Phase-space Spectral Line Deconfusion in Intensity Mapping

Line intensity mapping (LIM) is a promising tool to efficiently probe the three-dimensional large-scale structure by mapping the aggregate emission of a spectral line from all sources that trace the matter density field. Spectral lines from different redshifts can fall in the same observed frequency and be confused, however, which is a major challenge in LIM. In this work, we develop a line deconfusion technique in map space capable of reconstructing the three-dimensional spatial distribution of line-emitting sources. If multiple spectral lines of a source population are observable in multiple frequencies, using the sparse approximation, our technique iteratively extracts sources along a given line of sight by fitting the LIM data to a set of spectral templates. We demonstrate that the technique successfully extracts sources with emission lines present at a few σ above the noise level, taking into account uncertainties in the source modeling and presence of continuum foreground contamination and noise fluctuations. As an example, we consider a Tomographic Ionized-carbon Mapping Experiment/CarbON C II line in post-rEionisation and ReionisaTiOn epoch (TIME/CONCERTO)-like survey targeting [C II] at the epoch of reionization, and reliably reconstruct the 3D spatial distribution of the CO interlopers and their luminosity functions at 0.5 ≲ z ≲ 1.5. We also demonstrate a successful deconfusion for the Spectro-Photometer for the History of the Universe, Epoch of Reionization, and Ices Explorer (SPHEREx) mission in the near-infrared wavelengths. We discuss a formalism in which the reconstructed maps can be further cross-correlated with a (galaxy) tracer population to estimate the total interloper power. This technique is a general framework to extract the phase-space distribution of low-redshift interlopers, without the need of external information, for any line deconfusion problem.