Line of sight projections tend to dominate detections of dusty galaxies

In previous work, we found that a single-dish detection of a far-infrared (FIR) source, such as those seen by the Herschel Space Observatory, is often best reproduced by a combination of multiple sources, rather than by a single, extremely luminous galaxy. This result specifically is true when the Herschel detection has both a 24 micron and a 3.6 micron detection beneath it.

In this work, we aimed to determine whether those multiple bright counterparts are physically associated with each other, or whether they are simply line-of-sight projections. If physically associated, this bolsters the historical assumption that many of these bright, dusty galaxies are analogues to the local ULIRG population, which is almost entirely comprised of interacting galaxies.

We therefore matched all of our bright counterparts to a photometric redshift (or a spectroscopic redshift if available) for all available sources in the COSMOS field, using the L'aigle+2016 catalogue of photometric redshifts, and the GAMA team's catalogue of spectroscopic redshifts in the COSMOS field.

  Figure 2a of Scudder et al 2018 . This shows the redshift solutions for three dusty, bright galaxies whose light was blended together in a single Herschel detection. These three solutions do not overlap at all, and so the three galaxies must not be at consistent redshifts.

Figure 2a of Scudder et al 2018. This shows the redshift solutions for three dusty, bright galaxies whose light was blended together in a single Herschel detection. These three solutions do not overlap at all, and so the three galaxies must not be at consistent redshifts.

For each FIR detection, we plot the photometric redshift solutions, and calculate the likelihood that the two galaxies are likely to be at the same redshift. If the two solutions have no overlap, it is impossible for the two galaxies to coexist at the same redshift. We calculate the likelihood of coexistence for all possible pair permutations, and summarize in the Figure below.

  Figure 6 of Scudder et al. 2018 . This shows the joint probability across all bright pairings of galaxies. The vertical dotted line shows 50% probability. 70% of the sample has no overlap in their redshift solutions (0 percent likelihood of existing at the same redshifts.)

Figure 6 of Scudder et al. 2018. This shows the joint probability across all bright pairings of galaxies. The vertical dotted line shows 50% probability. 70% of the sample has no overlap in their redshift solutions (0 percent likelihood of existing at the same redshifts.)

We find that ~70% of the sample of dusty, bright galaxies are inconsistent with existing at the same redshift. This means that they are largely line-of-sight projection effects, and not galaxy interactions. Multiple unrelated galaxies are lining up to inflate the luminosity of the source detected by Herschel. Each galaxy can therefore be a less extreme version of what had been originally suggested by the amount of star formation required to cause such a luminous FIR detection.