About Me
Researcher on the intersection of astronomy, statistics and computer science.
Astrophysicist, Statistician, Computer Scientist
In Astrophysics, I research growing black holes, their demographics and their immediate environments. Before matter falls into a black hole, it swirls around it and heats up, creating radiation. This makes distant galaxies shine bright as Active Galactic Nuclei (AGN). I observe this radiation because it tells us how much the black hole is currently growing.
My research focuses on the time when this growth occured most (z=0.5-3). For my PhD project I reconstructed the total growth of black holes over cosmic time using a large sample of distant AGN (2000, CDFS, COSMOS, AEGIS, XMM-XXL). This also requires a good understanding of the observations (astro-statistics) and the obscuration of AGN.
In most AGN, much of the radiation is swallowed by thick columns of gas and dust near the black hole. In my research I try to understand these gas clouds, in particular their location, extent/covering and relation to the black hole. During my PhD I investigated different obscurer geometries. I could also place the best constraints to date on the intrinsic covering fraction of the obscurer (77% Compton-thin, 38% Compton-thick obscured). The large-scale gas in galaxies can also obscure — in two recent papers I constrained how important this effect is. To understand the nuclear obscurer is crucial to correctly infer the intrinsic emission and therefore the black hole growth. Also, the mechanisms making these clouds is currently unknown.
I have published in Statistics, where I focus on nested sampling Monte Carlo algorithms and their performance. Population studies (hierarchical Bayesian inference) interests me, as well as helping others with statistics problems.
I write a lot of software for various purposes (>100 github repos), many of them are also used by others: I am the author of the PyMultiNest package, and the Bayesian X-ray Astronomy (BXA) code. I think daily about new algorithms and solutions.
Since I begun publishing in 2014 my papers have received more than 100 citations. You can find a full list on ADS. Here I mention some aspects for each of the papers:
Year, Title, Authors | Astronomy aspects | Statistics aspects | Link |
---|---|---|---|
Buchner, Schulze & Bauer (2017): Galaxy gas as obscurer: I. GRBs x-ray galaxies and find a N_{H} ~ M* relation | We determined that the obscurer of long Gamma-ray Bursts (GRBs)
is their host galaxy, by analysing the X-ray spectrum of ~1000 GRBs
with better methods, correlation with host stellar mass,
and considering cosmological hydrodynamic simulations.
Supplimentary material: Video explanation • Accessible write-up of this study • Catalogue of Swift afterglows and their obscurations (FITS). | Inference about the properties of a population. Model comparison. Reconstructs a 3d shape from random probes (tomography). Analysis of the gas in Illustris cosmological simulations.. | |
Buchner & Bauer (2017): Galaxy gas as obscurer: II. Separating the galaxy-scale and nuclear obscurers of Active Galactic Nuclei | What fraction of AGN would be obscured if there
was no torus? 40%!
GRBs tell us how much gas there is in galaxies.
The host galaxies of GRBs and AGNs are slightly different,
but this can be bridged as we demonstrate in this paper.
The nuclear obscurer is then described in a new model,
the radiation-lifted torus, in which Eddington accretion
rate is the driving force of the changing obscurer.
Supplimentary material: Video explanation • Accessible write-up of this and the previous study. | Analysis of large cosmological simulations (EAGLE), non-parametric Kernel-density estimates. | |
Buchner et al. (2015): Obscuration-dependent evolution of Active Galactic Nuclei | This luminosity-function type study reconstructs the distribution and evolution of AGN in obscuration and luminosity using a novel robust non-parametric approach. We constrain the importance of Compton-thick AGN to the accretion history of the Universe, and the evolution of the obscured fraction.
Supplimentary material: Video explanation • XLF as table for download: Space density of AGN as f(Lx, z, NH) • Total Space density of AGN as f(Lx, z) • Plot for comparison. | Cencored inference about the properties of a population. Bayesian field inference. Reconstructs a 3d smooth function under selection effects without assuming a shape but only smoothness. Uses Stan to reconstruct the growth of black holes over cosmic time. | |
Buchner et al. (2014): X-ray spectral modelling of the AGN obscuring region in the CDFS: Bayesian model selection and catalogue | Comparison of various models for the obscurer in AGN. Through model comparison between a disk, sphere and toroidal geometry, with the latter preferred, the obscurer was found to be extended but not fully covering, even for the Compton-thick sub-population.
Supplimentary material: Vizier CDFS catalogue • BXA documentation | Presents several advancement in X-ray spectral analysis methods: Bayesian parametric analysis, comparison of models, Goodness-of-Fit, nested sampling vs. MCMC, vs likelihood contour error estimation | |
Buchner (2014): A statistical test for Nested Sampling algorithms | Statistics paper: Evaluation of MultiNest and similar algorithms
Supplimentary material: Code for RadFriends and UltraNest on Github. | Analyses several nested sampling algorithms (e.g. MultiNest) for flaws using a new statistical test. |
I am an active member on the Astrostatistics Facebook group, where we answer Astronomers questions about statistics, data mining, machine learning, programming, etc. I regularly answer questions, review papers on their use of statistics and write mini-tutorials. Similarly, I help out my colleagues with statistics questions in our astronomy institute.
I have written a minimal statistics checklist to help you identify and fix common errors/misinterpretation in your analysis, or of a paper you are refereeing.
You can find my statistics software and papers in the previous and next sections.
I write software to make my life and the life of my colleagues easier. Perhaps you can take advantage of it too:
Name | Description | Link |
---|---|---|
PyMultiNest** | Pythonic Bayesian inference and visualization for the MultiNest Nested Sampling Algorithm or MCMC. See also the tutorial, RMultiNest. | |
BXA** | Bayesian X-ray analysis (nested sampling for Xspec and Sherpa) | |
NWAY | nway -- Bayesian cross-matching of astronomical catalogues | |
UltraNest | Pythonic Nested Sampling Development Framework & UltraNest | |
simbad2kstars | Simbad to KStars import | |
SysCorr | Bayesian correlation swiss army knife | |
jbopt | A interface definition that lets you plug and play many algorithms against your likelihood function, including optimisation algorithms in scipy, OpenOpt, MultiNest, emcee, etc. | |
LightRayRider | Ray tracing of hydrodynamic simulations to compute column densities | |
languagecheck | Improve the language of your paper before submission | |
test-calculator | Online Scientific computations (with javascript) | |
athena-point-source-simulator | Simulating Compton-thick AGN for Athena | |
intersection | Ray tracing / Line intersection formulas for various 2d and 3d objects | |
spuren | A Desktop search engine kept fast and simple | |
DHCProbe | Send a DHCP request to DHCP server to check its configuration | |
imagehash** | A Python Perceptual Image Hashing Module | |
zwicky-morphological-analysis | Zwickys Morphological Analysis implemented in Python |
** very popular