Postdoctoral Fellow · Argelander Institute for Astronomy, University of Bonn
My research centres on extracting cosmological information from large-scale structure observables, with a particular focus on weak gravitational lensing and Fast Radio Bursts. I develop and apply statistical frameworks to map the distribution of matter and baryons in the Universe, test extensions of the standard cosmological model, and calibrate astrophysical processes — in particular baryonic feedback — that otherwise bias cosmological inference.
Mass along the line of sight deflects light from distant galaxies, coherently distorting their observed shapes. The resulting signal — cosmic shear — provides a direct probe of the projected matter distribution, sensitive to the growth of structure and the geometry of space-time. My work spans covariance modelling, higher-order statistics, photometric redshift systematics, and tests of modified gravity. I lead the covariance team within Rubin-LSST DESC and contribute to the Euclid covariance matrix.
Fast Radio Bursts are millisecond-duration radio transients observable at cosmological distances. Their pulses are dispersed by intervening ionised gas, imprinting a dispersion measure proportional to the integrated electron column density along each line of sight — making FRBs a direct tracer of the cosmic baryon distribution. I develop statistical frameworks to constrain the Hubble constant, test the equivalence principle, probe primordial non-Gaussianity, and measure baryonic feedback from FRB samples.
Energetic astrophysical processes — AGN feedback, supernovae, stellar winds — expel baryons from haloes, suppressing the matter power spectrum on small scales and introducing a significant systematic for weak lensing cosmology. Disentangling baryonic effects from the cosmological signal requires direct tracers of the baryon distribution. Using FRBs and weak lensing jointly, I demonstrated the first direct measurement of baryonic feedback with FRBs, rejecting no-feedback scenarios at greater than 3σ.
I have a broad interest in statistical inference, machine learning, and information theory. This spans simulation-based inference, functional derivative techniques, and information geometry — the differential-geometric study of parametric families of probability distributions — and their applications to cosmological parameter estimation and experimental design. I develop these tools both to answer specific cosmological questions and as methods research in their own right.
The Kilo-Degree Survey is a wide-field optical imaging survey designed to map cosmic shear across 1350 square degrees. For the KiDS-Legacy final analysis, I developed the end-to-end covariance model, publicly released as OneCovariance, and led analyses constraining dark energy, neutrino mass, spatial curvature, and modifications of gravity.
Euclid and the Legacy Survey of Space and Time (LSST) will each observe several billion galaxies, representing a step-change in statistical precision over current surveys. I lead the covariance team within the LSST Dark Energy Science Collaboration and contribute to the Euclid covariance matrix, extending the methods developed for KiDS to these next-generation datasets.
The Square Kilometre Array Observatory will detect FRBs at a rate orders of magnitude beyond current facilities. I co-lead the FRB science working group within the SKA cosmology community, developing forecasts and statistical frameworks for extracting constraints on the baryon distribution, the Hubble constant, and fundamental physics from the anticipated SKA FRB samples.
Leading-author papers. Full list on arXiv and ORCID.
Papers led by students I supervised or co-supervised. See also the Students page.
All public repositories are available on GitHub.
OneCovariance —
Complete covariance framework for projected large-scale structure observables. Used in the KiDS-Legacy
analysis and serving as the blueprint for Euclid and Rubin-LSST covariance matrices.
levin_bessel (PyLevin) —
Python/C++ library for fast numerical integration of products of up to three Bessel functions
via the Levin method. Published in the
Journal of Open Source Software.
CCL —
Core Cosmology Library: a Python/C library providing fast, accurate tools for cosmological
calculations, maintained by the LSST Dark Energy Science Collaboration.
Spaceborne —
End-to-end Fisher matrix and covariance forecast code for Euclid weak lensing and
galaxy clustering observables.
CLOE —
Cosmological Likelihood for Observables in Euclid: the official Euclid Consortium
likelihood and analysis pipeline.
TJPCov —
Two-point covariance estimation library for the LSST Dark Energy Science Collaboration.