Cosmological constraints from the power spectrum and bispectrum of 21cm intensity maps

Abstract: The 21cm emission of neutral hydrogen is a potential probe of the matter distribution in the Universe after reionisation. Cosmological surveys of this line intensity will be conducted in the coming years by the SKAO and HIRAX experiments, complementary to upcoming galaxy surveys. We present the first forecasts of the cosmological constraints from the combination of the 21cm power spectrum and bispectrum. Fisher forecasts are computed for the constraining power of these surveys on cosmological parameters, the BAO distance functions and the growth function. We also estimate the constraining power on dynamical dark energy and modified gravity. Finally we investigate the constraints on the 21cm clustering bias, up to second order. We consider the effects on the 21cm correlators of the telescope beam, instrumental noise, foreground avoidance, the Alcock-Paczynski effect and theoretical errors in the modelling of the correlators. Adding Planck priors, and marginalising over nuisance parameters, HIRAX achieves sub-percent precision on the $\Lambda$CDM parameters, with SKAO delivering slightly lower precision. The modified gravity parameter $\gamma$ is constrained at 1% (HIRAX) and 5% (SKAO). For the dark energy parameters $w_0,w_a$, HIRAX delivers percent-level precision while SKAO constraints are weaker. HIRAX achieves sub-percent precision on the BAO distance functions $D_A,H$, while SKAO reaches 1-2% for $0.6\lesssim z\lesssim 1$. The growth rate $f$ is constrained at a few-percent level for the whole redshift range of HIRAX and for $0.6\lesssim z\lesssim 1$ by SKAO. The different performances arise mainly since HIRAX is a packed inteferometer that is optimised for BAO measurements, while SKAO is not optimised for interferometer cosmology and operates better in single-dish mode, where the telescope beam limits access to the smaller scales that are covered by an interferometer.