Demonstrating the feasibility of line intensity mapping using mock data of galaxy clustering from simulations

Abstract

Visbal & Loeb (2010) have shown that it is possible to measure the clustering of galaxies by cross correlating the cumulative emission from two different spectral lines which originate at the same redshift. Through this cross correlation, one can study galaxies which are too faint to be individually resolved. This technique, known as intensity mapping, is a promising probe of the global properties of high redshift galaxies. Here, we test the feasibility of such measurements with synthetic data generated from cosmological dark matter simulations. We use a simple prescription for associating galaxies with dark matter halos and create a realization of emitted radiation as a function of angular position and wavelength over a patch of the sky. This is then used to create synthetic data for two different hypothetical instruments, one aboard the Space Infrared Telescope for Cosmology and Astrophysics (SPICA) and another consisting of a pair of ground based radio telescopes designed to measure the CO(1-0) and CO(2-1) emission lines. We find that the line cross power spectrum can be measured accurately from the synthetic data with errors consistent with the analytical prediction of Visbal & Loeb (2010). Removal of astronomical backgrounds and masking bright line emission from foreground contaminating galaxies do not prevent accurate cross power spectrum measurements.

Masking line foregrounds in intensity-mapping surveys

“…This is because we cannot avoid masking some CO sources when removing HCN. As noted in Visbal, Trac, & Loeb (2011), some of the information in the power spectrum is lost in this masking process, specifically that encapsulated in the amplitudes of the clustering and shot noise terms. On a brighter note, since we are only masking a percent or so of the pixels, the shape of the clustering term of the power spectrum does not change significantly when masked.…”

“…Unfortunately, the estimate of the change in power given in equation (4.1) differs significantly from our simulations, so this cannot be used to recover the unmasked power. However, cross correlations between different maps are always possible, and S14 and Visbal, Trac, & Loeb (2011) discuss the possibility of using information in another wavelength to isolate pixels which contain foreground galaxies, thus limiting the amount of signal lost in masking. G14 also state that the behavior of line-of-sight Fourier modes may be different for signal and foreground lines.…”

“…The typical method of simulating maps with realistic clustering is to perform an N-body dark matter simulation and populate the result with galaxies (Visbal, Trac, & Loeb 2011). However, this is computationally intensive and makes it difficult to simulate large sky areas.…”

“…We perform our study using simulated intensity maps. Visbal, Trac, & Loeb (2011) used N-body dark matter simulations for their analysis, but this technique is more numerically intensive than required for our purposes. We therefore describe below a different method for simulating intensity maps, in which we assume a model for the matter power spectrum is known a priori.…”

On Estimation of Contamination from Hydrogen Cyanide in Carbon Monoxide Line-intensity Mapping

“…Furthermore, as in Li et al (2016), we use halo catalogues from dark matter simulations rather than analytically derived galaxy distributions as in Breysse et al (2015). This approach affords greater flexibility in generating mock galaxies and sky data 1 Visbal et al (2011) calibrate many of their line-luminosity relations solely based on M82, and do not examine HCN(1-0). In addition, their focus was on CO(1-0)-CO(2-1) cross-correlation.…”