Research topics

My research focuses on developing statistical and data analysis methods that extract the maximum amount of information from cosmological data, with applications to current and next-generation surveys.

Field-level inference analyses cosmological data directly at the level of maps and fields, without relying on data compression. This approach captures the full information content of the observations, enabling higher-precision cosmological constraints and detailed reconstructions of the matter distribution across cosmic time.

• Field-level inference of cosmic shear with intrinsic alignments and baryons
  Porqueres, Heavens, Mortlock, Lavaux, Makinen (2023)
• Lifting weak lensing degeneracies with a field-based likelihood
  Porqueres, Heavens, Mortlock, Lavaux (2022)
• Bayesian forward modelling of cosmic shear data
  Porqueres, Heavens, Mortlock, Lavaux (2021)

• Euclid: Field-level inference of primordial non-Gaussianity and cosmic initial conditions
  Euclid Consortium; Andrews, Jasche et al. (2025)
• DISCO-DJ I: a differentiable Einstein-Boltzmann solver for cosmology
  Hahn, List, Porqueres (2023)
• Explicit Bayesian treatment of unknown foreground contaminations in galaxy surveys
  Porqueres, Ramanah, Jasche, Lavaux (2018)
• Imprints of the large-scale structure on AGN formation and evolution
  Porqueres, Jasche, Ensslin, Lavaux (2017)

• LyAl-Net: A high-efficiency Lyman-alpha forest simulation with neural networks
  Boonkongkird, Lavaux, Peirani, Dubois, Porqueres, Tsaprazi (2023)
• A hierarchical field-level inference approach to reconstruction from sparse Lyman-
  Porqueres, Hahn, Jasche, Lavaux (2020)
• Inferring high redshift large-scale structure dynamics from the Lyman-alpha forest
  Porqueres, Jasche, Lavaux, Ensslin (2019)

Simulation-based inference allows us to extract information from complex data compressions that could not be accurately modelled using traditional analytical approaches. Combining large simulation suites with neural networks, we can define an optimal data compression to retain maximal information for cosmological analyses.

• Hybrid Summary Statistics
  Makinen, Sui, Wandelt, Porqueres, Heavens (2024)
• Hybrid summary statistics: neural weak lensing inference beyond the power spectrum
  Makinen, Heavens, Porqueres, Charnock et al. (2024)
• The Cosmic Graph: Optimal Information Extraction from Large-Scale Structure using Catalogues
  Makinen, Charnock, Lemos, Porqueres, Heavens, Wandelt (2022)

High-order statistics can capture the non-Gaussian features of the cosmological datasets, providing more precision in the cosmology results. This can be achieved with high-order correlation functions and topological descriptors of the observations.

• Euclid preparation. LXXXV. Toward a DR1 application of higher-order weak lensing statistics
  Euclid Consortium; Vinciguerra, Bouche et al. (2025)

Intrinsic alignments are one of the main systematic effects of weak lensing analyses. Galaxy shapes are affected by the tidal fields of the cosmic structures where they form. Modelling this effect to sufficient accuracy is essential for weak lensing analyses.

• Accuracy requirements on intrinsic alignments for Stage-IV cosmic shear
  Paopiamsap, Porqueres, Alonso, Harnois-Deraps, Leonard (2023)

Information field theory is a fully Bayesian framework for signal processing and image reconstruction, which mathematically uses tools from field theory.

• NIFTy 3 - Numerical Information Field Theory - A Python framework for multicomponent signal inference on HPC clusters
  Steininger, Dixit, Frank et al. (2017)
• Cosmic expansion history from SNe Ia data via information field theory -- the charm code
  Porqueres, Ensslin, Greiner, Boehm et al. (2016)