Research
Research
Publications
A full list of publications can be found in NASA ADS.
Research topics
I am an observational cosmologist. I like to understand the data I work with in detail, including how they are reduced, calibrated, and turned into the measurements used for cosmological inference. At the same time, I develop new statistical tools to analyse these data in different ways and to stress-test our cosmological model.
I have extensive experience with weak gravitational lensing, which uses the small distortions induced in the observed shapes of distant galaxies to trace the matter distribution in the Universe, without relying directly on how galaxies trace that matter. Through my work in the Dark Energy Survey and Euclid, I have contributed to all stages of weak-lensing cosmological analyses. In particular, I have worked on redshift calibration, shear-catalogue production, weak-lensing mass-map reconstruction, and the development of new methods to extract cosmological information beyond standard two-point analyses, using higher-order statistics and simulation-based inference.
I am also interested in complementary probes of large-scale structure, particularly CMB lensing and the thermal Sunyaev–Zel’dovich effect. CMB lensing traces the integrated matter distribution over a broad range of redshifts, while the thermal Sunyaev–Zel’dovich effect probes the hot ionized gas in galaxy groups and clusters. On their own, and especially in combination with galaxy weak lensing, these observables can improve cosmological constraints, help identify systematic effects, and provide information about baryonic feedback and the distribution of matter and gas across a wide range of scales and cosmic time.
Weak-lensing inference beyond two-point statistics
My work on higher-order weak-lensing statistics is motivated by the goal of making these analyses as robust and mature as standard two-point approaches, while exploiting their ability to capture more of the information contained in the fields and substantially improve cosmological constraints. I began by developing theoretical models and explicit-likelihood analyses for second- and third-order moments of weak-lensing mass maps, and later moved toward simulation-based inference, which makes it possible to combine a much broader range of non-Gaussian summaries.
I co-lead the team that developed the DES Year 3 simulation-based inference analyses, combining moments, scattering transforms, wavelet phase harmonics, persistent homology, and learned compression. We are now extending this programme to DES Y6/DECADE and Euclid DR1. In parallel, I work on the tools needed to make these analyses reliable: data compression, baryonic-feedback modelling, intrinsic alignments, source clustering, interpretation of the summary statistics, and validation against observational and astrophysical systematics.
Simulation-based inference
Higher-order analyses and systematics
Weak-lensing surveys, maps, and calibration
My work on weak-lensing surveys covers redshift calibration, shear catalogues, mass-map reconstruction, and systematic-error testing. In DES, I developed clustering-redshift methods, co-led the construction and validation of the Year 3 shear catalogue and mass maps, and later led the combined DES Y3+DECADE mass map. Working directly on these data products shapes how I design and validate the cosmological analyses that use them.
Shape catalogues and mass maps
Redshift calibration
CMB lensing, tSZ, and baryonic physics
I use CMB secondary anisotropies as complementary probes of large-scale structure. CMB lensing traces the integrated matter distribution to higher redshift than galaxy weak lensing, while the thermal Sunyaev–Zel’dovich effect traces the thermal pressure of ionized gas in groups and clusters. Their cross-correlations with galaxy weak lensing therefore extend the analysis across a wider range of redshifts, physical scales, and halo masses, while helping to break parameter degeneracies and test observational systematics, foreground contamination, and models of baryonic feedback.
My work includes cross-correlations of DES weak-lensing data with ACT and Planck Compton-y maps, used to constrain halo pressure profiles and feedback models. I am also developing map-level baryonification and forward-modelling tools to generate realistic joint weak-lensing and tSZ mocks for simulation-based inference. I currently lead the production of the South Pole Telescope five-year main-field tSZ map and am extending my SBI framework to joint analyses of weak lensing, CMB lensing, and tSZ.