NASIM: Revealing the low surface brightness Universe from legacy VISTA data

TL;DR

NASIM is an automated pipeline designed to enhance the detection of faint, diffuse structures in near-infrared VISTA data, enabling new science in low surface brightness astronomy.

Contribution

It introduces NASIM, a novel data reduction pipeline optimized for VISTA observations, improving the recovery of low surface brightness features over traditional methods.

Findings

NASIM effectively recovers diffuse structures in deep VISTA data.

Compared to conventional pipelines, NASIM preserves low surface brightness features better.

Demonstrated success in analyzing galaxy outskirts, LSB galaxies, and intracluster light.

Abstract

Near-infrared imaging is a powerful technique in observational astronomy, but the bright background, primarily from the Earth\'s atmosphere, makes the detection of faint features particularly challenging. To recover low surface brightness (LSB) structures in such data, we present NASIM (Near-infrared Automated low Surface brightness reduction In Maneage), a fully automated and reproducible data reduction pipeline optimised for VISTA/VIRCAM observations. NASIM builds on GNU Astronomy Utilities (Gnuastro) to effectively remove large-scale instrumental artefacts while preserving faint, diffuse emission. As a key science application, we focus on deep Ks-band observations of the Euclid Deep Field South (KEDFS), one of the deepest VISTA/VIRCAM datasets and a high-priority field for synergy with current and future facilities, including Euclid, JWST, LSST, Roman, Spitzer, and ALMA. With VIRCAM no longer operational, KEDFS now stands as a unique legacy dataset. We release selected tiles from the KEDFS survey and highlight science cases, including galaxy outskirts, LSB galaxies, and intracluster light, that demonstrate NASIM\'s ability to recover diffuse structures. A direct comparison with conventional VISTA data reduction pipelines demonstrates the advantages of NASIM in preserving diffuse emission without compromising compact source detection. All quantitative results presented in this paper are fully reproducible with Maneage (commit 4d32667).