July 28, 2025 to August 1, 2025
Perimeter Institute for Theoretical Physics
America/Toronto timezone

Tracers with a Twist: Reshaping 21cm and Galaxy Clustering with Baryonic Physics

Not scheduled
1m
PI/1-100 - Theatre (Perimeter Institute for Theoretical Physics)

PI/1-100 - Theatre

Perimeter Institute for Theoretical Physics

190
Flash Talk

Speaker

Calvin Osinga (University of Maryland)

Description

We measure the spatial distribution of visible tracers to infer the large-scale structure (LSS) of matter and constrain cosmological parameters. Observations of blue and red galaxies are traditionally used for this purpose, although upcoming 21cm surveys, measuring the distribution of atomic hydrogen (HI), can provide a competitive alternative. By providing a different picture of LSS, these HI observations can improve the cosmological constraints from galaxy surveys alone and deepen our understanding of LSS. However, constructing the models necessary to interpret measurements from galaxy and HI surveys requires knowledge of how galactic physics shape the spatial distribution of HI and galaxies. I will discuss two results from the hydrodynamical simulation IllustrisTNG that demonstrate the significant influence of galactic physics on the clustering of HI, blue galaxies, and red galaxies, even to the largest scales probed by IllustrisTNG. First, the clustering of all three tracers decreases with time across z $\leq$ 1 even at the largest scales, contrary to the general expectation that clustering increases with time as gravity pulls structures closer together. We demonstrate that this phenomenon arises from processes associated with quenching (especially AGN feedback) altering the kinds of structures that each tracer tends to occupy. Second, we show that the bulk velocity alignment of massive halos and the HI within them decreases with redshift due to the suppression of HI in these halos’ centers. This effect can decouple the HI and matter velocity fields and introduce a redshift-dependent velocity bias that, if neglected, may cause ~5% errors in LSS models out to quasi-linear scales.

Author

Calvin Osinga (University of Maryland)

Co-author

Benedikt Diemer (University of Maryland)

Presentation materials

There are no materials yet.