Assessing the Impact of Source Confusion for GREX-PLUS based on Deep JWST NIRCam Imaging

TL;DR

This study evaluates how source confusion affects the depth and completeness of GREX-PLUS infrared observations by simulating images based on JWST data, revealing limitations and feasibility for faint galaxy studies.

Contribution

It provides a detailed analysis of source confusion impacts on GREX-PLUS imaging depth and completeness, using simulations based on JWST data and Monte Carlo methods.

Findings

GREX-PLUS images are shallower than JWST images due to unresolved sources and PSF wings.

Depth improvement continues with longer integrations but at diminishing returns.

Confusion reduces detection completeness, but statistical studies of faint galaxies remain feasible.

Abstract

We investigate the effects of source confusion expected in observations with GREX-PLUS, a JAXA L-class space infrared telescope mission candidate with a wide-field infrared camera covering 2-8 um with a field of view of 0.50 deg$^2$. For the deep imaging band near 4 um, we calculate the GREX-PLUS PSF and ghost based on the latest optical design, and consider two representative imaging performance cases with PSF FWHM values of 0.9 and 1.2 arcsec. We construct simulated GREX-PLUS images at different depths by convolving JWST NIRCam imaging data from JADES, GLASS, CEERS, and COSMOS-Web with the PSF+ghost kernel. Comparing the limiting magnitudes estimated from random aperture photometry using the same aperture sizes, we find that the simulated GREX-PLUS images are shallower than the original JWST images, with larger deviations for deeper original JWST images. This likely reflects unresolved faint sources and extended PSF+ghost wings from bright sources, which elevate background fluctuations in blank regions. Nevertheless, the limiting magnitudes continue to improve with increasing integration time down to ~27 mag, without a clear plateau at depths comparable to the planned GREX-PLUS deep survey, although the improvement becomes progressively less efficient toward longer integrations. Based on Monte Carlo simulations, we estimate detection completeness and correct the number counts for magnitude bias and incompleteness, finding that confusion-induced blending can reduce the completeness even at magnitudes well above the nominal 5-sigma depth. The completeness-corrected number counts agree well with the JWST-based number counts down to around the detection limit. Overall, our results suggest that statistical studies of faint galaxies remain feasible for GREX-PLUS; however, survey planning should account for less efficient depth improvement toward longer integrations due to source confusion.