The Role of Galaxy Mergers in the High Redshift Universe with Deep JWST Extragalactic Fields

We analyze the redshift evolution of major galaxy merger pair fractions using deep multi-band photometry from the ASTRODEEP-JWST catalogs, covering seven extragalactic fields (CEERS, Abell2744, NGDEEP, PRIMER-COSMOS, PRIMER-UDS, JADES-GN, and JADES-GS). Our sample includes ∼500,000 sources with photometric redshifts and NIRCam-HST photometry, selected with a signal-to-noise ratio (S/N) > 3 in the F444W band. We classify galaxies into red and blue populations based on their F160W − F444W color, adopting a threshold of 2.0 mag, and compute pair fractions for major mergers (stellar mass ratio ≥ 1 : 4) within a projected separation of 5–50 kpc, over the redshift range 0 < z < 10.5. We identify galaxy pairs with projected physical separations in the range 5–50 kpc and restrict the line-of-sight comoving separation to be ≤ 50 kpc, based on photometric redshifts and angular distances.  Companion identification accounts for photometric redshift uncertainties through adaptive redshift binning and applies both transverse and line-of-sight distance thresholds. The analysis is performed in three stellar mass bins (8.0 < log M⋆/M⊙ < 8.5, 8.5 < log M⋆/M⊙ < 9.0, 9.0 < log M⋆/M⊙ < 10.0), and statistical uncertainties are estimated via bootstrap resampling.  Our results show that pair fractions are generally lower than those in recent JWST studies, especially in the z = 2–6 range, likely due to our stricter sample selection and the absence of completeness weighting. However, we identify a sharp increase in pair fractions at z ∼ 8–9 across all stellar masses and for both red and blue populations, an unexpected feature not seen in previous literature, where high-redshift trends are typically declining. While this spike may reflect an actual increase in early merger activity, it could also arise from photometric redshift uncertainties, small-number statistics, low signal-to-noise sources, or the lack of mass completeness corrections. Future work, including completeness weighting and visual validation, will be required to confirm its physical origin.