Alumni 2019

Justus Zorn     (Germany)                                                                                                                                               18.12.2019

Cherenkov Camera and Analysis Development for Highest-Energy Gamma-Ray Astronomy  ( thesis pdf, 60 MB )

Imaging atmospheric Cherenkov telescopes are used for the detection of highest-energy γ-rays. This thesis focuses on two experiments equipped with such telescopes: The operating High Energy Stereoscopic System (H.E.S.S.) and the future Cherenkov Telescope Array (CTA). Four of the five H.E.S.S. cameras saw a major electronics upgrade a few years ago enabling improved readout and analysis techniques mainly at the highest energies. The Compact High-Energy Camera (CHEC) is a design for the Small-Sized Telescopes of CTA focusing on the detection of γ-rays with energies exceeding 1 TeV. The first part of the thesis is dedicated to the characterisation of two CHEC prototype cameras developed successively: CHEC-M and CHEC-S. I present results of laboratory and on-telescope measurements for both cameras. In the case of CHEC-S, I focus on those parameters that had been shown to be performance-limiting in CHEC-M and which were therefore addressed in the design iteration for CHEC-S. The second part is devoted to the upgraded H.E.S.S. cameras. I present results of Monte-Carlo simulation studies, analysis developments, and performance measurements using full-waveform readout. In the former case I demonstrate a general consistency between simulations and measurements, in the latter case I show that the use of full-sampled waveform readout improves the performance, especially at the highest energies. In the last part, I present a new analysis of the Galactic γ-ray source HESS J1646–458 which is associated with Westerlund 1, the most massive stellar cluster in our Galaxy.

Supervisor:    Jim Hinton (MPIK)

Steffi Xiang-Ting Yen     (USA)                                                                                                                                          09.12.2019

Peering into the Milky Way disk: Gaia's Perspective of our Galaxy's Open Clusters  ( thesis pdf, 6 MB )

In the pre-Gaia era, one of the largest collections of open clusters and their parameters was the Milky Way Star Cluster (MWSC) catalog, which consisted of 2808 open clusters. This sample was nearly complete up to a distance of about 1.8~kpc from the Sun, with the exception of a subset of old nearby clusters. However, with the unprecedented precision of astrometric and photometric data from Gaia, a more accurate census on the number of true open clusters can be achieved, also providing improved cluster parameters and detection of new clusters. With this aim, I developed an automated cluster characterization pipeline to consistently determine cluster membership from astrometry and cluster parameters via isochrone fitting to cluster multi-band photometry. Using Gaia DR1/TGAS data, I analyzed 24 nearby open clusters and found evidence for the non-existence of some clusters. With Gaia DR2 data, I reanalyzed the full MWSC cluster sample, successfully obtaining cluster memberships and parameters for 1873 clusters and denying the existence of 912 clusters. This is the first study, to date, to homogeneously analyze the largest catalog of open clusters. My results show that unlike previously thought, the open cluster census is very incomplete, even at the smallest distances, and thus a dedicated search for new clusters is required in order to gain a full understanding of the open cluster population.

Supervisor:    Sabine Reffert  (LSW)

Faiezeh Shabani     (Iran)                                                                                                                                                    12.11.2019

A Study of Spiral Structures and Lopsidedness in LEGUS Disk Galaxies  ( thesis pdf, 30 MB )

We present a detailed study of spiral structure and lopsidedness in four LEGUS disk galaxies. We first search for a possible age gradient of star clusters across the spiral arms, as predicted by the spiral density wave theory. We find a clear age sequence across the grand-design arms of NGC 1566, indicating that the density wave theory is the primary mechanism for the formation of spiral arms in this galaxy. In contrast, we find no offset in the location of star clusters with different ages in M51a and NGC 628, suggesting that tidal interactions and swing amplification theory might be the dominant scenarios to generate spiral patterns in these galaxies, respectively. Later, we search for a link between the asymmetric distribution of the star clusters along the spiral arms and the physical properties of NGC 3344, NGC 1566, NGC 628, and M51a. For this purpose, we construct 2D spatially resolved stellar mass and SFR surface density maps of the galaxies applying a pixel-by-pixel spectral energy distribution fitting technique. We do not observe any lopsidedness in the distribution of stellar mass surface density of the two arms in our target galaxies. On the contrary, our derived resolved SFR surface density maps of NGC 3344, NGC 1566, and the inner regions of M51a confirm the presence of lopsidedness between the two arms in a similar fashion to the distribution of the young star clusters. In the case of NGC 628, we do not observe any correlation between the lopsidedness in the distribution of star clusters and its SFR surface density.

Supervisor:    Eva Grebel  (ARI)

Mikhail Kovalev   (Russia)                                                                                                                                                11.11.2019

NLTE analysis of the Gaia-ESO spectroscopic survey     ( thesis pdf, 5 MB )

In this work, we develop the spectral analysis pipeline that combines the NLTE synthetic spectra with the Payne spectral model (Ting et al.2018). We compute two spectral grids, one grid with all elements treated in LTE and a second grid with iron, magnesium, titanium and manganese modelled in NLTE, to study the NLTE effects on the determination of the stellar parameters and chemical abundances. This pipeline is applied to spectra from the third public data release of the Gaia-ESO (Gilmore et al. 2012). We validate the method on the subsample of standard stars and cluster’s members and find significant differences between NLTE and LTE in the metal-poor regime. All clusters are homogeneous in Fe and Ti, but several globular clusters showed significant dispersion in [Mg/Fe]. The depletion of the mean [Mg/Fe] in several globular clusters, compared to field stars of the same metallicity, may indicate their ex-situ formation history. We also combine our NLTE results with Gaia DR2 (Gaia Collaboration et al.2018) astrometric data to study Galactic chemo-dynamic evolution. We apply chemical separation to select high-[α/Fe] and low-[α/Fe] disk populations, with halo stars selected using kinematics. We confirm previous results like super-solar Mg abundance in the metal-poor regime and decrease of [Mg/Fe] to the solar abundances in the metal-rich regime. We find a constant NLTE [Mg/Fe]∼0.3 dex ratio for both metal-poor disk and halo with relatively small star-to-star scatter. The halo stars with solar-like [Mg/Fe] are probably accreted from the other galaxies. The observed eccentricity distribution for high-[α/Fe] disk population rules out a violent thick disk formation mechanisms like direct accretion and dynamic heating. The measured chemical and kinematic gradients and velocity dispersions of the high-[α/Fe] population can be explained by the gas-rich merger scenario with the non-negligible contribution from the radial migration.

Supervisor:    Maria Bergemann (MPIA)

Taras Panamarev    (Kazakhstan)                                                                                                                                        06.11.2019

Dynamical interaction of supermassive black holes with the surrounding stellar system  ( thesis pdf, 10 MB )

We use high resolution direct N-body simulation methods to study the stellar dynamics in the Galactic centre (GC) and in active galactic nuclei (AGN). We follow the evolution of the GC from the assumed in-situ formation of the nuclear star cluster (NSC) up to 5 Gyr using one million particles taking into account single and binary stellar evolution. We investigate 3D stellar density distributions, obtain rate of tidal disruption events, rate of hypervelocity star ejections and estimate the number of extreme mass-ratio inspirals. We examine the survivability of binary stars in the NSC and discuss the contribution of binaries with compact objects in presence of pulsars and Supernovae Ia rates in the GC. We use more simplified model to study the stellar dynamics in AGN. The analysis of two simulations with 128000 particles (with and without the accretion disk (AD)) shows that the interaction of the NSC with the AD leads to formation of a stellar disc in the central part of the NSC. We derive the mass and size of the formed stellar disc and discuss possible existence of such discs in some nearby galaxies.

Supervisor:    Andreas Just / Andreas Quirrenbach  (ARI/LSW)

Grigorios Katsoulakos      (Greece)                                                                                                                                   04.11.2019

Nonthermal Processes Near Supermassive Black Holes  ( thesis pdf, 20 MB )

In recent years, γ-ray astronomy has made considerable progress in the exploration of the extragalactic γ-ray sky. In particular, active galaxies, whose relativistic jets/outflows are significantly inclined with respect to the line-of-sight, have revealed remarkable flaring activity at γ-ray energies. The observed rapid variability of the γ-ray emission, comparable to timescales of the light travel time across the black hole horizon, provides a strong motivation for testing radiative scenarios associated with the vicinity of the central supermassive black hole. In this doctoral study, we explore the so-called black hole magnetospheric scenario. Accordingly, strong particle acceleration may occur within the black hole magnetosphere in regions of unscreened electric fields (gaps). This can happen either at the null surface across which the charge density changes sign or at the stagnation surface which separates the inwardly from the outwardly moving matter. The acceleration of leptons is accompanied by γ-ray emission via inverse Compton scattering of the ambient (disk) soft photons as well as curvature radiation. This thesis explores the potential of these processes to account for the observed γ-ray features. By developing and studying an one-dimensional, steady model for magnetospheric particle acceleration and emission, as well as, estimating the terminal Lorentz factors of the accelerated charges and the maximum extractable gap power, we find that magnetospheric processes can be responsible for the observed, rapidly variable very-high-energy γ-ray emission in the radio galaxy M87.

Supervisor:    Frank Rieger  (ITA/MPIK)

Hestor Hiss     (BrAsil/Germany)                                                                                                                                          15.10.2019

Measuring the Thermal State of the Intergalactic Medium   ( thesis pdf, 10 MB )

The thermal evolution of the low density intergalactic medium (IGM) is a major diagnostic tool for understanding the nature and evolution of the predominant component of baryonic matter in the universe. In this study I present different approaches for measuring the thermal state of the IGM at different ages of the universe, in order to understand how it is affected by reionization processes. The main observable used to probe the thermal state of the gas is the so-called Lyman-alpha forest. This observable consists of a series of absorption lines in the spectra of distant quasi-stellar objects (QSOs) which arise due to the presence of residual intervening neutral hydrogen in the IGM between the observer and the QSO. Decomposing the Lyman-alpha forest into discrete absorption profiles allows one to explore how the distribution of Lyman-alpha absorption line widths and column densities (b-NHI distribution) depends on the thermal state of the gas, which is characterized by a temperature-density relation. In this thesis, I quantify the parameters of the temperature-density relation using high quality UVES and HIRES QSO spectra and state of the art cosmological hydrodynamic simulations. In the first part of this study, I apply a traditional cutoff Fitting method to the b-NHI distribution of the QSO spectra. Using simulations, I calibrate how the position of the cutoff in the b-NHI distribution relates to the thermal state of the IGM. I find that the thermal evolution of the IGM shows clear signatures of He II reionization at 2 < z < 3.4. In the second part of this thesis, I present a novel statistical method for constraining the thermal state of the IGM using the full shape of the b-NHI distribution. I show that this method is more accurate and precise than the traditional cutoff Fitting approach, by applying it to mock data realizations. I confirm this by applying it to observational data at z = 2. Finally, using this novel method, I quantify for the first time the parameters of the temperature-density relation at low redshift (z = 0.1) using the b-NHI distribution, and find broad agreement with theoretical expectations. Overall, this thesis demonstrates that the b-NHI distribution is a powerful statistical tool for studying the intergalactic medium and can place strong constraints on the evolution of its thermal state.

Supervisor:    Joe Hennawi  (MPIA)

Mayte Alfaro Cuello    (Chile)                                                                                                                                          25.07.2019

The Nucleus of the Sagittarius Dwarf Spheroidal Galaxy: M54   ( thesis pdf, 20 MB )

Nuclear star clusters are the densest stellar systems in the universe, hosted by galaxies across the entire Hubble sequence, including a high fraction of dwarf galaxies. The most massive, chemically complex globular clusters in the Milky Way exhibit similar characteristics as nuclear star clusters in dwarf galaxies. This raised the idea that these globular clusters are actually former nuclei of galaxies accreted by the Milky Way. In this context, M54 – the nuclear star cluster of the Sagittarius dwarf spheroidal galaxy (Sgr dSph) – offers a unique opportunity to understand this presumed direct connection between globular clusters and nuclear star clusters, and low-mass galaxy nucleation. The Sgr dSph is currently being disrupted by the tidal field of the Milky Way, leaving a long stellar stream as evidence of its advanced degree of disruption. M54 still lies at the center of its host, becoming a potential stripped nucleus, and presenting an outstanding example of this class of objects. From its discovery – long before the detection of the Sgr dSph – M54 was classified as a globular cluster, the second most massive in the Milky Way after ω Cen. M54 shows a high spread in iron abundance of its member stars, pointing towards an extended and complex star formation history. This Thesis presents a large Multi-Unit Spectroscopic Explorer (MUSE) data set covering a region of ∼2.5 times the effective radius of M54. The single spectra of more than 6 500 member stars extracted from the exceptional data set led to the recovery of the star formation history of this nuclear star cluster through age and metallicity information. This allowed disentangling the presence of – at least – three stellar subpopulations, whose kinematics show clear differences. The chemo-dynamical characterization of these subpopulations suggests that they originated in different star formation events. This work shows the complexity of M54 which appears to be a nuclear star cluster in a highly disturbed environment rather than a simple globular cluster. The evidence suggests that M54 is the result of the two proposed mechanisms for the formation of nuclear stars clusters, that happen at different stages of M54’s evolution: (i) at least two globular clusters are driven to the center of the host and merge to form a single high-mass cluster with a large age and metallicity spread, followed by (ii) in-situ star formation from enriched gas in the nucleus. The unprecedented details of this study help to understand low-mass galaxy nuclei, for which less information is available in contrast to the higher mass regime.

Supervisor:    Nadine Neumayer   (MPIA)

Sarah leslie    (Australia)                                                                                                                                                 24.07.2019

The cosmic evolution of star-forming galaxies   ( thesis pdf, 20 MB )

Galaxies separate into passive and star-forming galaxies, where star-forming galaxies show a tight correlation between stellar mass and star formation rate (SFR), referred to as the star forming main sequence. In this thesis, we have measured the average galaxy star formation rate – stellar mass relation out to z~5 in the COSMOS (Cosmic Evolution Survey) field. This survey includes deep radio observations taken as part of the VLA-COSMOS 3 GHz large program that provide a dust-unbiased view of star-formation. To measure SFRs over a wide range of galaxy masses, including galaxies too faint to be detected individually, we employed a stacking analysis on the 3 GHz data. We found a flattening of the star-forming main sequence at high masses that can be explained by the increasing fraction of bulge-dominated galaxies which follow a shallower SFR – stellar mass relation than disk-dominated galaxies. As bulges grow more prominent in the low-redshift galaxy population, the flattening of the main sequence becomes more significant. We found that galaxy environment, probed by X-ray-groups and local galaxy number density, has no significant effect on the shape of the star-forming main sequence at z>0.3. We have compared SFRs derived from publicly available mid-infrared (MIR), far-infrared (FIR), radio, and ultraviolet (UV) photometry for massive star-forming galaxies selected consistently at z~0 and z~0.7. We probed the dust properties of these massive disk galaxies by analysing how their UV luminosity depends on galaxy inclination. By comparing our observed trends with radiative transfer model predictions we constrained the average opacity and clumpiness of the dust. We found that UV attenuation has increased between z~0 and z~0.7 by a factor of 3.5. A higher fraction of clumpy dust around nascent star-forming regions can explain the substantial UV attenuation at z~0.7. If the gas and dust geometry at high-redshift are significantly different than inferred from our current models, this would have significant implications for our SFR calibrations relying on the UV and IR emission from galaxies. Reproducing the spatial distribution of galaxy components like stellar mass, newly formed stars, dust, metals, and gas will be a key objective for future theories of galaxy evolution.

Supervisor:    Eva Schinnerer (MPIA)

Priscilla Chauke    (South Africa)                                                                                                                                     24.07.2019

Determining Properties of LEGA-C Galaxies through Spectral Star-formation History Reconstruction   ( thesis pdf, 15 MB )

Over the past decade, photometric and spectroscopic surveys have enabled us to obtain an integrated view of galaxy evolution. We have measured the cosmic star formation history, and we know that about half of the stars we observe formed before the Universe was half of its current age. However, crucial knowledge of individual galaxy evolution has been limited because the detailed stellar population properties that we know about galaxies, such as ages, metallicities and kinematics, have mostly been obtained from nearby galaxies, which contain mostly old stellar populations. To advance our understanding of galaxy evolution, measurements have to be extended to more distant galaxies, which have a wider variation in age and star formation (SF) activity. In this thesis, I investigate the stellar population properties of a galaxy population at half the Universe’s current age (redshift z ~ 1), using high quality, high resolution spectra from the (recent) LEGA-C Survey. I present an algorithm that uses full-spectrum fitting to reconstruct the star formation histories (SFHs) of LEGA-C galaxies. First, I investigate the dependance of individual SFHs on stellar mass, stellar velocity dispersion and SF activity. The individual SFHs of high-mass (velocity dispersion) quiescent and star-forming populations are increasingly divergent towards lower redshifts, which indicates a strong correlation between current SF activity and past SF activity. Second, I trace the stellar mass evolution of LEGA-C galaxies between redshifts z ~ 1 and z = 3 (when the Universe was a sixth of its age). Galaxies that are similar in stellar mass at one redshift, have diverse evolutionary paths that lead to a wide range of stellar masses at another redshift, whether they are being traced backward or forward in time. Last, I investigate rejuvenation in quiescent galaxies, that is, when galaxies that are quiescent, re-ignite their SF before becoming quiescent again. Although, rejuvenation events do not contribute significantly to the growth of quiescent galaxies, a majority of rejuvenated galaxies lie in the ‘green valley’, where galaxies are thought to be in the (one-way) transition phase from the blue cloud to the red sequence. Reconstructing SFHs has allowed us to trace the individual pathways along which galaxies evolve, and investigate the physical processes that drive them. Measurements from this work, as well as other LEGA-C studies, will be important for future studies at even higher redshifts because they will act as a benchmark to connect populations at higher redshifts to the nearby Universe.

Supervisor:    Arjen van de Wel  (MPIA/ U. Gent)

Sebastian Bustamante    (Colombia)                                                                                                                                  24.07.2019

Modelling supermassive black hole spins and the metallicity evolution of merging galaxies in a cosmological context   ( thesis pdf, 10 MB )

Galaxy mergers are a key process in the current paradigm of hierarchical galaxy formation. They are responsible for many galaxy transformations. For instance, they can drive changes in galaxy morphology and colour, boost the star formation activity and trigger strong inflows of pristine gas that dilute the central gas metallicity and feed the central supermassive black hole. To which extent and how galaxy mergers affect some of the galaxy properties are still open questions. In the first part of this thesis, we use simulations and observations to study the processes of merger-induced gas metallicity dilution and star formation enhancement. We find that merging and post-merger galaxies constitute a prominent outlier population in the fundamental metallicity relation, a relation that links the galaxy stellar mass with the star formation rate and the gas metallicity. In the second part of this thesis, we present a sub-grid model of supermassive black hole spin evolution in cosmological simulations of galaxy formation. We find that the accretion discs of low-mass black holes accrete mass in a coherent fashion, resulting in high spin values. On the other hand, high-mass black holes have lower spins because gas accretion is either chaotic or completely suppressed, in which case black hole binary coalescence resulting from galaxy mergers is the only relevant evolution channel. Finally, we test a hypothesis in which the black hole energy feedback transitions from a thermal mode to a kinetic mode at the onset of self-gravity fragmentation in the accretion disc, and is thus coupled to the spin evolution of the black hole. We find that our conjecture reproduces the galaxy colour bimodality and the colour-morphology relation.

Supervisor:    Volker Springel (HITS)

Ying Chi "Gigi" Leung     (Hong Kong)                                                                                                                                         16.07.2019

Constraining the nature of dark matter in galaxies with multi-tracer dynamical models   ( thesis pdf, 100 MB )

The detailed mass distribution in galaxies provides important constraints on the nature of dark matter (DM), especially in relation to the baryonic content and feedback efficiency of a galaxy. In this thesis I use multiple kinematic tracers and a diverse set of dynamical models to simultaneously constrain DM density profiles, halo shapes and the evolutionary history of galaxies. I first show that the most common and advanced stellar dynamical models can reproduce the circular velocities (as traced independently by molecular gas rotation curves), to within ⇠10% accuracy. I further use high resolution observations to understand the sources (gravitational, feedback driven) of high velocity dispersion ionised gas. By incorporating realistic birth conditions for globular clusters (GCs) and flexible, self-consistent velocity distribution functions for the Fornax dSph, I am able to understand the survival of its five GCs. The comprehensive evolutionary model suggests that Fornax has a large DM core (&1.5 kpc) and has undergone a past merger of mass ratio ⇠1:2 to 1:5. Finally, by combining stellar and gas kinematic tracers together in a single dynamical model, I provide evidence that the isolated dwarf irregular galaxy WLM has a DM halo that has both an inner density core (! ⇠ 0.3±0.1), and a prolate axis ratio of 2:1. The recovered orbit structure (tangential anisotropic) is very similar to nearby dSph galaxies - suggesting that internal processes rather than tidal origin may lead to this dynamical configuration. The DM halo profile is consistent with the ⇤CDM cosmological picture when baryonic feedback is included. The prolate geometry is difficult for MOND and at the same time challenges self-interacting DM (SIDM) theories to create a thermalised DM core of the observed size, without sphericalising the halo. From both the dynamical models on WLM and Fornax, I am able to provide constraints on the particle mass of Bose-Einstein condensate DM models to 1.1 − 1.3 ⇥ 10−22 eV/c2, and interaction cross section for (velocity independent) SIDM particles of 0.8 . "/mSIDM . 3.1 cm2/g - though it remains to be seen that these can produce the proper core size and shape in the DM halos we find. Application of these new techniques and models to more galaxies will provide even tighter constraints on dark matter particle models.

Supervisor:    Glenn van de Ven  (MPIA)

Natascha Manger    (Germany)                                                                                                                                         16.07.2019

High Resolution Simulations of Structure Formation in Turbulent Protoplanetary Disks: A Case Study of the Vertical Shear Instability   ( thesis pdf, 20 MB )

Disks around young stars are the birth place of planetary systems like our own solar system. Thus, the study of turbulent processes in protoplanetary disks is not only important to understand the transport of angular momentum to explain for example the angular momentum deficit of our own sun, but also to understand how large scale structures emerge, which are recently regularly observed and which also represent a crucial puzzle piece in the understanding of how dust grains can grow into planetesimals via gravoturbulent processes. In this thesis, I conduct high resolution studies of three-dimensional global models of turbulent protoplanetary disks using the magneto-hydrodynamics code PLUTO. I focus my studies on the Vertical Shear Instability (VSI), which has been shown to operate efficiently at disk radii beyond a few AU in typical protoplanetary disks. I show that vortices with radial diameters of around 1.5 local pressure scale heights and aspect ratios χ > 8 form in VSI turbulent disks and that these vortices can survive more than 500 orbits. The vortices are forming irrespective of the underlying disk density gradient and aspect ratio and can therefore act as pressure traps for small to medium sized particles over a wide range of the disk. I also show evidence that these dusty vortices are compatible with detections of dust concentrations by current sub-mm interferometers. These findings therefore present a crucial puzzle piece which will help the understanding under which conditions and how early after the formation of a disk around a young star planetesimals can form via gravoturbulent planetesimal formation.

Supervisor:    Hubert Klahr (MPIA)

Marcelo tala Pinto    (Chile)                                                                                                                                             16.07.2019

Perspectives of the Radial Velocity Method: Physical Modeling of the Wavelength Solution & Exoplanetary Detections around Giant Stars   ( thesis pdf, 40 MB )

First, I present the design and construction of the Front-end of the Waltz Spectrograph. I designed upgrades in the system to optimize its mechanical stability and light coupling efficiency. To improve the instrument performance, I changed the original spectrograph fiber for one with slightly larger core and better focal ratio degradation, providing a spectral resolving power of 60.000, a mean spectral sampling of 2.6 pixels and only 5% losses due to focal ratio degradation. In the second part I present a ray tracing software that calculates the optical path of individual rays through an echelle spectrograph from the slit to the detector. By including the effects of the environment on the physical properties of the optical elements that compose the spectrograph, I am able to reproduce some of the trends observed in the time series of the spectral line positions of the calibration data. Finally, I report the discovery of two planets orbiting the stars HD 25723 and 17 Sco and two planet candidates orbiting 3 Cnc and 44 Uma. Also, I investigate the planet occurrence rates as a function of evolutionary stage for two surveys, Lick and EXPRESS, concluding that there is no strong effect of stellar evolution in planet occurrence rates.

Supervisor:   Andreas Quirrenbach (LSW)

Sven Buder    (Germany)                                                                                                                                                      16.07.2019

Spectroscopic Analysis and Chemodynamic Exploration of the Milky Way with Million-Star Survey   ( thesis pdf, 100 MB )

The Milky Way is traditionally depicted as a composite of three main stellar components: the halo, disk, and bulge. The closer we look and the more information we gather, this picture becomes more complicated. The components are in fact overlapping in stellar properties, such as positions, kinematics, ages, and chemical composition. The most promising way to further our understanding of the formation of the Milky Way, the main aim of Galactic Archaeology, is to explore chemical information and stellar ages with large stellar surveys.
In Chapter 1, we motivate this exploration in more detail and describe the challenges that come with the collection of big data from spectroscopic surveys. In Chapter 2, we introduce the large-scale spectroscopic `Galactic Archaeology with HERMES' (GALAH) survey, whose spectra were analysed in the course of this Thesis in order to estimate stellar properties. The very large data flow of stellar surveys has been referred to as the industrial revolution of Galactic archaeology and the analysis requires new efficient and automated techniques. This Thesis describes the spectroscopic analyses of the more than 650,000 stars in the GALAH survey, which deliver up to 30 abundances with unprecedented accuracy on such scales, based on the papers by Buder et al. (2018) and Buder et al. (in prep. b). In Chapter 3, we use the chemical information from GALAH together with dynamical information and stellar ages to analyse the Galactic disk in the solar neighborhood, which has shown to be assembled from two populations. This work is published by Buder et al. (2019) and we find that the two populations of the disk can be separated more clearly when using stellar chemical composition and age, rather than phase-space information. In Chapter 4, we use chemodynamic information and stellar ages to analyse the transition between the disk and halo. This work will be submitted by Buder et al. (in prep. a) and we confirm that the old disk overlaps significantly with the halo. With our performed chemodynamic decompositions we are able to link the disk population that is enhanced in alpha-element abundances with the high-alpha halo population identified by Nissen & Schuster (2010). We further show that the accreted halo population of the `Sausage', identified by Belokurov et al. (2018), is strongly correlated with the low-alpha halo population identified by Nissen & Schuster (2010). Our stellar age estimates suggest that the halo components and the oldest high-alpha disk stars are coeval and show no strong age gradient, which would rule out several formation scenarios of the Milky Way. In Chapter 5, we conclude the work of this Thesis and outline further ways to continue the research in the field of Galactic archaeology with large stellar surveys.

Supervisor:  Karin Lind (MPIA)

Matthias Samland    (Germany)                                                                                                                                             16.07.2019

High-Contrast Imaging Characterization of Exoplanets   ( thesis pdf, 20 MB )

Direct imaging of exoplanetary systems and the spectral characterization of exoplanetary atmospheres are amongst the most challenging, as well as rapidly developing fields in astronomy, propelled by new technologies and observational strategies. In this thesis, I contributed to the atmospheric analysis of exoplanets, the development of new algorithms to find faint planet signatures in the data, and the improvement of the fidelity of obtained exoplanet spectra. I performed atmospheric analyses of directly imaged planets observed with the planet imaging instrument VLT/SPHERE. For this purpose, I wrote a statistical inference code (BACON, Bayesian Atmospheric CharacterizatiON), which uses self-consistently computed model atmospheres to derive atmospheric parameters. The planets I studied in this thesis are: 51 Eridani b, one the coldest methane-rich directly imaged planets; PDS 70 b, the first young planet discovered inside the gap of its host star’s transition disk; HIP 65426 b, a planet of similar spectral type to PDS 70 b, but hotter and older; and GJ 504 b, a colder methane-rich companion which, depending on its age, could be a planet or brown dwarf. The new algorithm I developed to detect planets in high-contrast imaging data shifts the focus from an image analysis interpretation of the data, towards a time-domain analysis approach. I show that with this technique (TRAP, Temporal Reference Analysis for Exoplanets), an improvement of up to a factor of six in signal-to-noise can be achieved at very small angular separations between the planet and host star. Furthermore, I adapted the CHARIS instrument pipeline to use with SPHERE-IFS. This pipeline opens new possibilities for improving the quality of spectra obtained for exoplanets using SPHERE. Using this pipeline, I confirm the low flux emitted at around 1 micron previously obtained for 51 Eridani b, consistent with the absorption due to methane and water opacities predicted by models. Lastly, I discuss the future prospects for my work and how these approaches can be combined into a single framework.

Supervisor:   Wolfgang Brandner (MPIA)

Hans-Paul  Baehr    (USA)                                                                                                                                                   05.07.2019

Formation Criteria and Initial Constraints on Objects Formed in Gravitationally Unstable Disks   ( thesis pdf, 60 MB )

Early protoplanetary disks are cool and massive and thus subject to gravitational instabilites and fragmentation of the disk into dense clumps of gas. These fragments are massive enough to become gas giant planets and brown dwarfs in the distant regions of the disks beyond 50 au where traditional planet formation scenarios have trouble creating planetary cores fast enough to explain directly observed planets. I used high-resolution three-dimensional hydrodynamic simulations to model the collapse of self-gravitating disks to constrain the formation location of these fragments and characterize their initial gas and particle masses to compare to directly observed planets and brown dwarfs. I find the traditional cooling criterion, which constrains the formation location to the outer disk, is converged in these simulations and overall masses are consistent with massive gas giants bordering on brown dwarfs. The concentration of solid material in these fragments leads to an increase of the overall metallicity of the fragment and a solid core several tens of Earth masses. To model fragmentation with full disk simulations, I have also implemented a multigrid self-gravity solver in the PLUTO code which uses adaptive mesh refinement to resolve both the disk and fragments.

Supervisor:  Hubert Klahr (MPIA)

Stefan Brems    (Germany)                                                                                                                                                    03.07.2019

Stellar Imaging and Spectroscopy for the Discovery of Extrasolar Planets   ( thesis pdf, 20 MB )

Observing planets during their formation is a challenging, but also important task if one wants to learn about their evolution. In this thesis I explain how I have contributed to various direct imaging surveys, as well as the RVSPY radial velocity survey, which aim at finding planets around young stars. The contributions include appropriate target selection, data analysis and calibration of the instruments, in particular for the NaCo-ISPY and RVSPY surveys. I present systems I analyzed throughout the course of the NaCo-ISPY direct imaging survey which showed promising signals. Further investigation showed that all signals are most likely not caused by physical companions – with the exception of HD 101412, which is a good candidate for a system composed of a star orbited by two brown dwarfs or low mass stellar objects (< 0.1 solar masses). I explain what makes us believe this and what are our next steps to conclude on the system's nature. Further I present the current status as well as some early results of the recent RVSPY survey.

I also present results from the analysis of stellar radial velocity jitter as function of age and observational timescale. The main result is a greater than exponential decrease of the radial velocity jitter with increasing stellar age. Additionally the radial velocity jitter increases by about a factor of two when probing timescales of years instead of hours or days.

Supervisor:  Andreas Quirrenbach   (LSW)

Yulong Zhuang    (China)                                                                                                                                                      26.06.2019

Diversity of Galactic Stellar Metallicity Gradients and their Origin   ( thesis pdf, 25 MB )

In this thesis we use integral field spectroscopic observations, semi-analytic models and state of the art cosmological simulations to study the diverse formation and evolutionary mechanisms setting stellar population gradients in the inner regions of galaxies. We find a clear link between the shape of the radial stellar metallicity profiles and the stellar Surface mass density profiles within the inner (< 1 Re) regions of galaxies; independent of their mass (10^9 ≤ M_gal ≤ 10^12.5) or morphological type (E to Sd). The stellar metallicity profiles are well described by a Sérsic function, and we show that the Sérsic indices of the stellar mass density and metallicity profiles are tightly correlated, with nZ = 1.51 x nS ~ 0.53 and intrinsic scatter of ~ 0.25. The expected signatures of radial migration (bars, spiral arms, mergers) do not appear to correlate with the metallicity gradient strength, nor do the present-day stellar kinematics. We show with a semi-analytic chemical evolution model that the angular momentum of the infalling gas is an important factor in developing the metallicity and surface mass density correlations. Together, These results suggest that the in-situ, timeintegrated stellar mass build-up is the predominant factor setting the metallicity gradients of the inner regions of galaxies.

Supervisor:  Ryan Leaman / Glenn van de Ven   (MPIA)

Daniel Rahner    (Germany)                                                                                                                                                  18.06.2019

Stellar feedback and the self-regulation of star formation in giant molecular clouds: a new semi-analytic approach  ( thesis pdf, 13 MB )

Stars and the gas between them - the interstellar medium - are intrinsically coupled. Massive stars form in clouds of molecular gas and illuminate them with their radiation, thus creating regions of ionized hydrogen (HII regions) and photodissociation regions. Eventually these stars destroy their parent clouds via powerful feedback mechanisms: stellar winds, ionizing radiation, and supernova explosions. These feedback processes are a crucial self-regulation mechanism of star formation, since, as soon as the first massive stars have formed, further star formation is suppressed.

Stellar winds, radiation, and supernova feedback interact with each other in a highly non-linear manner. This complexity poses a problem not only for purely analytic approaches but also for three-dimensional hydrodynamical simulations due to the high computational cost. Here, I present a novel, semi-analytic, one-dimensional model, called WARPFIELD, which allows the cost-efficient simulation of the effects of stellar feedback from a massive star cluster on its natal giant molecular cloud (GMC). With WARPFIELD we can show that the strength of each feedback process depends strongly on time and environment. For a large range of GMC and star cluster properties we can also demonstrate that stellar feedback can naturally explain the observed inefficiency of star formation.

Supervisor:  Simon Glover (ITA)

Armelle Jardin-Blicq    (France)                                                                                                                                         12.06.2019

The TeV γ-ray emission of the Galactic Plane. HAWC and H.E.S.S. observations of the Galactic Plane and detailed study of the region surrounding 2HWC J1928+177  ( thesis pdf, 100 MB )

TeV γ-ray astronomy is led by two main ground-based facilities: Imaging Atmospheric Cherenkov Telescopes, and Water Cherenkov Detectors. This thesis focuses on H.E.S.S., an array of five Cherenkov telescopes, and HAWC, a wide field of view γ -ray observatory, that are very complementary despite their technical differences. Part of this thesis is dedicated to the HAWC high energy upgrade with the outrigger array that was developed and deployed in the time frame of this Ph.D. I introduce the calibration for the charge and time reconstruction of the photomultiplier tubes that equip each of the small outrigger tanks. I present a comparison of the galactic plane as seen by HAWC and H.E.S.S. in the region where they overlap. I highlight the differences which arise from intrinsic properties of the instruments and from their dedicated data analysis, and show that background estimation is a major ingredient. The region of the galactic plane around the source 2HWC J1928+177 is studied in more details. A multi-component fit is performed and the best fit model includes two point sources for 2HWC J1930+188 and the new source HWC J1932+192, one extended source of size σ = 0.17° for 2HWC J1928+177 and 2 extra extended sources that account for larger scale emission.

Supervisor:  Jim Hinton (MPIK)

Neven Tomicic    (Croatia)                                                                                                                                                      15.05.2019

Probing the cold phase of the interstellar medium and star formation in nearby galaxies   ( thesis pdf, 30 MB )

Properly measuring the spatial distribution of the star formation rate (SFR) in galaxies helps us to understand the driving forces for the star formation in galaxies, effects on their interstellar media (ISM), and their evolution. However, this is hindered by the uncertainties in estimating SFRs and calibrating the SFR prescriptions. These uncertainties are caused by not properly measuring the attenuation of light, probing large spatial scales, or averaging over large sample of galaxies. Additionally, the physical factors that set the efficiency whith which galaxies convert gas into stars (star formation efficiency; SFE), and their role in galactic evolution, are not yet fully understood. Variations in the SFE are difficult to disentangle from uncertainties of estimated SFRs.
The main goal of this thesis is to use optical integral field unit (IFU) observations of nearby galaxies, in order to probe the cold phase of their ISM at sub-kpc scales. We aim to measure the attenuation, thus to properly estimate the SFRs and calibrate the SFR prescriptions. We also estimate variations of SFE across the disk of an interacting galaxy. Using IFU observations of the outskirts of the Andromeda galaxy (M31) at sub-kpc scales (from 10 pc to kpc), we derive the Balmer line attenuation. By comparing attenuation with the dust mass surface density, we derive the 3-dimensional spatial distribution of dust and ionized gas in M31. Our results indicate that the vertical dust/gas distribution from the central areas of nearby galaxies differs from the outskirts of M31. From this evidence, we hypothesize that the vertical dust/gas distribution in galactic disks varies as a function of the galactocentric distance. Following that, we use extinction corrected Balmer line emission as a reference SFR tracer in a combination with ultraviolet and near-infrared images, to calibrate hybrid SFR prescriptions. We find that the hybrid SFR prescriptions do not change with spatial scale or with the subtraction of a diffuse component. However, our SFR prescriptions observed in M31 differ from the prescriptions in the literature. This indicates that the SFR prescriptions are not universal and that they may vary with the inclination and the galactocentric radius, due to varying dust/gas distributions.
Our IFU observations of the interacting galaxy NGC 2276 are used to investigate how the early phase of galactic interaction affects the ISM, and SFE across its disk. Although NGC 2276 shows a significant asymmetrically elevated SFR surface density, and asymmetric stellar distribution, it does not show an unusual gas phase metallicity gradient or shock ionization. On the other hand, we probed the SFE at sub-kpc scales (0.5 kpc) across NGC 2276’s disk to trace the origin of its elevated and asymmetric SFR and found more than two orders of magnitude variation in SFE. This is significantly larger than what is seen in nearby galaxies. We speculate hat this is caused by both tidal forces exerted from a neighboring galaxy and ram pressure affecting NGC 2276.

Supervisor:  Eva Schinnerer   (MPIA)

Anna-Christina Eilers    (Germany)                                                                                                                                         03.05.2019

Unravelling 13 billion years of cosmic history with spectroscopic studies: from the Milky Way to the epoch of reionization  (thesis pdf, 40MB)

In this thesis we trace the cosmic history from the Epoch of Reionization to the local universe by means of several spectroscopic studies. In the first part, we analyze 34 quasar spectra at 5.8 < zem < 6.5 and measure the redshift evolution of the opacity of the intergalactic medium (IGM) within the Lyα as well as the Lyβ forest to set new constraints on the timing and morphology of the reionization epoch. We find evidence for an extended reionization process down to z ∼ 5, and, while the observed scatter in the Lyα forest optical depth can be well reproduced by current state-of-the-art simulations including spatial fluctuations in the temperature field or the ultraviolet background, we find a strong mismatch between simulations and observations in the Lyβ forest opacity, suggesting an inversion of the thermal state of the post-reionization IGM.
We also measure the sizes of the quasars’ proximity zones, which are regions of enhanced ionization in the vicinity of the quasars, ionized by their own radiation. We find a dependency of the proximity zone sizes to the quasars’ lifetime, whicpresents a novel method to estimate the lifetime of individual quasars, providing unprecedented constraints on the formation and growth of supermassive black holes in the early universe. We discover three quasars with very short lifetimes, i.e. t_Q ∼ 10^4 − 10^5 yr, that pose significant challenges to all current black hole formation theories.
In the second part of this thesis we explore the structure and dynamics of the Milky Way. We develop a new data-driven model to determine precise parallaxes by combining multi-band photometry and spectroscopy to make global kinematic maps of our Galaxy from & 45, 000 luminous red giant stars with only . 10% parallax uncertainties. Our map extends to Galactocentric distances of 25 kpc, well beyond the reach of parallax estimates by the Gaia mission. Making use of these new spectrophotometric parallaxes, we determine the most precise measurement to date of the circular velocity curve of the Milky Way over a wide range of Galactocentric distances. Based on Jeans modeling in an axisymmetric gravitational potential we find that the velocity curve is gently but significantly declining.

Supervisor:  Joe Hennawi   (MPIA/UCSB)

Vikas Josji    (India)                                                                                                                                                                  06.02.2019

Reconstruction and Analysis of Highest Energy γ-Rays and its Application to Pulsar Wind Nebulae      ( thesis pdf, 35 MB )

The High Altitude Water Cherenkov (HAWC) γ-ray observatory detects cosmic- and γ-rays in the TeV energy range. HAWC was recently upgraded with a sparse detector array (the outrigger array), which increases the instrumented area by a factor of 4-5 and will improve the sensitivity at energies greater than 10 TeV. This thesis consists of a number of contributions towards the improvement of the performance of HAWC at the highest energies and the study of a prominent high energy source, 2HWC J2019+367. To decide on components of the outrigger array, simulation input is provided. A new Monte Carlo template-based reconstruction method for air shower arrays is developed. It reconstructs the core location and energy of γ-ray showers. The goodness of fit of the method is utilised to separate the cosmic- and γ-ray showers. This method significantly improves the HAWC shower reconstruction and combines the reconstruction of HAWC and the outrigger array. In-depth spectral and morphological studies of 2HWC J2019+367 are performed. 2HWC J2019+367 shows a hint of energy-dependent morphology. A new HAWC source is discovered in the vicinity associated with VER J2016+371. The preferred direction of the X-ray and TeV emission indicates their association, and their combined spectral modelling show that 2HWC J2019+367 is likely to be the TeV pulsar wind nebula of PSR J2021+3651.

Supervisor:  Jim Hinton   (MPIK)

Tobias Schmidt     (Germany)                                                                                                                                                      05.02.2019

Constraints on Quasar Emission Properties from the HeII and HI Transverse Proximity Effect      ( thesis pdf, 10 MB )

Bright quasars are powerful sources of ionizing radiation and have profound impact on the Intergalactic Medium. In particular, they create regions with enhanced ionization and therefore reduced Lyman α forest absorption in their surroundings. Observing this so-called transverse proximity effect along background sightlines provides a view of the foreground quasar from different vantage points, and hence at different lookback times compared to the line-of-sight toward Earth. One can thus constrain the emission history (lifetime, age) and emission geometry (obscuration, opening angle) of the foreground quasar based purely on geometric and light travel time arguments. Both quantities are so far poorly constrained by observations but fundamental for the understanding of Active Galactic Nuclei. To investigate the HeII transverse proximity effect, we conducted an optical spectroscopic foreground quasar survey around 22 HST/COS sightlines, leading to a sample of 20 foreground quasars. We find statistical evidence for the the HeII transverse proximity effect and infer a constraint on the quasar lifetime of > 25 Myr. From a detailed modeling, based on cosmological hydrodynamical simulations and a dedicated photoionization model including quasar obscuration and finite quasar lifetime, we derive joint constraints on age and obscuration of individual objects, indicating that one quasar is old and unobscured (tage ≈ 25 Myr, Ωobsc < 30 %) while three other are either young (tage < 10 Myr) or highly obscured (Ωobsc > 70 %). However, the models also reveal that the large scatter intrinsic to the HeII Lyα forest prohibits further progress in the field. I therefore developed a novel method that uses large numbers of HI Lyα forest spectra to map the 3D light echo of individual quasars. An end-to-end test confirms that such tomographic observations can constrain the age of hyperluminous quasars to better than 20%, requiring only 1 – 2 nights on existing 8 – 10 m facilities. The method bears potential to also constrain the quasar emission geometry and the full lightcurve over the past 100 Myr, rendering it a viable tool to investigate quasar properties.

Supervisor:  Joseph Hennawi   (MPIA/ UCSB)

Antonio D'Isanto   (Italy)                                                                                                                                                          31.01.2019

Probabilistic photometric redshift estimation in massive digital sky surveys via machine learning      ( thesis pdf, 15 MB )

The problem of photometric redshift estimation is a major subject in astronomy, since the need of estimating distances for a huge number of sources, as required by the data deluge of the recent years. The ability to estimate redshifts through spectroscopy does not scale with this avalanche of data. Photometric redshifts provide the required redshift estimates at the cost of some precision. The success of several forthcoming missions is highly dependent on the availability of photometric redshifts. The purpose of this thesis is to provide innovative methods for photometric redshift estimation. Two models are proposed. The first is fully-automatized, based on the combination of a convolutional neural network with a mixture density network, to predict probabilistic multimodal redshifts directly from images. The second model is features-based, performing a massive combination of photometric parameters to apply a forward selection in a huge feature space. The proposed models perform very efficiently compared to some of the most common models used in the literature. An important part of the work is dedicated to the correct estimation of the errors and prediction quality. The proposed models are very general and can be applied to different topics in astronomy and beyond.

Supervisor:  Kai Polsterer  (HITS)

Katharina M. J. Wollenberg   (Germany)                                                                                                                                 16.01.2019

Diversity of Population III Star Formation :  influence of rotation, turbulence, and magnetic fields on the fragmentation behavior of Population III protostellar disks and implications for the later evolution of the star-forming environment  ( thesis pdf, 120 MB,  corrigendum )

The aim of this thesis is to improve our understanding of the fragmentation behavior of Population III protostellar disks under the influence of rotation, turbulence, and magnetic fields. We further evaluate consequences that may be inferred for the later evolution of the star-forming halo and its surroundings with respect to protostellar ejections and in terms of the impact of radiative feedback on later chemical enrichment of neighboring halos. In the main part of this thesis, we follow the collapse of a primordial gas cloud until the formation of the first protostar and the creation of a highly gravitationally unstable protostellar disk system. We find that turbulence promotes the fragmentation of the protostellar disk and both rotation and magnetic fields can provide some stabilization against it. While the total mass growth of the collection of protostars is only mildly affected by rotation and turbulence, magnetic fields can have such a strong impact on the dynamical evolution of the disk system that accretion onto the protostars is highly disturbed and their mass growth is significantly reduced. In spite of all the differences, the disk generally breaks up into a protostellar cluster that develops a top-heavy mass function. Interactions between protostars in the cluster are highly dynamical and lead to a considerable number of protostellar ejections. We demonstrate that some of these ejected Population III protostars, even if they continue to accrete for some longer period after they have left the disk environment, continue to have masses of M < 0.8 solar mass. Hence, they have lifetimes longer than the current age of the Universe and thus describe Population III candidates that could still be observable today. In another project, we assess the role of photoevaporation of a pristine halo by a near-by Population III star prior to the supernova explosion of that star. We demonstrate that it is crucial for realistic simulations of metal enrichment to account for the photoevaporation as the radiation ablates and thins out the outer halo layers and thus makes the halo more susceptible to mixing with the metals from the supernova ejecta. In this thesis, we use both analytical estimations and numerical simulations. Further tests are conducted to investigate the performance of our numerical methods and the sensitivity of our results to the numerical resolution. We demonstrate that general trends, in particular when effects of turbulence are examined, cannot be reliably deduced from only a single numerical run. Instead a statistical analysis of an ensemble of realizations based on the same initial conditions needs to be considered.

Supervisor:  Simon Glover (ITA)