Improving cosmological analyses of HI clustering by reducing stochastic noise

• DOI: 10.48550/arxiv.2405.18559
• Publication Date: 2024-05-28
• Authors: Simon Foreman; Andrej Obuljen; Marko Simonović
• Keywords: Cosmology and Nongalactic Astrophysics (astro-ph.CO); FOS: Physical sciences; Astrophysics - Cosmology and Nongalactic Astrophysics

High-number-density tracers of large-scale structure, such as the HI-rich galaxies measured by 21 cm intensity mapping, have low sampling noise, making them particularly promising cosmological probes. At large scales, this noise can be subdominant to other scale-independent contributions power spectrum, arising from nonlinear bias. This has important consequences for constraints obtained tracers, since it indicates that using spectrum does not lead optimal even in linear regime. In paper, we provide a conservative estimate possible improvement constraining 21cm survey if one were use an analysis strategy (such field-level analysis), where only true enters error budget. We find improvements uncertainties on some parameters 50%, depending redshift, foreground cleaning efficiency, scales used analysis, and instrumental noise. One byproduct our work is measurements bias stochasticity neutral hydrogen IllustrisTNG simulation over wide range redshifts; simple fitting formulas these measurements. Our results motivate further exploration new techniques insights into current future surveys.