CO Contamination 1–10%

Carbon monoxide rotational line emission from molecular gas in star-forming galaxies contaminates intensity mapping and spectroscopic surveys that target other spectral features, particularly the 21 cm hydrogen signal and the [CII] line at high redshift. CO lines from different rotational transitions (J=1-0 at 115 GHz, J=2-1 at 230 GHz, and higher-J transitions) are redshifted into the observing bands of surveys targeting lower-frequency signals from higher redshifts, creating an interloper contamination that is difficult to separate without multi-frequency data. Current estimates of CO contamination in intensity mapping experiments range from 1 to 10 percent of the target signal power, depending on redshift and angular scale, with the uncertainty in the CO luminosity function at high redshift being the dominant source of modeling error. This contamination must be accurately characterized and removed to extract reliable cosmological constraints from next-generation intensity mapping surveys.

Successive Collision Theory affects the CO contamination problem through the pre-existing stellar populations that modified the high-redshift CO luminosity function relative to ΛCDM expectations. In SCT, the pre-existing stellar generations from the colliding pockets had already processed significant amounts of carbon into CO through molecular cloud formation and stellar nucleosynthesis in their prior epochs. When the collision thermalized these populations, the processed molecular gas — including CO — was mixed into the debris field and contributed to the molecular gas reservoir available at high redshift. The effective CO luminosity function at z > 3 is therefore elevated above what pure post-collision stellar evolution produces, increasing the contamination fraction in intensity mapping bands that probe these redshifts. The magnitude of the CO excess traces the same pre-existing stellar content that produces the early massive galaxies and the metallicity floor — all reflecting the same pre-collision stellar populations.

The angular momentum organization of the debris field also affects the spatial statistics of CO contamination. In ΛCDM, CO interloper galaxies are clustered according to standard hierarchical halo occupation, producing a specific angular power spectrum of contamination that depends on the galaxy clustering bias. In SCT, the CO-emitting galaxies at high redshift are organized along the angular momentum strata of the collision debris, producing a contamination power spectrum with non-Poissonian angular correlations on large scales correlated with the cosmic web orientation. This spatially correlated contamination is harder to separate from the target signal than uncorrelated noise, potentially inflating the effective noise floor in intensity mapping beyond what ΛCDM-based contamination models predict. SCT therefore motivates more conservative contamination treatments in intensity mapping data analysis and the development of foreground cleaning algorithms that account for large-scale angular correlation in the interloper population.