I once did science professionally…
Most of the information here assumes some kind of knowledge already. I’m sorry about that for those who do not fully understand what I’ve written down. If you have any question, please don’t hesitate to ask me! It is mostly in negative chronological order: newest first! Most of this should be somewhere in my publications list too.
Self-regulated galaxy formation
From the simulations described below, and work by others, insights have emerged about how galaxies form and evolve.
The build-up of stellar mass inside them is regulated on the large scales, through a balance of inflow of gas (which is determined by the cosmology and the cooling properties of gas) and outflows (which are the result of massive stars and Active Galactic Nuclei).
This implies that what happens deep inside the dense gaseous regions of the galaxy, including the star formation process itself (!) is to a large extent irrelevant for the star formation properties of the galaxy as a whole.
There are some important implications of this picture. Even though self-regulation is becoming widely accepted, these implications are often overlooked. As such, observations and simulations are often interpreted erroneously.
Translating simulations into observations
In the work started at STScI I mainly focus on making mock observations from different inds of simulations. We use semi-analytic
models as well as hydrodynamic simulations. We simulate deep images and analyze these as if they were real images to find back the properties of the
galaxies on the images. We then compare the resulting galaxy population with the galaxy population in the input models and investigate how different
they are and why. This is very useful when planning future observations (e.g. with JWST) and for interpreting current observations of Hubble and other
The environment of galaxies and their host dark matter halos
For now, just a link to the website corresponding to a paper.
Stochastic star formation and its influence on galaxies
Most, or at least many, stars are formed in clusters. Stars in star forming regions and in star clusters follow a certain mass distribution that
is fairly well known. As stars of different masses create different elements and put different amounts of energy in the gas surrounding them, the
precise, galaxy wide distribution of stellar masses is an important quantity for galaxy evolution. Now the clusters in which these stars form themselves
follow a mass distribution that strongly favours low mass clusters. In the lowest mass clusters the largest possible stellar masses cannot exist (the
most massive star would be more massive than its hosting cluster!), so once all stars in all clusters are taken together, there are fewer massive
stars than you would guess from the (universal) mass distribution in the star clusters.
The OverWhelmingly Large Simulations Project
The project for my PhD is part of the OverWhelmingly Large Simulations project. OWLS consists of a large suite of cosmological N-body/SPH simulations, using a modified version of Volker Springel’s Gadget code. The power of the projects lies in the large variation of subgrid models for the unresolved physics. For the non-experts: simulations have a limited mass resolution (and a limited spatial resolution). Therefore, many processes that are important in the formation of galaxies (e.g. forming stars, growing a supermassive black hole in the center, exploding supernovae, …) can not be followed explicitly. Therefore one makes assumptions about how these processes work out on the larger scales that can be resolved. These models are called subgrid models, and the most important aspect of OWLS is the large variation in these models.
In particular, the following parameters are varied:
- Size of the box (mainly 25 and 100 Mpc/h, comoving)
- Mass resolution
- Cosmological parameters
- Star formation law
- The effective equation of state for high density gas
- The way supernova feedback works
- The way (and also whether at all) supermassive black holes grow and feed energy back into the galaxy
- The cooling function of the gas
- The reionization of the universe at high redshift
My project within OWLS: The Galaxies
My project focused mainly on the galaxies that form inside a the simulation box, a beautiful example of which you see depicted above. Selecting these galaxies automatically is not as trivial as it may seem, as a computer cannot look at such images as we can.
A part of my topic is to select groups of particles that make up something together that an observer might call a galaxy. A comparison of several of these methods, from very easy (by linking particles that are close enough together), along physically motivated (group together gravitationally bound structures) all the way to creating mock observations of the simulation (and selecting galaxies with the tools observers would use for the same purpose) is an important part of my project.
Once the galaxies are identified I try to explain their physical properties (total mass, mass in stars, ages, metallicity, star formation rate, luminosity, etc…) and relate them to the input physics of the specific run. Example questions are “How is the star formation rate of a galaxy related to the wind velocity due to supernova explosions?” and “What kind of processes do we need to explain the observed, red and dead (no star formation) elliptical galaxies?”
Getting observables (magnitude, color, reddening due to dust) for galaxies is also something of my concern. I implemented an adapted version of the Bruzual and Charlot (2003) population synthesis code to determine the luminosity of the stellar component of the simulated universe. This light then travels towards us and encounters gas and dust. What this gas and dust do to the light is vaguely known and I am now trying to consistently model effects, in order to come up with realistic luminosities and colors of the galaxies.
The full thesis is here, too!
Master’s Thesis: Star Clusters in M51
In my MSc project we investigated several aspects of the star clusters in M51, mainly related to their luminosities and
radii. Broadband imaging observations (in this case HST/ACS, B,V and I imaging) on a small wavelength range are not
sufficient to in detail derive ages, masses, extinctions and metallicity for all clusters separately. Therefore, it may be wise to only look at the statistics of the population, like radius distributions, luminosity functions and correlations between radius and magnitude (which both can be derived from the data, of course). So, that’s what we did and some rather nice results followed.
In my fourth year I followed a course on General Relativity, of which this report on lensing by cosmic strings marked the end.
The only course more or less related to my later work was about galaxies. A report on synthetic spectral modeling of galaxies resulted from this.
The course on stellar mass loss culminated in an abstract about the research of the influence of stellar mass loss on the evolution of star clusters.
A report on the calculation of the velocity distribution of particles in gasses at extremely high temperatures, which I’ve written during my third year.
A report on the analysis of the X-ray spectrum of an AGN, using SPEX. Also third year stuff.
A text on the numerical integration of partial differential equations from my second year.
A text on some methods used in observational astrophysics, belonging to the corresponding second year course.