Refine
Year of publication
- 2015 (24) (remove)
Document Type
- Doctoral Thesis (24) (remove)
Language
- English (24) (remove)
Is part of the Bibliography
- yes (24)
Keywords
- regional climate model (2)
- Aerosole (1)
- Arktis (1)
- Asteroiden (1)
- Astronomie (1)
- Astrophysik (1)
- Brillouin Streuung (1)
- Brillouin scattering (1)
- Budgetstudie (1)
- Budgetstudien (1)
Institute
- Institut für Physik und Astronomie (24) (remove)
In this thesis we utilize resolved stellar populations to improve our understanding of galaxy formation and evolution. In the first part we improve a method for metallicity determination of faint old stellar systems, in the second and third part we analyze the individual history of six nearby disk galaxies outside the Local Group.
A New Calibration of the Color Metallicity Relation of Red Giants for HST data:
It is well known, that the color distribution of stars on the the Red Giant Branch (RGB) can be used to determine metallicities of old stellar populations that have only shallow photometry. Based on the largest sample of globular clusters ever used for such studies, we quantify the relation between metallicity and color in the widely used HST ACS filters F606W and F814W.
We use a sample of globular clusters from the ACS Globular Cluster Survey and measure their RGB color at given absolute magnitudes to derive the color-metallicity relation. We find a clear relation between metallicity and RGB color; we investigate the scatter and the uncertainties in this relation and show its limitations. A comparison with isochrones shows reasonably good agreement with BaSTI models, a small offset to Dartmouth models, and a larger offset to Padua models.
Even for the best globular cluster data available, the metallicity of a simple stellar population can be determined from the RGB alone only with an accuracy of 0.3 dex for [M/H]<-1, and 0.15 dex for [M/H]>-1. For mixed populations, as they are observed in external galaxies, the uncertainties will be even larger due to uncertainties in extinction, age, etc. Therefore caution is necessary when interpreting photometric metallicities.
The Structural History of Nearby Low Mass Disk Galaxies:
We study the individual evolution histories of three nearby, low-mass, edge-on galaxies (IC5052, NGC4244, NGC5023).
Using the color magnitude diagrams of resolved stellar populations, we construct star count density maps for populations of different ages and analyze the change of structural parameters with stellar age within each galaxy.
The three galaxies show low vertical heating rates, which are much lower than the heating rate of the Milky Way. This indicates that heating agents, as giant molecular clouds and spiral structure are weak in low mass galaxies.
We do not detect a separate thick disk in any of the three galaxies, even though our observations cover a larger range in equivalent surface brightness than any integrated light study. While scaleheights increase with age, each population can be well described by a single disk. Only two of the galaxies contain a very weak additional component, which we identify as the faint halo. The mass of these faint halos is less than 1% of the mass of the disk.
All populations in the three galaxies exhibit no or only little flaring. While this finding is consistent with previous integrated light studies, it poses strong constraints on galaxy formation models, because most theoretical simulations often find strong flaring due to interactions or radial migration.
Furthermore, we find breaks in the radial profiles of all three galaxies. The radii of these breaks are independent of age, and the break strength is decreasing with age in two of the galaxies (NGC4244 and NGC5023). This is consistent with break formation models, that combine a star formation cutoff with radial migration. The differing behavior of IC5052 can be explained by a recent interaction or minor merger.
The Structural History of Massive Disk Galaxies:
We extend the structural analysis of stellar populations with distinct ages to three massive galaxies, NGC891, NGC4565 and NGC7814. While confusion effects due to the high stellar number densities in their central region, and the prominent dust lanes inhibit an detailed analysis of the radial profiles, we can study their vertical structure.
These massive galaxies also have a slower heating than the Milky Way, comparable to the low mass galaxies. This can be traced back to their already thick young populations and thick layers of their interstellar medium.
We do not find a clear separate thick disk in any of these three galaxies; all populations can be described by a single disk plus a S\'ersic bulge/halo component. In contrast to the low mass galaxies, we cannot rule out the presence of thick disks in the massive galaxies, because of the strong influence of the halo, that might hide the possible contribution of the thick disk to the vertical star count profiles. However, the faintness of the possible thick disks still points to problems in the earlier ubiquitous findings of thick disks in external galaxies.
Most of the baryonic matter in the Universe resides in a diffuse gaseous phase in-between galaxies consisting mostly of hydrogen and helium. This intergalactic medium (IGM) is distributed in large-scale filaments as part of the overall cosmic web. The luminous extragalactic objects that we can observe today, such as galaxies and quasars, are surrounded by the IGM in the most dense regions within the cosmic web. The radiation of these objects contributes to the so-called ultraviolet background (UVB) which keeps the IGM highly ionized ever since the epoch of reionization.
Measuring the amount of absorption due to intergalactic neutral hydrogen (HI) against extragalactic background sources is a very useful tool to constrain the energy input of ionizing sources into the IGM. Observations suggest that the HI Lyman-alpha effective optical depth, τ_eff, decreases with decreasing redshift, which is primarily due to the expansion of the Universe. However, some studies find a smaller value of the effective optical depth than expected at the specific redshift z~3.2, possibly related to the complete reionization of helium in the IGM and a hardening of the UVB. The detection and possible cause of a decrease in τ_eff at z~3.2 is controversially debated in the literature and the observed features need further explanation.
To better understand the properties of the mean absorption at high redshift and to provide an answer for whether the detection of a τ_eff feature is real we study 13 high-resolution, high signal-to-noise ratio quasar spectra observed with the Ultraviolet and Visual Echelle Spectrograph (UVES) at the Very Large Telescope (VLT). The redshift evolution of the effective optical depth, τ_eff(z), is measured in the redshift range 2.7≤z≤3.6. The influence of metal absorption features is removed by performing a comprehensive absorption-line-fitting procedure.
In the first part of the thesis, a line-parameter analysis of the column density, N, and Doppler parameter, b, of ≈7500 individually fitted absorption lines is performed. The results are in good agreement with findings from previous surveys.
The second (main) part of this thesis deals with the analysis of the redshift evolution of the effective optical depth. The τ_eff measurements vary around the empirical power law τ_eff(z)~(1+z)^(γ+1) with γ=2.09±0.52. The same analysis as for the observed spectra is performed on synthetic absorption spectra. From a comparison between observed and synthetic spectral data it can be inferred that the uncertainties of the τ_eff values are likely underestimated and that the scatter is probably caused by high-column-density absorbers with column densities in the range 15≤logN≤17. In the real Universe, such absorbers are rarely observed, however. Hence, the difference in τ_eff from different observational data sets and absorption studies is most likely caused by cosmic variance. If, alternatively, the disagreement between such data is a result of an too optimistic estimate of the (systematic) errors, it is also possible that all τ_eff measurements agree with a smooth evolution within the investigated redshift range. To explore in detail the different analysis techniques of previous studies an extensive literature comparison to the results of this work is presented in this thesis.
Although a final explanation for the occurrence of the τ_eff deviation in different studies at z~3.2 cannot be given here, our study, which represents the most detailed line-fitting analysis of its kind performed at the investigated redshifts so far, represents another important benchmark for the characterization of the HI Ly-alpha effective optical depth at high redshift and its indicated unusual behavior at z~3.2.
Synchronization of large ensembles of oscillators is an omnipresent phenomenon observed in different fields of science like physics, engineering, life sciences, etc. The most simple setup is that of globally coupled phase oscillators, where all the oscillators contribute to a global field which acts on all oscillators. This formulation of the problem was pioneered by Winfree and Kuramoto. Such a setup gives a possibility for the analysis of these systems in terms of global variables. In this work we describe nontrivial collective dynamics in oscillator populations coupled via mean fields in terms of global variables. We consider problems which cannot be directly reduced to standard Kuramoto and Winfree models.
In the first part of the thesis we adopt a method introduced by Watanabe and Strogatz. The main idea is that the system of identical oscillators of particular type can be described by a low-dimensional system of global equations. This approach enables us to perform a complete analytical analysis for a special but vast set of initial conditions. Furthermore, we show how the approach can be expanded for some nonidentical systems. We apply the Watanabe-Strogatz approach to arrays of Josephson junctions and systems of identical phase oscillators with leader-type coupling.
In the next parts of the thesis we consider the self-consistent mean-field theory method that can be applied to general nonidentical globally coupled systems of oscillators both with or without noise. For considered systems a regime, where the global field rotates uniformly, is the most important one. With the help of this approach such solutions of the self-consistency equation for an arbitrary distribution of frequencies and coupling parameters can be found analytically in the parametric form, both for noise-free and noisy cases.
We apply this method to deterministic Kuramoto-type model with generic coupling and an ensemble of spatially distributed oscillators with leader-type coupling. Furthermore, with the proposed self-consistent approach we fully characterize rotating wave solutions of noisy Kuramoto-type model with generic coupling and an ensemble of noisy oscillators with bi-harmonic coupling.
Whenever possible, a complete analysis of global dynamics is performed and compared with direct numerical simulations of large populations.
Spots on stellar surfaces are thought to be stellar analogues of sunspots. Thus, starspots are direct manifestations of strong magnetic fields. Their decay rate is directly related to the magnetic diffusivity, which itself is a key quantity for the deduction of an activity cycle length. So far, no single starspot decay has been observed, and thus no stellar activity cycle was inferred from its corresponding turbulent diffusivity.
We investigate the evolution of starspots on the rapidly-rotating K0 giant XX Triangulum. Continuous high-resolution and phase-resolved spectroscopy was obtained with the robotic 1.2-m STELLA telescope on Tenerife over a timespan of six years. With our line-profile inversion code iMap we reconstruct a total of 36 consecutive Doppler maps. To quantify starspot area decay and growth, we match the observed images with simplified spot models based on a Monte-Carlo approach.
It is shown that the surface of XX Tri is covered with large high-latitude and even polar spots and with occasional small equatorial spots. Just over the course of six years, we see a systematically changing spot distribution with various time scales and morphology such as spot fragmentation and spot merging as well as spot decay and formation.
For the first time, a starspot decay rate on another star than the Sun is determined. From our spot-decay analysis we determine an average linear decay rate of D = -0.067±0.006 Gm^2/day. From this decay rate, we infer a turbulent diffusivity of η_τ = (6.3±0.5) x 10^14 cm^2/s and consequently predict an activity cycle of 26±6 years. The obtained cycle length matches very well with photometric observations.
Our time-series of Doppler maps further enables to investigate the differential rotation of XX Tri. We therefore applied a cross-correlation analysis. We detect a weak solar-like differential rotation with a surface shear of α = 0.016±0.003. This value agrees with similar studies of other RS CVn stars.
Furthermore, we found evidence for active longitudes and flip-flops. Whereas the more active longitude is located in phase towards the (unseen) companion star, the weaker active longitude is located at the opposite stellar hemisphere. From their periodic appearance, we infer a flip-flop cycle of ~2 years. Both activity phenomena are common on late-type binary stars.
Last but not least we redetermine several astrophysical properties of XX Tri and its binary system, as large datasets of photometric and spectroscopic observations are available since its last determination in 1999. Additionally, we compare the rotational spot-modulation from photometric and spectroscopic studies.
In this thesis, I study ultrafast dynamics in perovskite oxides using time resolved broadband spectroscopy. I focus on the observation of coherent phonon propagation by time resolved Brillouin scattering: following the excition of metal transducer films with a femtosecond infrared pump pulse, coherent phonon dynamics in the GHz frequency range are triggered. Their propagation is monitored using a delayed white light probe pulse. The technique is illustrated on various thin films and multilayered samples. I apply the technique to investigate the linear and nonlinear acoustic response in bulk SrTiO_3, which displays a ferroelastic phase transition from a cubic to a tetragonal structural phase at T_a=105 K. In the linear regime, I observe a coupling of the observed acoustic phonon mode to the softening optic modes describing the phase transition. In the nonlinear regime, I find a giant slowing down of the sound velocity in the low temperature phase that is only observable for a strain amplitude exceeding the tetragonality of the material. It is attributed to a coupling of the high frequency phonons to ferroelastic domain walls in the material. I propose a new mechanism for the coupling of strain waves to the domain walls that is only effective for high amplitude strain. A detailed study of the phonon attenuation across a wide temperature range shows that the phonon attenuation at low temperatures is influenced by the domain configuration, which is determined by interface strain. Preliminary measurements on magnetic-ferroelectric multilayers reveal that the excitation fluence needs to be carefully controlled when dynamics at phase transitions are studied.
The lives of more than 1/6 th of the world population is directly affected by the caprices of the South Asian summer monsoon rainfall. India receives around 78 % of the annual precipitation during the June-September months, the summer monsoon season of South Asia. But, the monsoon circulation is not consistent throughout the entire summer season. Episodes of heavy rainfall (active periods) and low rainfall (break periods) are inherent to the intraseasonal variability of the South Asian summer monsoon. Extended breaks or long-lasting dryness can result in droughts and hence trigger crop failures and in turn famines. Furthermore, India's electricity generation from renewable sources (wind and hydro-power), which is increasingly important in order to satisfy the rapidly rising demand for energy, is highly reliant on the prevailing meteorology. The major drought years 2002 and 2009 for the Indian summer monsoon during the last decades, which are results of the occurrence of multiple extended breaks, emphasise exemplary that the understanding of the monsoon system and its intraseasonal variation is of greatest importance. Although, numerous studies based on observations, reanalysis data and global model simulations have been carried out with the focus on monsoon active and break phases over India, the understanding of the monsoon intraseasonal variability is only in the infancy stage. Regional climate models could benefit the comprehension of monsoon breaks by its resolution advantage.
This study investigates moist dynamical processes that initiate and maintain breaks during the South Asian summer monsoon using the atmospheric regional climate model HIRHAM5 at a horizontal resolution of 25 km forced by the ECMWF ERA Interim reanalysis for the period 1979-2012. By calculating moisture and moist static energy budgets the various competing mechanisms leading to extended breaks are quantitatively estimated. Advection of dry air from the deserts of western Asia towards central India is the dominant moist dynamical process in initiating extended break conditions over South Asia. Once initiated, the extended breaks are maintained due to many competing mechanisms: (i) the anomalous easterlies at the southern flank of this anticyclonic anomaly weaken the low-level cross-equatorial jet and thus the moisture transport into the monsoon region, (ii) differential radiative heating over the continental and the oceanic tropical convergence zone induces a local Hadley circulation with anomalous rising over the equatorial Indian Ocean and descent over central India, and (iii) a cyclonic response to positive rainfall anomalies over the near-equatorial Indian Ocean amplifies the anomalous easterlies over India and hence contributes to the low-level divergence over central India.
A sensitivity experiment that mimics a scenario of higher atmospheric aerosol concentrations over South Asia addresses a current issue of large uncertainty: the role aerosols play in suppressing monsoon rainfall and hence in triggering breaks. To study the indirect aerosol effects the cloud droplet number concentration was increased to imitate the aerosol's function as cloud condensation nuclei. The sensitivity experiment with altered microphysical cloud properties shows a reduction in the summer monsoon precipitation together with a weakening of the South Asian summer monsoon. Several physical mechanisms are proposed to be responsible for the suppressed monsoon rainfall: (i) according to the first indirect radiative forcing the increase in the number of cloud droplets causes an increase in the cloud reflectivity of solar radiation, leading to a climate cooling over India which in turn reduces the hydrological cycle, (ii) a stabilisation of the troposphere induced by a differential cooling between the surface and the upper troposphere over central India inhibits the growth of deep convective rain clouds, (iii) an increase of the amount of low and mid-level clouds together with a decrease in high-level cloud amount amplify the surface cooling and hence the atmospheric stability, and (iv) dynamical changes of the monsoon manifested as a anomalous anticyclonic circulation over India reduce the moisture transport into the monsoon region. The study suggests that the changes in the total precipitation, which are dominated by changes in the convective precipitation, mainly result from the indirect radiative forcing. Suppression of rainfall due to the direct microphysical effect is found to be negligible over India. Break statistics of the polluted cloud scenario indicate an increase in the occurrence of short breaks (3 days), while the frequency of extended breaks (> 7 days) is clearly not affected. This disproves the hypothesis that more and smaller cloud droplets, caused by a high load of atmospheric aerosols trigger long drought conditions over central India.
Semi-empirical sea-level models (SEMs) exploit physically motivated empirical relationships between global sea level and certain drivers, in the following global mean temperature. This model class evolved as a supplement to process-based models (Rahmstorf (2007)) which were unable to fully represent all relevant processes. They thus failed to capture past sea-level change (Rahmstorf et al. (2012)) and were thought likely to underestimate future sea-level rise. Semi-empirical models were found to be a fast and useful tool for exploring the uncertainties in future sea-level rise, consistently giving significantly higher projections than process-based models.
In the following different aspects of semi-empirical sea-level modelling have been studied. Models were first validated using various data sets of global sea level and temperature. SEMs were then used on the glacier contribution to sea level, and to infer past global temperature from sea-level data via inverse modelling. Periods studied encompass the instrumental period, covered by tide gauges (starting 1700 CE (Common Era) in Amsterdam) and satellites (first launched in 1992 CE), the era from 1000 BCE (before CE) to present, and the full length of the Holocene (using proxy data). Accordingly different data, model formulations and implementations have been used. It could be shown in Bittermann et al. (2013) that SEMs correctly predict 20th century sea-level when calibrated with data until 1900 CE. SEMs also turned out to give better predictions than the Intergovernmental Panel on Climate Change (IPCC) 4th assessment report (AR4, IPCC (2007)) models, for the period from 1961–2003 CE.
With the first multi-proxy reconstruction of global sea-level as input, estimate of the human-induced component of modern sea-level change and projections of future sea-level rise were calculated (Kopp et al. (2016)). It turned out with 90% confidence that more than 40 % of the observed 20th century sea-level rise is indeed anthropogenic. With the new semi-empirical and IPCC (2013) 5th assessment report (AR5) projections the gap between SEM and process-based model projections closes, giving higher credibility to both. Combining all scenarios, from strong mitigation to business as usual, a global sea-level rise of 28–131 cm relative to 2000 CE, is projected with 90% confidence. The decision for a low carbon pathway could halve the expected global sea-level rise by 2100 CE.
Present day temperature and thus sea level are driven by the globally acting greenhouse-gas forcing. Unlike that, the Milankovich forcing, acting on Holocene timescales, results mainly in a northern-hemisphere temperature change. Therefore a semi-empirical model can be driven with northernhemisphere temperatures, which makes it possible to model the main subcomponent of sea-level change over this period. It showed that an additional positive constant rate of the order of the estimated Antarctic sea-level contribution is then required to explain the sea-level evolution over the Holocene. Thus the global sea level, following the climatic optimum, can be interpreted as the sum of a temperature induced sea-level drop and a positive long-term contribution, likely an ongoing response to deglaciation coming from Antarctica.
The non-linear behaviour of the atmospheric dynamics is not well understood and makes the evaluation and usage of regional climate models (RCMs) difficult. Due to these non-linearities, chaos and internal variability (IV) within the RCMs are induced, leading to a sensitivity of RCMs to their initial conditions (IC). The IV is the ability of RCMs to realise different solutions of simulations that differ in their IC, but have the same lower and lateral boundary conditions (LBC), hence can be defined as the across-member spread between the ensemble members.
For the investigation of the IV and the dynamical and diabatic contributions generating the IV four ensembles of RCM simulations are performed with the atmospheric regional model HIRHAM5. The integration area is the Arctic and each ensemble consists of 20 members. The ensembles cover the time period from July to September for the years 2006, 2007, 2009 and 2012. The ensemble members have the same LBC and differ in their IC only. The different IC are arranged by an initialisation time that shifts successively by six hours. Within each ensemble the first simulation starts on 1st July at 00 UTC and the last simulation starts on 5th July at 18 UTC and each simulation runs until 30th September. The analysed time period ranges from 6th July to 30th September, the time period that is covered by all ensemble members. The model runs without any nudging to allow a free development of each simulation to get the full internal variability within the HIRHAM5.
As a measure of the model generated IV, the across-member standard deviation and the across-member variance is used and the dynamical and diabatic processes influencing the IV are estimated by applying a diagnostic budget study for the IV tendency of the potential temperature developed by Nikiema and Laprise [2010] and Nikiema and Laprise [2011]. The diagnostic budget study is based on the first law of thermodynamics for potential temperature and the mass-continuity equation. The resulting budget equation reveals seven contributions to the potential temperature IV tendency.
As a first study, this work analyses the IV within the HIRHAM5. Therefore, atmospheric circulation parameters and the potential temperature for all four ensemble years are investigated. Similar to previous studies, the IV fluctuates strongly in time. Further, due to the fact that all ensemble members are forced with the same LBC, the IV depends on the vertical level within the troposphere, with high values in the lower troposphere and at 500 hPa and low values in the upper troposphere and at the surface. By the same reason, the spatial distribution shows low values of IV at the boundaries of the model domain.
The diagnostic budget study for the IV tendency of potential temperature reveals that the seven contributions fluctuate in time like the IV. However, the individual terms reach different absolute magnitudes. The budget study identifies the horizontal and vertical ‘baroclinic’ terms as the main contributors to the IV tendency, with the horizontal ‘baroclinic’ term producing and the vertical ‘baroclinic’ term reducing the IV. The other terms fluctuate around zero, because they are small in general or are balanced due to the domain average.
The comparison of the results obtained for the four different ensembles (summers 2006, 2007, 2009 and 2012) reveals that on average the findings for each ensemble are quite similar concerning the magnitude and the general pattern of IV and its contributions. However, near the surface a weaker IV is produced with decreasing sea ice extent. This is caused by a smaller impact of the horizontal 'baroclinic' term over some regions and by the changing diabatic processes, particularly a more intense reducing tendency of the IV due to condensative heating. However, it has to be emphasised that the behaviour of the IV and its dynamical and diabatic contributions are influenced mainly by complex atmospheric feedbacks and large-scale processes and not by the sea ice distribution.
Additionally, a comparison with a second RCM covering the Arctic and using the same LBCs and IC is performed. For both models very similar results concerning the IV and its dynamical and diabatic contributions are found. Hence, this investigation leads to the conclusion that the IV is a natural phenomenon and is independent from the applied RCM.
This thesis investigates the application of polyelectrolyte multilayers in plasmonics and picosecond acoustics. The observed samples were fabricated by the spin-assisted layer-by-layer deposition technique that allowed a precise tuning of layer thickness in the range of few nanometers.
The first field of interest deals with the interaction of light-induced localized surface plasmons (LSP) of rod-shaped gold nanoparticles with the particles' environment. The environment consists of an air phase and a phase of polyelectrolytes, whose ratio affects the spectral position of the LSP resonance.
Measured UV-VIS spectra showed the shift of the LSP absorption peak as a function of the cover layer thickness of the particles. The data are modeled using an average dielectric function instead of the dielectric functions of air and polyelectrolytes. In addition using a measured dielectric function of the gold nanoparticles, the position of the LSP absorption peak could be simulated with good agreement to the data.
The analytic model helps to understand the optical properties of metal nanoparticles in an inhomogeneous environment.
The second part of this work discusses the applicability of PAzo/PAH and dye-doped PSS/PAH polyelectrolyte multilayers as transducers to generate hypersound pulses. The generated strain pulses were detected by time-domain Brillouin scattering (TDBS) using a pump-probe laser setup. Transducer layers made of polyelectrolytes were compared qualitatively to common aluminum transducers in terms of measured TDBS signal amplitude, degradation due to laser excitation, and sample preparation.
The measurements proved that fast and easy prepared polyelectrolyte transducers provided stronger TDBS signals than the aluminum transducer. AFM topography measurements showed a degradation of the polyelectrolyte structures, especially for the PAzo/PAH sample.
To quantify the induced strain, optical barriers were introduced to separate the transducer material from the medium of the hypersound propagation. Difficulties in the sample preparation prohibited a reliable quantification. But the experiments showed that a coating with transparent polyelectrolytes increases the efficiency of aluminum transducers and modifies the excited phonon distribution.
The adoption of polyelectrolytes to the scientific field of picosecond acoustics enables a cheap and fast fabrication of transducer layers on most surfaces. In contrast to aluminum layers the polyelectrolytes are transparent over a wide spectral range. Thus, the strain modulation can be probed from surface and back.
The origin of cosmic rays was the subject of several studies for over a century. The investigations done within this dissertation are one small step to shed some more light on this mystery.
Locating the sources of cosmic rays is not trivial due to the interstellar magnetic field. However, the Hillas criterion allows us to arrive at the conclusion that supernova remnants are our main suspect for the origin of galactic cosmic rays. The mechanism by which they are accelerating particles is found within the field of shock physics as diffusive shock acceleration. To allow particles to enter this process also known as Fermi acceleration pre-acceleration processes like shock surfing acceleration and shock drift acceleration are necessary. Investigating the processes happening in the plasma shocks of supernova remnants is possible by utilising a simplified model which can be simulated on a computer using Particle-in-Cell simulations.
We developed a new and clean setup to simulate the formation of a double shock, i.e., consisting of a forward and a reverse shock and a contact discontinuity, by the collision of two counter-streaming plasmas, in which a magnetic field can be woven into. In a previous work, we investigated the processes at unmagnetised and at magnetised parallel shocks, whereas in the current work, we move our investigation on to magnetised perpendicular shocks.
Due to a much stronger confinement of the particles to the collision region the perpendicular shock develops much faster than the parallel shock. On the other hand, this leads to much weaker turbulence. We are able to find indications for shock surfing acceleration and shock drift acceleration happening at the two shocks leading to populations of pre-accelerated particles that are suitable as a seed population to be injected into further diffusive shock acceleration to be accelerated to even higher energies. We observe the development of filamentary structures in the shock ramp of the forward shock, but not at the reverse shock. This leads to the conclusion that the development of such structures in the shock ramp of quasi-perpendicular collisionless shocks might not necessarily be determined by the existence of a critical sonic Mach number but by a critical shock speed.
The results of the investigations done within this dissertation might be useful for further studies of oblique shocks and for studies using hybrid or magnetohydrodynamic simulations. Together with more sophisticated observational methods, these studies will help to bring us closer to an answer as to how particles can be accelerated in supernova remnants and eventually become cosmic rays that can be detected on Earth.
Optical frequency combs (OFC) constitute an array of phase-correlated equidistant spectral lines with nearly equal intensities over a broad spectral range. The adaptations of combs generated in mode-locked lasers proved to be highly efficient for the calibration of high-resolution (resolving power > 50000) astronomical spectrographs. The observation of different galaxy structures or the studies of the Milky Way are done using instruments in the low- and medium resolution range. To such instruments belong, for instance, the Multi Unit Spectroscopic Explorer (MUSE) being developed for the Very Large Telescope (VLT) of the European Southern Observatory (ESO) and the 4-metre Multi-Object Spectroscopic Telescope (4MOST) being in development for the ESO VISTA 4.1 m Telescope. The existing adaptations of OFC from mode-locked lasers are not resolvable by these instruments.
Within this work, a fibre-based approach for generation of OFC specifically in the low- and medium resolution range is studied numerically. This approach consists of three optical fibres that are fed by two equally intense continuous-wave (CW) lasers. The first fibre is a conventional single-mode fibre, the second one is a suitably pumped amplifying Erbium-doped fibre with anomalous dispersion, and the third one is a low-dispersion highly nonlinear optical fibre. The evolution of a frequency comb in this system is governed by the following processes: as the two initial CW-laser waves with different frequencies propagate through the first fibre, they generate an initial comb via a cascade of four-wave mixing processes. The frequency components of the comb are phase-correlated with the original laser lines and have a frequency spacing that is equal to the initial laser frequency separation (LFS), i.e. the difference in the laser frequencies. In the time domain, a train of pre-compressed pulses with widths of a few pico-seconds arises out of the initial bichromatic deeply-modulated cosine-wave. These pulses undergo strong compression in the subsequent amplifying Erbium-doped fibre: sub-100 fs pulses with broad OFC spectra are formed. In the following low-dispersion highly nonlinear fibre, the OFC experience a further broadening and the intensity of the comb lines are fairly equalised. This approach was mathematically modelled by means of a Generalised Nonlinear Schrödinger Equation (GNLS) that contains terms describing the nonlinear optical Kerr effect, the delayed Raman response, the pulse self-steepening, and the linear optical losses as well as the wavelength-dependent Erbium gain profile for the second fibre. The initial condition equation being a deeply-modulated cosine-wave mimics the radiation of the two initial CW lasers. The numerical studies are performed with the help of Matlab scripts that were specifically developed for the integration of the GNLS and the initial condition according to the proposed approach for the OFC generation. The scripts are based on the Fourth-Order Runge-Kutta in the Interaction Picture Method (RK4IP) in combination with the local error method.
This work includes the studies and results on the length optimisation of the first and the second fibre depending on different values of the group-velocity dispersion of the first fibre. Such length optimisation studies are necessary because the OFC have the biggest possible broadband and exhibit a low level of noise exactly at the optimum lengths. Further, the optical pulse build-up in the first and the second fibre was studied by means of the numerical technique called Soliton Radiation Beat Analysis (SRBA). It was shown that a common soliton crystal state is formed in the first fibre for low laser input powers. The soliton crystal continuously dissolves into separated optical solitons as the input power increases. The pulse formation in the second fibre is critically dependent on the features of the pulses formed in the first fibre. I showed that, for low input powers, an adiabatic soliton compression delivering low-noise OFC occurs in the second fibre. At high input powers, the pulses in the first fibre have more complicated structures which leads to the pulse break-up in the second fibre with a subsequent degradation of the OFC noise performance. The pulse intensity noise studies that were performed within the framework of this thesis allow making statements about the noise performance of an OFC. They showed that the intensity noise of the whole system decreases with the increasing value of LFS.
The main goal of this cumulative thesis is the derivation of surface emissivity data in the infrared from radiance measurements of Venus. Since these data are diagnostic of the chemical composition and grain size of the surface material, they can help to improve knowledge of the planet’s geology. Spectrally resolved images of nightside emissions in the range 1.0-5.1 μm were recently acquired by the InfraRed Mapping channel of the Visible and InfraRed Thermal Imaging Spectrometer (VIRTIS-M-IR) aboard ESA’s Venus EXpress (VEX). Surface and deep atmospheric thermal emissions in this spectral range are strongly obscured by the extremely opaque atmosphere, but three narrow spectral windows at 1.02, 1.10, and 1.18 μm allow the sounding of the surface. Additional windows between 1.3 and 2.6 μm provide information on atmospheric parameters that is required to interpret the surface signals. Quantitative data on surface and atmosphere can be retrieved from the measured spectra by comparing them to simulated spectra. A numerical radiative transfer model is used in this work to simulate the observable radiation as a function of atmospheric, surface, and instrumental parameters. It is a line-by-line model taking into account thermal emissions by surface and atmosphere as well as absorption and multiple scattering by gases and clouds. The VIRTIS-M-IR measurements are first preprocessed to obtain an optimal data basis for the subsequent steps. In this process, a detailed detector responsivity analysis enables the optimization of the data consistency. The measurement data have a relatively low spectral information content, and different parameter vectors can describe the same measured spectrum equally well. A usual method to regularize the retrieval of the wanted parameters from a measured spectrum is to take into account a priori mean values and standard deviations of the parameters to be retrieved. This decreases the probability to obtain unreasonable parameter values. The multi-spectrum retrieval algorithm MSR is developed to additionally consider physically realistic spatial and temporal a priori correlations between retrieval parameters describing different measurements. Neglecting geologic activity, MSR also allows the retrieval of an emissivity map as a parameter vector that is common to several spectrally resolved images that cover the same surface target. Even applying MSR, it is difficult to obtain reliable emissivity maps in absolute values. A detailed retrieval error analysis based on synthetic spectra reveals that this is mainly due to interferences from parameters that cannot be derived from the spectra themselves, but that have to be set to assumed values to enable the radiative transfer simulations. The MSR retrieval of emissivity maps relative to a fixed emissivity is shown to effectively avoid most emissivity retrieval errors. Relative emissivity maps at 1.02, 1.10, and 1.18 μm are finally derived from many VIRTIS-M-IR measurements that cover a surface target at Themis Regio. They are interpreted as spatial variations relative to an assumed emissivity mean of the target. It is verified that the maps are largely independent of the choice of many interfering parameters as well as the utilized measurement data set. These are the first Venus IR emissivity data maps based on a consistent application of a full radiative transfer simulation and a retrieval algorithm that respects a priori information. The maps are sufficiently reliable for future geologic interpretations.
This work investigates the influence of the Coriolis force on mass motion related to the Rheasilvia impact basin on asteroid (4) Vesta's southern hemisphere. The giant basin is 500km in diameter, with a centre which nearly coincides with the rotation axis of Vesta. The Rheasilvia basin partially overlaps an earlier, similarly large impact basin, Veneneia.
Mass motion within and in the vicinity of the Rheasilvia basin includes slumping and landslides, which, primarily due to their small linear extents, have not been noticeably affected by the Coriolis force. However, a series of ridges related to the basin exhibit significant curvature, which may record the effect of the Coriolis force on the mass motion which generated them.
In this thesis 32 of these curved ridges, in three geologically distinct regions, were examined. The mass motion velocities from which the ridge curvatures may have resulted during the crater modification stage were investigated. Velocity profiles were derived by fitting inertial circles along the curved ridges and considering both the current and past rotation states of Vesta. An iterative, statistical approach was used, whereby the radii of inertial circles were obtained through repeated fitting to triplets of points across the ridges. The most frequently found radius for each central point was then used for velocity derivation at that point.
The results of the velocity analysis are strongly supportive of a Coriolis force origin for the curved ridges. Derived velocities (29.6 ± 24.6 m/s) generally agree well with previously published predictions from numerical simulations of mass motion during the impact process. Topographical features such as local slope gradient and mass deposition regions on the curved ridges also independently agree with regions in which the calculated mass motion accelerates or decelerates.
Sections of constant acceleration, deceleration and constant velocity are found, showing that mass motion is being governed by varying conditions of topography, regolith structure and friction. Estimates of material properties such as the effective viscosities (1.9-9.0·10⁶ Pa·s) and coefficients of friction (0.02-0.81) are derived from the velocity profile information in these sections. From measured accelerations of mass motions on the crater wall, it is also shown that the crater walls must have been locally steeper at the time of the mass motion.
Together with these novel insights into the state and behaviour of material moving during the modification stage of Rheasilvia's formation, this work represents the first time that the Coriolis Effect on mass motions during crater formation has been shown to result in diagnostic features preserved until today.
The high-latitudinal thermospheric processes driven by the solar wind and Interplanetary Magnetic Field (IMF) interaction with the Earth magnetosphere are highly variable parts of the complex dynamic plasma environment, which represent the coupled Magnetosphere – Ionosphere – Thermosphere (MIT) system. The solar wind and IMF interactions transfer energy to the MIT system via reconnection processes at the magnetopause. The Field Aligned Currents (FACs) constitute the energetic links between the magnetosphere and the Earth ionosphere. The MIT system depends on the highly variable solar wind conditions, in particular on changes of the strength and orientation of the IMF.
In my thesis, I perform an investigation on the physical background of the complex MIT system using the global physical - numerical, three-dimensional, time-dependent and self-consistent Upper Atmosphere Model (UAM). This model describes the thermosphere, ionosphere, plasmasphere and inner magnetosphere as well as the electrodynamics of the coupled MIT system for the altitudinal range from 80 (60) km up to the 15 Earth radii.
In the present study, I developed and investigated several variants of the high-latitudinal electrodynamic coupling by including the IMF dependence of FACs into the UAM model. For testing, the various variants were applied to simulations of the coupled MIT system for different seasons, geomagnetic activities, various solar wind and IMF conditions. Additionally, these variants of the theoretical model with the IMF dependence were compared with global empirical models. The modelling results for the most important thermospheric parameters like neutral wind and mass density were compared with satellite measurements. The variants of the UAM model with IMF dependence show a good agreement with the satellite observations. In comparison with the empirical models, the improved variants of the UAM model reproduce a more realistic meso-scale structures and dynamics of the coupled MIT system than the empirical models, in particular at high latitudes. The new configurations of the UAM model with IMF dependence contribute to the improvement of space weather prediction.
Function by structure
(2015)
Forcing Earth’s sea level
(2015)
Organic bulk heterojunction (BHJ) solar cells based on polymer:fullerene blends are a promising alternative for a low-cost solar energy conversion. Despite significant improvements of the power conversion efficiency in recent years, the fundamental working principles of these devices are yet not fully understood. In general, the current output of organic solar cells is determined by the generation of free charge carriers upon light absorption and their transport to the electrodes in competition to the loss of charge carriers due to recombination.
The object of this thesis is to provide a comprehensive understanding of the dynamic processes and physical parameters determining the performance. A new approach for analyzing the characteristic current-voltage output was developed comprising the experimental determination of the efficiencies of charge carrier generation, recombination and transport, combined with numerical device simulations.
Central issues at the beginning of this work were the influence of an electric field on the free carrier generation process and the contribution of generation, recombination and transport to the current-voltage characteristics. An elegant way to directly measure the field dependence of the free carrier generation is the Time Delayed Collection Field (TDCF) method. In TDCF charge carriers are generated by a short laser pulse and subsequently extracted by a defined rectangular voltage pulse. A new setup was established with an improved time resolution compared to former reports in literature. It was found that charge generation is in general independent of the electric field, in contrast to the current view in literature and opposed to the expectations of the Braun-Onsager model that was commonly used to describe the charge generation process. Even in cases where the charge generation was found to be field-dependend, numerical modelling showed that this field-dependence is in general not capable to account for the voltage dependence of the photocurrent. This highlights the importance of efficient charge extraction in competition to non-geminate recombination, which is the second objective of the thesis.
Therefore, two different techniques were combined to characterize the dynamics and efficiency of non-geminate recombination under device-relevant conditions. One new approach is to perform TDCF measurements with increasing delay between generation and extraction of charges. Thus, TDCF was used for the first time to measure charge carrier generation, recombination and transport with the same experimental setup. This excludes experimental errors due to different measurement and preparation conditions and demonstrates the strength of this technique. An analytic model for the description of TDCF transients was developed and revealed the experimental conditions for which reliable results can be obtained. In particular, it turned out that the $RC$ time of the setup which is mainly given by the sample geometry has a significant influence on the shape of the transients which has to be considered for correct data analysis.
Secondly, a complementary method was applied to characterize charge carrier recombination under steady state bias and illumination, i.e. under realistic operating conditions. This approach relies on the precise determination of the steady state carrier densities established in the active layer. It turned out that current techniques were not sufficient to measure carrier densities with the necessary accuracy. Therefore, a new technique {Bias Assisted Charge Extraction} (BACE) was developed. Here, the charge carriers photogenerated under steady state illumination are extracted by applying a high reverse bias. The accelerated extraction compared to conventional charge extraction minimizes losses through non-geminate recombination and trapping during extraction. By performing numerical device simulations under steady state, conditions were established under which quantitative information on the dynamics can be retrieved from BACE measurements.
The applied experimental techniques allowed to sensitively analyse and quantify geminate and non-geminate recombination losses along with charge transport in organic solar cells. A full analysis was exemplarily demonstrated for two prominent polymer-fullerene blends.
The model system P3HT:PCBM spincast from chloroform (as prepared) exhibits poor power conversion efficiencies (PCE) on the order of 0.5%, mainly caused by low fill factors (FF) and currents. It could be shown that the performance of these devices is limited by the hole transport and large bimolecular recombination (BMR) losses, while geminate recombination losses are insignificant. The low polymer crystallinity and poor interconnection between the polymer and fullerene domains leads to a hole mobility of the order of 10^-7 cm^2/Vs which is several orders of magnitude lower than the electron mobility in these devices. The concomitant build up of space charge hinders extraction of both electrons and holes and promotes bimolecular recombination losses.
Thermal annealing of P3HT:PCBM blends directly after spin coating improves crystallinity and interconnection of the polymer and the fullerene phase and results in comparatively high electron and hole mobilities in the order of 10^-3 cm^2/Vs and 10^-4 cm^2/Vs, respectively. In addition, a coarsening of the domain sizes leads to a reduction of the BMR by one order of magnitude. High charge carrier mobilities and low recombination losses result in comparatively high FF (>65%) and short circuit current (J_SC ≈ 10 mA/cm^2). The overall device performance (PCE ≈ 4%) is only limited by a rather low spectral overlap of absorption and solar emission and a small V_OC, given by the energetics of the P3HT.
From this point of view the combination of the low bandgap polymer PTB7 with PCBM is a promising approach. In BHJ solar cells, this polymer leads to a higher V_OC due to optimized energetics with PCBM. However, the J_SC in these (unoptimized) devices is similar to the J_SC in the optimized blend with P3HT and the FF is rather low (≈ 50%). It turned out that the unoptimized PTB7:PCBM blends suffer from high BMR, a low electron mobility of the order of 10^-5 cm^2/Vs and geminate recombination losses due to field dependent charge carrier generation.
The use of the solvent additive DIO optimizes the blend morphology, mainly by suppressing the formation of very large fullerene domains and by forming a more uniform structure of well interconnected donor and acceptor domains of the order of a few nanometers. Our analysis shows that this results in an increase of the electron mobility by about one order of magnitude (3 x 10^-4 cm^2/Vs), while BMR and geminate recombination losses are significantly reduced. In total these effects improve the J_SC (≈ 17 mA/cm^2) and the FF (> 70%). In 2012 this polymer/fullerene combination resulted in a record PCE for a single junction OSC of 9.2%.
Remarkably, the numerical device simulations revealed that the specific shape of the J-V characteristics depends very sensitively to the variation of not only one, but all dynamic parameters. On the one hand this proves that the experimentally determined parameters, if leading to a good match between simulated and measured J-V curves, are realistic and reliable. On the other hand it also emphasizes the importance to consider all involved dynamic quantities, namely charge carrier generation, geminate and non-geminate recombination as well as electron and hole mobilities. The measurement or investigation of only a subset of these parameters as frequently found in literature will lead to an incomplete picture and possibly to misleading conclusions.
Importantly, the comparison of the numerical device simulation employing the measured parameters and the experimental $J-V$ characteristics allows to identify loss channels and limitations of OSC. For example, it turned out that inefficient extraction of charge carriers is a criticical limitation factor that is often disobeyed. However, efficient and fast transport of charges becomes more and more important with the development of new low bandgap materials with very high internal quantum efficiencies. Likewise, due to moderate charge carrier mobilities, the active layer thicknesses of current high-performance devices are usually limited to around 100 nm. However, larger layer thicknesses would be more favourable with respect to higher current output and robustness of production. Newly designed donor materials should therefore at best show a high tendency to form crystalline structures, as observed in P3HT, combined with the optimized energetics and quantum efficiency of, for example, PTB7.