A New Statistical Study Of Galaxy-pair Populations In The COSMOS Field Based On Chance-projection Probabilities

We present a new statistical study of the evolution of galaxy-pair fractions in the COSMOS survey up to z 1. We use a simple method based on chanceprojection probabilities to indirectly sample relatively isolated galaxy pairs. The method subtracts contaminations of chance-project pairs according to simulated random galaxy distributions. We for the ?rst time assess and empirically avoid the screen effect in high-redshift pair statistics. Our sampled pairs statistically have projected separations(rp) from ～ 20 h-175 kpc to ～ 30 h-175 kpc. By taking these pairs as evolutionary counterparts of the nearby CPG pairs[1–3]of similar rp, which is supported by our additional statistics, we ?nd that the pair fractions evolve with redshift as(1 + z)mwith m = 0.93 ± 0.20. This result argues for a signi?cant medium-fast pair-fraction increase toward higher z. We ?nd that sampling of populations of high-redshift pairs of rp< rsc pis likely strongly biased by the screen effect if using sky-survey object catalogs generated with currently available algorithms(e.g. the SExtractor), with rsc pvarying with the redshift in the range ～ 15 h-175 kpc to ～ 25 h-175 kpc. Furthermore, current object identi?cation from space-based sky-survey images may have over-deblended high-redshift galaxies and generated false galaxy pairs of rp rsc p. These issues raise serious doubts about previous results suggesting fast pair-fraction evolution, m ≈ 2.2- 3.1. In addition, our study shows strong and complex evolution of galaxy environment in the vicinities of relatively isolated pairs. We thus cau-tion about directly interpreting pair-fraction increase(re?ected in the m) as the galaxy merger-rate increase. Conversion of pair-fraction increase to merger-rate increase involves the merger timescales(τm) at various redshifts, but τmmay evolve in an unknown complex pattern under the in?uences of the evolving close galaxy environments of the pairs.Overall, our study uncovers several problems that should be interesting for future research of high-redshift pairs, mainly the in?uences of the screen effect and cosmic environments of pairs on pair statistics and τm.