Refine
Year of publication
Document Type
- Article (2860)
- Doctoral Thesis (510)
- Postprint (139)
- Other (76)
- Review (52)
- Monograph/Edited Volume (34)
- Preprint (17)
- Conference Proceeding (13)
- Habilitation Thesis (12)
- Master's Thesis (6)
- Part of a Book (3)
- Bachelor Thesis (2)
- Report (2)
- Moving Images (1)
Keywords
- climate change (52)
- Holocene (44)
- erosion (29)
- permafrost (28)
- Himalaya (27)
- remote sensing (26)
- Climate change (23)
- Tibetan Plateau (22)
- Earthquake source observations (21)
- Pollen (21)
Institute
- Institut für Geowissenschaften (3727) (remove)
Terrestrial landscape dynamics are dominated by the production, mobilisation, transfer and deposition of sediment. Numerous chemical elements are carried by sediments, making them a key component for ecological processes, as soil constitution and thus plants and animal ecosystems depends on them, and by extension the human species. They are also essential for climate evolution and regulation as marine sedimentation acts as a carbon sink. However, the processes at the origin of their production, mobilisation, transfer can occur suddenly with high-energy content – such as volcanic eruption, mass wasting or flooding events and wildfires – shaking ecosystems and shaping landforms. Besides, in the last era, the human species has shown its ability to disturb landscape dynamics and change sediments cycles. Thus, there is a need for predictive understanding of the processes involved. This relies on un-derstanding of the mechanisms of key processes and their controls, and knowledge of the state and evolution of the Earth’s surface. Classic approaches to these challenges include empirical observations and numerical modeling of geochemical fluxes and surface processes, as well as the study of terrestrial sedimentary archives to better understand the parameters at stake in landscape dynamics and climate changes and the different actions and retroactions between the production, mobilisation, transfer and deposition of sediments which ultimately shape landscapes. Environmental seismology complements these approaches.
Environmental seismology is the science topic investigating the source functions and propagation properties of seismic vibrations triggered by processes happening at or near the Earth’s surface, below and above it - cryosphere, hydrosphere, atmosphere, human habitat, biosphere, etc., to obtain insight into all these physical processes. Indeed, from mass wasting event to rivers, from wild species to hu-man, all these processes are generating seismic waves. Environmental seismology is a rather recent field, with new branches rapidly expanding and at various stages of scientific progress. This thesis is motivated by the goal of learning more on two major natural process hazards (river bedload transport and mass wasting) as well as on human-generated acoustic hazard, while covering the axis of funda-mental research progression, from data exploration and method and theory development to proof of concept, with the twin aims of developing a better understanding of the operation of these specific processes, and of advancing the methods we have at our disposal to study them.
First, I provide a benchmark for assessment of the reliability of existing seismic bedload model inver-sion to retrieve bedload flux from seismic data. Bedload flux measurements are essential to better understand river dynamics, and this can be achieved with environmental seismology. However, due to a lack of well-constrained validation data, the accuracy of the resulting inversions is unknown. I address this gap in Chapter 2.2, reporting a seismic field experiment, and comparing the results to high-quality independent bedload measurements to constrain a seismic bedload model. The study shows that the quality of bedload flux estimates from seismic data strongly depends on the quality of the input data for the model. Direct measurements of relevant parameters, chiefly seismic ground proper-ties needed for the Green's function and the grain size distribution of the moving bedload, considerably improve the model quality over generic approaches using empirical or theoretical functions. I also pro-vide a numerical tool to facilitate the use of water turbulence and bedload seismic inversion models: seismic models for bedload flux and water turbulence require painstaking work to constrain parame-ters describing the ground properties by active seismic study or analysis of passive seismic data, and the grain size distribution via independent measurements. Reasonable predictions can be achieved by using a Monte Carlo approach to optimize the free parameters with respect to the target parameters. The validation of the tool, in Chapter 2.3, with independent measurements of water depth and bedload flux at a study site on the Eshtemoa river in Israël makes it available for reliable use at other sites. The work reported in this chapter has been published in Lagarde et al. 2021 and Dietze et al. 2019b.
In a second study, reported in Chapter 3, I investigate the formation of a failure plane prior to a rock-slide. A better understanding of the dynamics of the preparation phase is essential to determine the timing, volume and mobilization mechanism of a rock slope failure, and this can be achieved with envi-ronmental seismology. I take advantage of a network of seismic stations close to an instable slope recording cracking signals prior to the slope failure, and use a machine learning technique based on hidden Markov models to isolate these signals from the seismic data, retrieving the cumulative number of cracking events in a period of 20 days prior to a large rockslide and 10 days after. The trajectory of the cumulative number of cracks shifts from a rather linear shape in the two weeks prior to the rock-slide to an S-shaped development in the last 27 h before failure. I interpret this change as a switch from initial distributed cracking to localised damage accumulation in the hours prior to the failure. I develop a simple physical model to explain the temporal evolution of crack activity during the S-shape phase, revealing the importance of an internal parameter, the total crack boundary length as the dominant control parameter of failure plane evolution. This study has been published as Lagarde et al. 2023.
Third, I develop a model converting acoustic signals to seismic signals. A part of the acoustic vibrations generated on the Earth’s surface is converted to seismic signals at the ground interface. Consequently, noise pollution may be translated to slope fatigue and rock micro (or macro) fracturing resulting in a degrading effect on landforms. Moreover, this pollution can have negative impacts, such as physical, physiological as well as psychological effects on animal species. At present, the impact of seismic pol-lution generated by acoustic sources is difficult to evaluate. In Chapter 4, I improve and implement a model converting the acoustic pressure generated by a source in the atmosphere to the corresponding seismic signal for a receiver within the ground. The ground is considered to be a porous elastic medium in which wave behaviour can be approximated by the Biot-Stoll model. The model is extended for use of a temporal pressure pulse as an input, and to produce output on a plan 2D map, where the wind effect on the acoustic to seismic coupling can be reproduced. I invest extensive effort in making the model user-friendly, as the project aims at reaching a large audience, comprising, for example, geo-morphologists, biologists and sociologists. Finally, the model is subjected to synthetic testing as well as a qualitative comparison of the predicted ground particle velocity and the seismic signal of a real heli-copter flight as a source of acoustic input.
These studies advance understanding of the operation of specific natural processes in channels and on hillslopes, and bring us closer to designing functioning early warning systems for mass-wasting and flood events. This thesis also raises questions that have not been considered before, such as the con-tribution of human acoustic pollution to the seismic hum and its impact on the natural environment, or the importance of cracks in the self-development of the failure plane prior to slope. Together, these studies question general assumptions usually made regarding the triggering of mass wasting or the hillslope-channel connectivity. Beyond this, the thesis covers the axis of fundamental research progres-sion, from data exploration and method and theory development to proof of concept, and shows how in the rapidly developing field of environmental seismology, an active awareness of progress can help strengthen and accelerate general advances.
Clusty is a new open source toolbox dedicated to earthquake clustering based on waveforms recorded across a network of seismic stations. Its main application is the study of active faults and the detection and characterization of faults and fault networks. By using a density-based clustering approach, earthquakes pertaining to a common fault can be recognized even over long fault segments, and the first-order geometry and extent of active faults can be inferred. Clusty implements multiple techniques to compute a waveform based network similarity from maximum cross-correlation coefficients at multiple stations. The clustering procedure is designed to be transparent and parameters can be easily tuned. It is supported by a number of analysis visualization tools which help to assess the homogeneity within each cluster and the differences among distinct clusters. The toolbox returns graphical representations of the results. A list of representative events and stacked waveforms facilitate further analyses like moment tensor inversion. Results obtained in various frequency bands can be combined to account for large magnitude ranges. Thanks to the simple configuration, the toolbox is easily adaptable to new data sets and to large magnitude ranges. To show the potential of our new toolbox, we apply Clusty to the aftershock sequence of the M-w 6.9 25 October 2018 Zakynthos (Greece) Earthquake. Thanks to the complex tectonic setting at the western termination of the Hellenic Subduction System where multiple faults and faulting styles operate simultaneously, the Zakynthos data set provides an ideal case-study for our clustering analysis toolbox. Our results support the activation of several faults and provide insight into the geometry of faults or fault segments. We identify two large thrust faulting clusters in the vicinity of the main shock and multiple strike-slip clusters to the east, west and south of these clusters. Despite its location within the largest thrust cluster, the main shock does not show a high waveform similarity to any of the clusters. This is consistent with the results of other studies suggesting a complex failure mechanism for the main shock. We propose the existence of conjugated strike-slip faults in the south of the study area. Our waveform similarity based clustering toolbox is able to reveal distinct event clusters which cannot be discriminated based on locations and/or timing only. Additionally, the clustering results allows distinction between fault and auxiliary planes of focal mechanisms and to associate them to known active faults.
Impact of Late Pleistocene climate variability on paleo-erosion rates in the western Himalaya
(2022)
It has been proposed that at short timescales of 10(2)-10(5) yr, climatic variability can explain variations in sediment flux, but in orogens with pronounced climatic gradients rate changes caused by the oscillating efficiency in rainfall, runoff, and/or sediment transport and deposition are still not well-constrained.
To explore landscape responses under variable climatic forcing, we evaluate time windows of prevailing sediment aggradation and related paleo-erosion rates from the southern flanks of the Dhauladhar Range in the western Himalaya.
We compare past and present Be-10-derived erosion rates of well-dated Late Pleistocene fluvial landforms and modern river sediments and reconstruct the sediment aggradation and incision history based on new luminescence data.
Our results document significant variations in erosion rates ranging from 0.1 to 3.4 mm/yr over the Late Pleistocene.
We find that, during times of weak monsoon intensity, the moderately steep areas (hillslope angles of 27 +/- 13 degrees) erode at lower rates of 0.1-0.4 mm/yr compared to steeper (>40 degrees) crestal regions of the Dhauladhar Range that erode at 0.8-1.3 mm/yr.
In contrast, during several millennia of stronger monsoon intensity, both the moderately steep and high slope areas record higher erosion rates (>1-3.4 mm/yr). Lithological clast-count analysis shows that this increase of erosion is focused in the moderately steep areas, where Lesser Himalayan rocks are exposed.
Our data thus highlight the highly non-linear response of climatic forcing on landscape evolution and suggest complex depositional processes and sedimentary signals in downstream areas. (C) 2021 Elsevier B.V. All rights reserved.
Bastnsite [REE(CO3)F] is the main mineral of REE ore deposits in carbonatites. Synthetic bastnsite-like compounds were precipitated from aqueous solutions by many different methods, but previous attempts to model magmatic crystallization of bastnsite from hydrous calciocarbonatite melts were unsuccessful.
Here we present the first experimental evidence that bastnsite and two other REE carbonates, burbankite, and lukechangite, can crystallize from carbonatite melt in the synthetic system La(CO3)F-CaCO3-Na2CO3 at temperatures between 580 and 850 degrees C and a pressure 100 MPa.
The experiments on starting mixtures of reagent-grade CaCO3, Na2CO3, La-2(CO3)(3), and LaF3 were carried out in cold-seal rapid-quench pressure vessels.
The studied system is an isobaric pseudoternary join of a quinary system where CO2 and fluorides act as independent components. Liquidus phases in the run products are calcite, nyerereite, Na carbonate, bastnsite, burbankite solid solution (Na,Ca)(3)(Ca,La)(3)(CO3)(5), and lukechangite Na3La2(CO3)(4)F. Calcite and bastnite form a eutectic in the boundary join La(CO3)F-CaCO3 at 780 +/- 20 degrees C and 58 wt% La(CO3)F. Phase equilibria in the boundary join La(CO3)F-Na2CO3 are complicated by peritectic reaction between Ca-free end-member of burbankite solid solution petersenite (Pet) and lukechangite (Luk) with liquid (L): Na4La2(CO3)(5) (Pet) + NaF (L) = Na3La2(CO3)(4)F (Luk) + Na2CO3 (Nc).
The right-hand-side assemblage becomes stable below 600 +/- 20 degrees C. In ternary mixtures, bastnsite (Bst), burbankite (Bur), and calcite (Cc) are involved in another peritectic reaction: 2La(CO3)F (Bst) + CaCO3 (Cc) + 2Na(2)CO(3) (L) = Na2CaLa2(CO3)(5) (Bur) + 2NaF (L). Burbankite in equilibrium with calcite replaces bastnsite below 730 +/- 20 degrees C. Stable solidus assemblages in the pseudoternary system are: basnsite-burbankite-fluorite-calcite, basnasite-burbankite-fluorite-lukechangite, bastnsite-burbankite-lukechangite, burbankite-lukechangite-nyerereitecalcite, and burbankite-lukechangite-nyerereite-natrite. Addition of 10 wt% Ca-3(PO4)(2) to one of the ternary mixtures resulted in massive crystallization of La-bearing apatite and monazite and complete disappearance of bastnsite and burbankite.
Our results confirm that REE-bearing phosphates are much more stable than carbonates and fluorocarbonates.
Therefore, primary crystallization of the latter from common carbonatite magmas is unlikely. Possible exceptions are carbonatites at Mountain Pass that are characterized by very low P2O5 concentrations (usually at or below 0.5 wt%) and extremely high REE contents in the order of a few weight percent or more. In other carbonatites, bastnsite and burbankite probably crystallized from highly concentrated alkaline carbonate-chloride brines that were found in melt inclusions and are thought to be responsible for widespread fenitization around carbonatite bodies.
The chaotic spatio-temporal electrical activity during life-threatening cardiac arrhythmias like ventricular fibrillation is governed by the dynamics of vortex-like spiral or scroll waves. The organizing centers of these waves are called wave tips (2D) or filaments (3D) and they play a key role in understanding and controlling the complex and chaotic electrical dynamics. Therefore, in many experimental and numerical setups it is required to detect the tips of the observed spiral waves. Most of the currently used methods significantly suffer from the influence of noise and are often adjusted to a specific situation (e.g. a specific numerical cardiac cell model). In this study, we use a specific type of deep neural networks (UNet), for detecting spiral wave tips and show that this approach is robust against the influence of intermediate noise levels. Furthermore, we demonstrate that if the UNet is trained with a pool of numerical cell models, spiral wave tips in unknown cell models can also be detected reliably, suggesting that the UNet can in some sense learn the concept of spiral wave tips in a general way, and thus could also be used in experimental situations in the future (ex-vivo, cell-culture or optogenetic experiments).
The European Alps are amongst the regions with highest glacier mass loss rates over the last decades. Under the threat of ongoing climate change, the ability to predict glacier mass balance changes for water and risk management purposes has become imperative. This raises an urgent need for reliable glacier models. The European Alps do not only host glaciers, but also numerous caves containing carbonate formations, called speleothems. Previous studies have shown that those speleothems also grew during times when the cave was covered by a warm-based glacier. In this thesis, I utilise speleothems from the European Alps as archives of local, environmental conditions related to mountain glacier evolution.
Previous studies have shown that speleothem isotope data from the Alps can be strongly affected by in-cave processes. Therefore, part of this thesis focusses on developing an isotope evolution model, which successfully reproduces differences between contemporaneous growing speleothems. The model is used to propose correction approaches for prior calcite precipitation effects on speleothem oxygen isotopes (δ18O). Applications on speleothem records from caves outside of the Alps demonstrate that corrected δ18O agrees better with other records and climate model simulations.
Existing speleothem growth histories and carbon isotope (δ13C) records from Alpine caves located at different elevations are used to infer soil vs. glacier cover and the thermal regime of the glacier over the last glacial cycle. The compatibility with glacier evolution models is statistically assessed. A general agreement between speleothem δ13C-derived information on soil vs. glacier presence and modelled glacier coverage is found. However, glacier retreat during Marine Isotope Stage (MIS) 3 seems to be underestimated by the model. Furthermore, speleothem data provides evidence of surface temperature above the freezing point which is, however, not fully reproduced by the simulations.
History of glacier cover and their thermal regime is explored for the high-elevation cave system Melchsee-Frutt in the Swiss Alps. Based on new (MIS 9b – MIS 7b, MIS 2) and available speleothem δ13C (MIS 7a – 5d) data, warm-based glacier cover is inferred for MIS 8, 7d, 6, and 2. Also a short period of cold-based ice coverage is found for early MIS 6. In a detailed multi-proxy analysis (δ18O, δ13C, Mg/Ca and Sr/Ca), millennial-scale changes in the glacier-related source of the water infiltrating in the karst during MIS 8 and 7d are found and linked to Northern Hemisphere climate variability.
While speleothem records from high-elevation cave sites in the Alps exhibit huge potential for glacier reconstruction, several limitations remain, which are discussed throughout this thesis. Ultimately, recommendations are given to further leverage subglacial speleothems as an archive of glacier dynamics.
From June to August 2021, we deployed a dense seismic nodal network across the Hengill geothermal area in southwest Iceland to image and characterize faults and high-temperature zones at high resolution.
The nodal network comprised 498 geophone nodes spread across the northern Nesjavellir and southern Hverahlio geothermal fields and was complemented by an existing permanent and temporary backbone seismic network of a total of 44 short-period and broadband stations.
In addition, we recorded distributed acoustic sensing data along two fiber optic telecommunication cables near the Nesjavellir geothermal power plant with commercial interrogators.
During the time of deployment, a vibroseis survey took place around the Nesjavellir power plant.
Here, we describe the network and the recorded datasets.
Furthermore, we showsome initial results that indicate a high data quality and highlight the potential of the seismic records for various follow up studies, such as high-resolution event location to delineate faults and body- and surface-wave tomographies to image the subsurface velocity structure in great detail.
Lakes act as important sinks for inorganic and organic sediment components. However, investigations of sedimentary carbon budgets within glacial lakes are currently absent from Arctic Siberia. The aim of this paper is to provide the first reconstruction of accumulation rates, sediment and carbon budgets from a lacustrine sediment core from Lake Rauchuagytgyn, Chukotka (Arctic Siberia). We combined multiple sediment biogeochemical and sedimentological parameters from a radiocarbon-dated 6.5m sediment core with lake basin hydroacoustic data to derive sediment stratigraphy, sediment volumes and infill budgets. Our results distinguished three principal sediment and carbon accumulation regimes that could be identified across all measured environmental proxies including early Marine Isotope Stage 2 (MIS2) (ca. 29-23.4 ka cal BP), mid-MIS2-early MIS1 (ca. 23.4-11.69 ka cal BP) and the Holocene (ca. 11.69-present). Estimated organic carbon accumulation rates (OCARs) were higher within Holocene sediments (average 3.53 gOCm(-2) a(-1)) than Pleistocene sediments (average 1.08 gOCm(-2) a(-1)) and are similar to those calculated for boreal lakes from Quebec and Finland and Lake Baikal but significantly lower than Siberian thermokarst lakes and Alberta glacial lakes. Using a bootstrapping approach, we estimated the total organic carbon pool to be 0.26 +/- 0.02 Mt and a total sediment pool of 25.7 +/- 1.71 Mt within a hydroacoustically derived sediment volume of ca. 32 990 557m(3). The total organic carbon pool is substantially smaller than Alaskan yedoma, thermokarst lake sediments and Alberta glacial lakes but shares similarities with Finnish boreal lakes. Temporal variability in sediment and carbon accumulation dynamics at Lake Rauchuagytgyn is controlled predominantly by palaeoclimate variation that regulates lake ice-cover dynamics and catchment glacial, fluvial and permafrost processes through time. These processes, in turn, affect catchment and within-lake primary productivity as well as catchment soil development. Spatial differences compared to other lake systems at a trans-regional scale likely relate to the high-latitude, mountainous location of Lake Rauchuagytgyn.
Martian atmospheric spectral end-members retrieval from ExoMars Thermal Infrared (TIRVIM) data
(2022)
Key knowledge about planetary composition can be recovered from the study of thermal infrared spectral range datasets.
This range has a huge diagnostic potential because it contains diagnostic absorptions from a planetary surface and atmosphere. The main goal of this study is to process and interpret the dataset from the Thermal Infrared channel (TIRVIM) which is part of the Atmospheric Chemistry Suite of the ExoMars2016 Trace Gas Orbiter mission to find and characterize dust and water ice clouds in the atmosphere.
The method employed here is based on the application of principal component analysis and target transformation techniques to extract the independent variable components present in the analyzed dataset. Spectral shapes of both atmospheric dust and water ice aerosols have been recovered from the analysis of TIRVIM data.
The comparison between our results with those previously obtained on Thermal Emission Spectrometer (TES) data and with previous analysis on TIRVIM data, validates the methodology here applied, showing that it allows to correctly recover the atmospheric spectral endmembers present in the TIRVIM data.
Moreover, comparison with atmospheric retrievals on PFS, TES and IRIS data, allowed us to assess the temporal stability and homogeneity of dust and water ice components in the Martian atmosphere over a time period of almost 50 years.
The early exhumation history of the Tauern Window in the European Eastern Alps and its surface expression is poorly dated and quantified, partly because thermochronological and provenance information are sparse from the Upper Austrian Northern Alpine Foreland Basin. For the first time, we combine a single-grain double-dating approach (Apatite Fission Track and U-Pb dating) with trace-element geochemistry analysis on the same apatites to reconstruct the provenance and exhumation history of the late Oligocene/early Miocene Eastern Alps. The results from 22 samples from the Chattian to Burdigalian sedimentary infill of the Upper Austrian Northern Alpine Foreland Basin were integrated with a 3D seismic-reflection data set and published stratigraphic reports. Our highly discriminative data set indicates an increasing proportion of apatites (from 6% to 23%) with Sr/Y values <0.1 up-section and an increasing amount of apatites (from 24% to 38%) containing >1,000 ppm light rare-earth elements from Chattian to Burdigalian time. The number of U-Pb ages with acceptable uncertainties increases from 40% to 59% up-section, with mostly late Variscan/Permian ages, while an increasing number of grains (10%-27%) have Eocene or younger apatite fission track cooling ages. The changes in the apatite trace-element geochemistry and U-Pb data mirror increased sediment input from an >= upper amphibolite-facies metamorphic source of late Variscan/Permian age - probably the otztal-Bundschuh nappe system - accompanied by increasing exhumation rates indicated by decreasing apatite fission track lag times. We attribute these changes to the surface response to upright folding and doming in the Penninic units of the future Tauern Window starting at 29-27 Ma. This early period of exhumation (0.3-0.6 mm/a) is triggered by early Adriatic indentation along the Giudicarie Fault System.