As the James Webb Space Telescope (JWST) has become fully operational, early-release data are now available to begin building the tools and calibrations for precision point-source photometry and astrometry in crowded cluster environments. Here, we present our independent reduction of NIRCam imaging of the metal-poor globular cluster M 92, which were collected under Director’s Discretionary Early Release Science programme ERS-1334. We derived empirical models of the Point Spread Function (PSF) for filters F090W, F150W, F277W, and F444W, and find that these PSFs: (i) are generally under-sampled (FWHM ∼ 2 pixel) in F150W and F444W and severely under-sampled (FWHM ∼ 1 pixel) in F090W and F277W; (ii) have significant variation across the field of view, up to ∼15–20 per cent; and (iii) have temporal variations of ∼ 3–4 per cent across multi-epoch exposures. We deployed our PSFs to determine the photometric precision of NIRCam for stars in the crowded, central regions of M 92, measured to be at the ∼0.01 mag level. We use these data to construct the first JWST colour-magnitude diagrams of a globular cluster. Employing existing stellar models, we find that the data reach almost the bottom of the M 92 main sequence (∼0.1 M⊙), and reveal 24 white dwarf candidate members of M 92 in the brightest portion of the white dwarf cooling sequence. The latter are confirmed through a cross-match with archival HST UV and optical data. We also detect the presence of multiple stellar populations along the low-mass main sequence of M 92.

In this work, starting from the well-accepted relations in literature, we introduce a new formalism to compute the astrometric membership probabilities for sources in star clusters, and we provide an application to the case of the open cluster M 37. The novelty of our approach is a refined –and magnitude-dependent– modelling of the parallax distribution of the field stars. We employ the here-derived list of members to estimate the cluster’s mean systemic astrometric parameters, which are based on the most recent Gaia’s catalog (EDR3).