The feasibility of weak lensing and 21cm intensity mapping cross-correlation measurements

One of the most promising probes to complement current standard cosmological surveys is the H i intensity map, i.e. the distribution of temperature fluctuations in neutral hydrogen. In this paper we present calculations of the two-point function between HI (at redshift z< 1) and lensing convergence (κ). We also construct HI intensity maps from N-body simulations, and measure two-point functions between HI and lensing convergence. HI intensity mapping requires stringent removal of bright foregrounds, including emission from our Galaxy. The removal of large-scale radial modes during this HI foreground removal will reduce the HI-lensing cross-power spectrum signal, as radial modes are integrated to find the convergence; here we wish to characterize this reduction in signal. We find that after a simple model of foreground removal, the cross-correlation signal is reduced by ~50–70 per cent; we present the angular and redshift dependence of the effect, which is a weak function of these variables. We then calculate S/N of κHI detection, including cases with cut sky observations, and noise from radio and lensing measurements. We present Fisher forecasts based on the resulting two-point functions; these forecasts show that by measuring κ Δ THI correlation functions in a sufficient number of redshift bins, constraints on cosmology and HI bias will be possible.