triggering by either co-seismic or post-seismic slip on the mainshock fault.
Finally, I investigated the relationship between mainshock-induced stress changes and associated seismicity parameters, in particular those of the frequency-magnitude (Gutenberg-Richter) distribution and the temporal aftershock decay (Omori-Utsu law). For that purpose, I used my global data set of 127 mainshock-aftershock sequences with the calculated Coulomb Stress (ΔCFS) and the alternative receiver-independent stress metrics in the vicinity of the mainshocks and analyzed the aftershocks properties depend on the stress values. Surprisingly, the results show a clear positive correlation between the Gutenberg-Richter b-value and induced stress, contrary to expectations from laboratory experiments. This observation highlights the significance of structural heterogeneity and strength variations in seismicity patterns. Furthermore, the study demonstrates that aftershock productivity increases nonlinearly with stress, while the Omori-Utsu parameters c and p systematically decrease with increasing stress changes. These partly unexpected findings have significant implications for future estimations of aftershock hazard.
The findings in this thesis provides valuable insights into earthquake triggering mechanisms by examining the relationship between stress changes and aftershock occurrence. The results contribute to improved understanding of earthquake behavior and can aid in the development of more accurate probabilistic-seismic hazard forecasts and risk reduction strategies.