Toward a robust estimate of the merger rate evolution using near-IR photometry

We use a combination of deep, high angular resolution imaging data from the CDFS (HST/ACS GOODS survey) and ground-based near-IR K-s images to derive the evolution of the galaxy major merger rate in the redshift range 0: 2 <= z <= 1.2. We select galaxies solely on the basis of their J-band rest-frame absolute magnitude, which is a good tracer of the stellar mass. We find steep evolution with redshift, with the merger rate proportional to(1 + z)(3.43+/-0.49) for optically selected pairs and proportional to(1 + z)(2.18+/- 0.18) for pairs selected in the near-IR. Our result is unlikely to be affected by luminosity evolution that is relatively modest when using rest-frame J-band selection. The apparently more rapid evolution that we find in the visible is likely caused by biases relating to incompleteness and spatial resolution affecting the ground-based near-IR photometry, underestimating pair counts at higher redshifts in the near-IR. The major merger rate was similar to 5.6 times higher at z similar to 1.2 than at the current epoch. Overall, 41% x (0.5 Gyr/tau) of all galaxies with M-J <= -19.5 have undergone a major merger in the last similar to 8 Gyr, where tau is the merger timescale. Interestingly, we find no effect on the derived major merger rate due to the presence of the large-scale structure at z - 0.735 in the CDFS.