Interests

The Shape of the Power Spectrum: a new observable in cosmology.
State-of-the-art techniques focus on the Baryon Acoustic Oscillation (BAO) signal to measure the expansion rate of the Universe in units of the sound horizon scale. Additionally, the amplitude of the isotropic and anisotropic power spectrum signals can be used to constrain the growth of structure of the Universe, via the galaxy's peculiar velocities: the so-called Redshift Space Distortions (RSD). These two observables, BAO+RSD, have been key for extracting cosmological information from spectroscopic galaxy maps once a cosmology model is assumed. Further cosmology-relevant information has been for a long time unused, as it was complemented by CMB observations. This information is related to the shape of the transfer function of the matter power spectrum after recombination, and is also contained in the observed galaxy power spectrum. Such information can be compressed in just one extra cosmology parameter in addition to the usual BAO and RSD parameters: the Shape. By using all BAO, RSD and Shape, galaxy clustering measurements put competitive constraints on the amount of matter in the Universe, or the sum of the neutrino masses.

Ref. Brieden, GM, Verde 2023, Brieden, GM, Verde 2022a, Brieden, GM, Verde 2022b, Brieden, GM, Verde 2021a, Brieden, GM, Verde 2021b

Baryon Acoustic Oscillations of the large sale structure as a probe of the expansion history of the Universe. The Baryon Acoustic Oscillations (BAO) scale is the cornerstone for extracting cosmological information from large catalogues of galaxies. This scale was imprinted in the matter distribution on the early Universe by pressure waves traveling through the primordial plasma formed by coupled photons and baryons. By sampling massive galaxy catalogues we are able to identify such scales in the galaxy correlation function and measure the expansion history of the Universe at that epoch. We refer to the BAO scale as a standard ruler, as it is very insensitive to spurious effects, such as nonlinear evolution, or astrophysical processes (galaxy bias). With this powerful tool in hand, cosmologists can probe how the Universe has been expanding through different epochs.
New generation of galaxy surveys, such as DESI and Euclid will reach the precision limit imposed by systematics, which urges us to develop new approaches to keep them under control in order to extract the largest amount of information.
bi
Ref. Eisenstein et al. 2005, Eisenstein, Seo, Sirko & Spergel 2006, Hinton, Howlett & Davis 2020, Gil-Marin et al. 2020

Redshift Space Distortions as a probe of gravity.
Redshift Space Distortions (RSD) are the apparent shift in the position of galaxies along the line of sight of the observer caused by the peculiar motions of these galaxies. In cosmology we infer the radial positions of distant galaxies by measuring their spectra and comparing them with templates at rest. The shift between these it's called redshift. Assuming that the redshift is solely caused by the Hubble flow we infer the radial distance given a cosmological model. However, the velocity of galaxies has an extra component in addition to the Hubble flow because of the interaction with nearby galaxies. This extra component distorts the measured galaxy radial position. However, this distortion in positions in the galaxy catalogue is not statistically random but contains information about how dense the Universe is. By detecting and properly modelling this effect we are able to determine the growth of structure of the Universe, and therefore to test the theory General Relativity at cosmological scales.

Ref. Kaiser 1987, Percival & White 2008, McDonald 2009, Gil-Marin et al 2016, Gil-Marin et al 2018, Gil-Marin et al. 2020.

Information content of the bispectrum signal of galaxies.
The statistical distribution of a galaxy field is fully described by 2-point statistics (the power spectrum in Fourier space) iff the distribution of galaxies is Gaussian. In practice the positions of galaxies in the late-time Universe are strongly influenced by non-linear gravitational physics, and the galaxy field is actually fully non-Gaussian. Therefore, if we want to increase the amount of information we extract from a given galaxy catalogue, we need to look, not only at the power spectrum signal, but also ats the bispectrum signal (the 3-point correlation function in Fourier space). In order to do this one needs to address several challenges.
1. Building an efficient estimator is key because of the large amount of shapes and scales one needs to measure.
2. Developing reliable theoretical models in order to describe the effects of galaxy bias and redshift space distortions on the bispectrum signal, as well as modelling the effects of the covariance among triangles of similar shapes and scales.
3. Developing techniques to address the observational effects of an incomplete distribution of galaxies.

Ref: Verde et al. 2001, Scoccimarro et al. 1998, Gil-Marin 2011, Gil-Marin et al. 2015, Gil-Marin et al. 2018.

Blinding in Cosmology.
The current cosmological observations, from early- (CMB), to late-time Universe (LSS, SNe), strongly favour the standard model of cosmology, the flat-ΛCDM model, which provides a coherent and precise description. Within this model Dark Energy and Dark Matter components are poorly understood and may be considered as effective descriptions for the true model's ingredients which are yet to be found. Next generation of LSS surveys will provide a wealth set of data which will bring under a stress-test the flat-ΛCDM model. In order accomplish this goal we need to account for all potential sources of systematic errors sources and properly correct for them. In this process the experimenter's bias enters into play, where we, as observers, may stop looking or accounting for a specific systematic when the observations reproduce the expected flat-ΛCDM model. One promising way of avoiding experimenter's bias is to carry out blind data analyses, where the original data vector is transformed as to hide the true signal in a controlled way before running the analysis pipeline. If suitably implemented, this allows for studying the degeneracies of systematic effects with theoretical models while being blind to the underlying values of the model parameters. Once the analysis of the blinded data is analyzed, the analysis pipeline is frozen and either applied to the original data, or the results are unblinded.

Ref. Brieden et al. 2020

Research Interests