Alumni 2024
Brooke Polak (USA) 17.10.2024
The Secret Lives of Young Massive Star Clusters ( thesis pdf, 40 MB)
Many aspects of star cluster formation and evolution are unconstrained. This thesis presents models of star cluster formation — including gas and stellar dynamics, sub-grid star formation, stellar evolution, and stellar feedback — from turbulent gas clouds of initial mass 10⁴, 10⁵, and 10⁶ M⊙. The latter is one of the largest star-by-star cluster models to date and has provided many new insights into the formation and early evolutionary properties of young massive star clusters. First, I examine the global properties of each cloud, particularly noting the high star formation efficiency. In sufficiently dense and massive clus- ters, stellar feedback is unable to quench star formation as gravity overpowers it. Next, I identify a novel mechanism for the production of runaway stars in young clusters, the sub-cluster ejection scenario, in which a subset of stars in an infalling sub-cluster are ejected by a tidal interaction with the assembling cluster’s center of mass. Lastly, I analyze the presence of dynamical mass segregation in the models, where massive stars are more centrally clustered than low-mass stars. These models have no primordial mass segregation by construction. Young clusters can undergo early dynamical mass segregation during core collapse when the crossing time is substantially reduced.
Supervisor: Ralf Klessen (ITA)
Evert Nasedkin (Canada) 24.07.2024
Atmospheric Characterisation of Directly Imaged Exoplanets ( thesis pdf, 25 MB)
As long as we have gazed at the night sky we have wondered about the nature of planets both within our Solar System and beyond. The last three decades of observations have revealed a staggering diversity of exoplanets in terms of their size, temperature, and composition. Out of the thousands of known exoplanets, only a small handful can be directly imaged. These rare systems present a unique opportunity to characterise their atmospheres with high precision and broad wavelength coverage. In order to observe these planets, they must retain enough heat from their formation to be observable in thermal emission, using dedicated high-contrast imaging techniques to separate the faint signal of the companion from the nearby star which it orbits. Only young, giant exoplanets are bright enough to be directly observable. Their relative youth enables us to tie their present-day composition back to the mechanism via which they formed.
In this thesis, I present the tools and methods used to observe and characterise such directly imaged exoplanets and apply them to the benchmark system HR 8799. The development of the petitRADTRANS retrieval module allows for the rapid fitting of models to spectroscopic data in a Bayesian framework. This retrieval method is applied to the JWST Early Release Science target WASP-39 b, a hot, Saturn mass exoplanet. WASP-39 b is found to have an atmosphere enriched in metals, and the detections of CO2} and photochemically produced SO2 are confirmed. By comparing three different high-contrast imaging post-processing algorithms, the impacts of data processing techniques used to extract an exoplanet's spectrum on one's ability to infer atmospheric properties are explored. Accounting for the correlation between wavelength channels of spectroscopic measurements is found to be critical to producing unbiased parameter estimates. Having developed the data analysis and modelling framework, the atmospheres of the four HR 8799 planets are systematically characterised. Using new VLTI/GRAVITY observations, together with a broad range of archival data we compile the most complete spectra of these objects to date. The planets are fit using the petitRADTRANS retrieval framework, as well as using self-consistent radiative-convective equilibrium grids. Bulk properties of effective temperature, surface gravity, radius, metallicity, carbon-to-oxygen number ratio, mass, and bolometric luminosity are inferred for all four planets. Their atmospheres are found to be highly metal-rich, with stellar-to-superstellar C/O ratios. Future study will be necessary to link these atmospheric properties to their formation history, but it is clear that these enigmatic planets will remain a target of further observations for years to come.
Supervisor: Laura Kreidberg / Paul Molliere (MPIA)
Glen Hunter (UK) 24.07.2024
The consequence of dynamics in the interstellar medium ( thesis pdf, 10 MB)
The interstellar medium (ISM) of the Milky Way is in a constant state of motion. Between the large scale motions of galactic dynamics and the turbulence within giant molecular clouds (GMCs), the evolution of the ISM is guided by its underlying motions. The large scale dynamics perturb the diffuse warm neutral medium (WNM) of the Milky Way, starting the cascade of the formation of clouds and stars. In this thesis we make use of hydrodynamical simulations to explore the impact dynamics has on various quantities. The first part of this thesis focuses on the collisions of giant molecular clouds, exploring how star formation rates change with collisions. We verify the stabilitiy of “clumps” within our simulations and relating these properties to the alignment of the magnetic and velocity fields within our simulations. The second part of this thesis zooms out, looking at the structure of the Milky Way itself. We present a new, realistic model for the gravitational potential of the Milky Way and test out how reliable the axisymmetric assumption is when used in kinematic distance estimates.
Supervisor: Simon Glover (ITA)
Diego Sotillo Ramos (Spain) 19.07.2024
Milky Way and M31 analogues: insights from the cosmological simulation TNG50 ( thesis pdf, 22 MB)
In this thesis, I study key facets of Milky Way- and Andromeda-like (MW/M31-like) galaxies’ formation, evolution and structure in the cosmological context. To this aim, I use TNG50, the highest resolution run of the cutting-edge state-of-the-art IllustrisTNG suite of cosmological mangeto-hydrodynamical simulations. TNG50 simulates 198 MW/M31 analogs, in an unprecedented blend of high numerical resolution and sample size. I study the impact of major mergers on galaxy stellar disks, revealing that a significant fraction of the analyzed galaxies undergo a recent major merger (in the last 5 billion years) and are still disky at z = 0. Among these galaxies, for two-thirds of the cases, the merger destroys the disk, but a new one is able to form until z = 0. In the remaining galaxies, the disk survives the merger. I analyze and quantify stellar disk flaring, i.e., the increase of vertical stellar disk height with galactocentric distance, showing a great diversity of types and values across the galaxies of the sample. But it is a complex phenomenon, difficult to predict for an individual galaxy according to its z = 0 global structural properties or merger history. Finally, I also investigate the presence of very metal-poor stars in the stellar disks and other morphological components ofMW-like galaxies, yielding valuable insights into their origin and development: according to TNG50, there is a non-negligible fraction of these stars that populate the disk, and that are also very old, challenging therefore the notion that the stellar halo is the oldest component of the MW. Therefore, this thesis leverages the power of numerical tools like TNG50 to uncover pivotal aspects of the evolution and structure of MW/M31-like galaxies as well as their formation, providing valuable insights that should be tested with future observational studies.
Supervisor: Annalisa Pillepich (MPIA)
Wuji Wang (China) 18.07.2024
Three-dimensional view of circumgalactic to interstellar medium around distant radio galaxies ( thesis pdf, 40 MB)
High-redshift radio galaxies (HzRGs, z ≳ 2) represent a unique population of luminous active galactic nuclei (AGN), enabling the simultaneous examination of jet-mode and radiative-mode feedback in massive galaxies. This thesis focuses on the gaseous medium from circumgalactic to interstellar medium (CGM to ISM) scales around the most well-observed HzRGs sample, leveraging multiwavelength 3D information from integral field spectrographs. The lives of galaxies are shaped both by internal processes and their environment. The ISM and CGM are thus key components in galaxy evolution as sites of energy exchange, chemical enrichment and material cycling. Using MUSE, I conducted the first systematic analysis of ∼ 100 kpc intrinsic Lyα nebulae (corrected for H I absorption) for the eight HzRGs. Our characterization of CGM gas properties enables a cross-comparison to different quasar species. The ISM within HzRGs however has remained poorly understood prior to the JWST era. I have analyzed the warm ionized gas on galactic scales using JWST/NIRSpec IFU with sub-kpc resolution, revealing inefficient radiatively driven outflows in one HzRG, potentially implying that jet-mode feedback dominates. Overall, this thesis initiates the 3D analysis of distant AGN, with legacy datasets spanning from the rest frame UV to far infrared (e.g., ALMA).
Supervisor: Dominika Wylezalek (ARI)
Lukas Eisert (Germany) 18.07.2024
Inferring the assembly and merger histories of galaxies with the IllustrisTNG simulations and machine learning ( thesis pdf, 40 MB)
This thesis presents an investigation into galaxy formation and evolution, utilizing cutting-edge cosmological simulations and machine learning methodologies. Galaxy data from the full cosmological simulations TNG50 and TNG100 within the IllustrisTNG project are employed, and machine learning techniques are trained to extract the assembly and merging history of these simulated galaxies. These machine-learning models can then be applied to observational data, offering a novel method to connect simulations and observations. Initially, this is achieved by using scalars representing integrated observable galaxy features, from which we are able to infer scalars describing the assembly and merging history accurately. However, scalars encompass only a fraction of the complete observational data (images, spectra, IFUs, etc.), and accurately reconstructing scalars from simulations with the same observable bias as observations is not trivial. To address this, contrastive learning is employed to perform representation learning on both survey-realistic mocks of TNG and observed Hyper Suprime-Cam (HSC) image data (in r, g, and i bands), ensuring comparability between simulated and observed images. Remarkably, our findings demonstrate sufficient similarity between the simulated and observed images to justify the idea of simulation-based inference with images. Subsequently, an inference model trained on TNG data is successfully applied to HSC data, allowing the retrieval of information regarding the ex-situ fraction and the time and mass of the last major merger undergone by a galaxy. This interdisciplinary approach merges the domains of cosmological simulations, observational astronomy, and machine learning, offering a new perspective on galaxy formation and evolution. The developed methodologies not only enhance our ability to interpret observational data but also enable the assessment of the realism of cosmological simulations.
Supervisor: Annalisa Pillepich (MPIA)
Theodora Xylakis-Dornbusch (Greece) 09.07.2024
Investigation of the Galactic chemical enrichment history with searches for and chemical abundance analysis of metal-poor stars ( thesis pdf, 40 MB)
This PhD dissertation presents a method for the identification of metal-poor stars with Gaia BP/RP spectra. The metal-poor star selection method is based on flux ra- tios and was developed with Gaia BP/RP simulated synthetic spectra. In follow-up work the selection method was updated and applied to Gaia DR3 BP/RP spectra with E(B - V)<=1.5. Furthermore, 26 metal-poor candidates were selected for ob- servations, of which 100% had [Fe/H] < -2.0, 57% had [Fe/H] < -2.5, and 8% had [Fe/H] < -3.0. Finally, a catalog of stellar metallicities for 10 861 062 stars was assembled. Moreover, a kinematic analysis and a 1D LTE abundance analysis of limited-r stars was conducted. In addition, the lanthanide mass fractions (XLa) of all the to date known limited-r stars were calculated and compared to that of the KN AT2017gfo. The results showed that the abundance patterns of the neutron-capture elements of limited-r stars are different depending on whether [Ba/Eu] is below or above -0.3. Also, the XLa of the KN was found to be higher than the XLa’s of the limited-r stars and in the transition region between the latter and the XLa’s of the r-I and r-II stars. Finally, the current sample of limited-r stars is largely born in the Galaxy rather than being accreted.
Supervisor: Norbert Christlieb (LSW)
Maria Selina Nitschai (Germany) 02.07.2024
Dynamics of the Milky Way Disk and Spectroscopic Analysis of ω Centauri ( thesis pdf, 40 MB)
The Milky Way is the perfect laboratory to study galaxy evolution and formation due to our unique position inside its disk. Recent surveys and instruments can provide an extensive amount of data for our Galaxy that are the key to revealing its assembly history. In this thesis, I first combine Gaia EDR3 and APOGEE data throughout Galactocentric radii between 5.0 ≤ R ≤ 19.5 kpc and construct a dynamical model. Using the spherically-aligned Jeans Anisotropic Method I model the stellar velocities and the velocity dispersions of the Galactic disk. This model can capture the main kinematic features and give an accurate mass density of the Galaxy, making it a fundamental test for both galaxy dynamics in general and the mass distribution of the Milky Way. Further, I focus on Omega Centauri (ω Cen), the most massive globular cluster in the Milky Way, which has long been suspected to be the stripped nucleus of a dwarf galaxy that fell into the Galaxy a long time ago. This merger event was the last significant merger the Galaxy has experienced and is therefore an important event in its evolution. Firstly, I present a MUSE spectroscopic dataset with more than 300,000 extracted stellar spectra, reaching more than two magnitudes below the main sequence turn-off. This massive spectroscopic dataset will enable future studies that will transform our understanding of ω Cen. Secondly, I investigate the underlying metallicity distributions as well as the spatial variations of the populations within the cluster for the red giant branch stars. For that, I combine the new MUSE spectroscopy with new HST photometry and show that there appears to be no gradient in metallicity, indicating that the cluster is well mixed and any merge happened many years ago. Finally, I conclude with future studies that will investigate the stellar populations, ages, abundances, kinematics, and dynamics of ω Cen in great detail.
Supervisor: Nadine Neumayer (MPIA)
Zhiqiu Huang (China) 23.05.2024
Gamma-ray Bursts and Implications for Particle Acceleration at Ultra-relativistic Shocks ( thesis pdf, 20 MB)
Recent TeV detections of gamma-ray burst afterglows offer new insights into particle acceleration at relativistic shocks. Kinetic simulations have improved our understanding of shock microphysics, enhancing models of particle acceleration relevant to afterglows. We explore scenarios for determining the maximum achievable energy, comparing our findings with data from the H.E.S.S. source, GRB 190829A. This comparison reveals a tension between observations and theoretical expectations. Motivated by this, we developed a Monte Carlo code to revisit acceleration theory for relativistic shocks in uniform and non-uniform magnetic field configurations. In uniform fields, we demonstrate that acceleration requires only strong scattering on one side of the shock. Analytic solutions confirm this conclusion. For non-uniform fields, we consider a cylindrical magnetic-field structure typical of astrophysical jets. We find that curvature drifts enable repeated shock-crossings for particles of favourable charge, and neglecting losses extends the maximum energy to the system's confinement limit. These results challenge the misconception that ultra-relativistic shocks cannot serve as effective accelerators, offering a fresh perspective on relativistic shock acceleration. The findings suggest new features on maximum achievable energy and spectral index, indicating the need to revisit current knowledge on relativistic shocks. This could open promising avenues for producing ultra-high energy cosmic rays.
Supervisor: Brian Reville (MPIK)
Javier Moran Fraile (Spain) 30.04.2024
Simulating the dynamical interaction of white dwarf stars in binaries ( thesis pdf, 25 MB)
Binary stellar interactions that take place on dynamical timescales are some of the most challenging processes to model in astronomy, and are best described by multidimensional, multi-physics simulations. The focus of this thesis is on the numerical modeling of some of the least explored stellar interactions, with a spe- cial emphasis on those involving white dwarfs. This work presents three different studies using three-dimensional hydrodynamic simulations. In the first place, the emission of gravitational waves during common-envelope events is studied, esti- mating the chances for their detection with future space-based detectors. Secondly, the tidal disruption of a white dwarf by a neutron star is studied, showing how the accurate modeling of these events requires the inclusion of a multitude of physi- cal processes including magnetic fields, nuclear reactions and neutrino emission. Finally, through a third simulation, it is shown how mergers between low-mass white dwarfs, with total masses substantially below the Chandrasekhar limit, can lead to thermonuclear explosions under the right conditions. The results of this thesis stress how dynamical interactions between stars can produce a multitude of bright transients, and how the use of advanced multidimensional, multi-physics codes for their modeling will help improve our understanding of the physics and processes involved.
Supervisor: Friedrich Roepke (HITS)
Toni Peter (Germany) 23.04.2024
Understanding the Era of Reionization via Numerical Methods for Radiative Transfer ( thesis pdf, 22 MB)
The goal of this thesis are numerical studies of the era of reionization, which took place at about a 150 million years to a billion years after the Big Bang. Since reionization is a process driven by radiation, a major fraction of this work is dedicated to numerical methods for radiative transfer. In particular, we develop the Sweep method, which allows us to study reionization within large cosmological simulations. We begin by in- troducing the basics of the Sweep algorithm and its implementation in the simulation code Arepo. We discuss the motivation behind it, how it integrates with the rest of Arepo and perform a number of tests to assess its performance and physical accuracy. We find that the Sweep method does not only produce physically accurate results, but does so in a very efficient manner, even when applied to large simulations on a large number of processors. We then proceed by introducing the standalone radiative trans- fer postprocessing code Subsweep in which we add a variety of improvements to the original Sweep method, in particular the addition of sub-timesteps. We perform a number of additional tests to verify that Subsweep correctly solves a number of physi- cal problems and show that sub-timesteps can drastically improve performance of the Sweep algorithm when applied to problems with heterogeneous environments without sacrificing accuracy. Finally, we apply Subsweep to the cosmological simulation suite TNG in order to recreate the era of reionization in the TNG universe. We find that Subsweep allows us to study the spatial structure of reionization in detail and that we can reproduce the observational constraints on the history of reionization reasonably well.
Supervisor: Ralf Klessen (ITA)
Bastian Alejandro Reinoso Reinoso (Chile) 15.04.2024
Formation of massive black hole seeds through runaway stellar collisions and gas accretion in dense stellar systems ( thesis pdf, 30 MB)
The goal of this work is to gain a better understanding of the processes that lead to the formation of massive black hole seeds in the early Universe, in order to provide insights into the rapid emergence of the highest redshift quasars. Two different seeding mechanisms were studied via numerical simulations. The first mechanism explores the onset of runaway stellar collisions in dense clusters of Population III stars, focusing on understanding the role of an external potential for modelling the gas during the embedded phase. Stellar collision rates were also explored in a similar environment with the goal of confronting analytic estimates with numerical simulations. The study of this seeding mechanism demonstrates the plausibility of forming black hole seeds with > 1000 M⊙ through runaway stellar collisions that produce very massive stars. Furthermore, an analytic model for estimating the number of collisions in dense star clusters is presented, and the identification of a new collision channel involving perturbations on binary stars is reported. The second seeding mechanism explored in this work deals with the emergence of supermassive stars through the interplay of gas accretion and stellar collisions in environments resembling collapsed gas clouds in atomic cooling halos. The numerical implementation developed in this work allowed for a self-consistent treatment of stellar and gas dynamics for the exploration of this mechanism. The results show that the emergence of supermassive stars with 10^4 M⊙ is inevitable and a binary system of supermassive stars is the outcome in one third of the cases. This thesis concludes by summarizing and discussing the results found in these studies and commenting on the future work needed to improve the models presented here.
Supervisor: Ralf Klessen (ITA)
Francesco Conte (Italy) 01.02.2024
Gamma-ray emission and absorption in the inner few parsecs of the Galactic Centre ( thesis pdf, 15 MB)
Located 8 kpc away, the Galactic Centre is a rich environment for observing non-thermal phenomena such as a supermassive black hole, potential dark matter accumulations, supernova remnants, pulsar wind nebulae, clustered massive stars, and many more. It is a key target for both operational and next-generation TeV observatories like H.E.S.S., MAGIC, and CTA. Current telescopes, limited by a full-width half-maximum of 5 arcminutes, struggle to pinpoint the nature of gamma-ray sources amidst the Galactic Centre’s complexity. UV/visible observations are also compromised due to dust absorption and infrared re-emission. However, this study leverages the infrared radiation’s ability to absorb high-energy photons, using a model of the infrared field for spatial and spectral gamma-ray analyses. In this thesis I present the first 3D model for the infrared radiation field in the inner few parsecs. By studying the high-energy absorption, I find that if the central gamma-ray source and the large scale gamma-ray emission share the same cosmic-ray accelerator, then the central emitter is a ring of outer radius 2.5 pc that CTA will see as an extended source. In that case, the diffuse gamma-ray emission is expected to show a turn-off around 20 TeV rather than a power-law to 100 TeV.
Supervisor: Richard Tuffs (MPIK)