Improved UVIS Aperture Corrections derived from Focus Diverse PSF Maps

TL;DR

This work addresses the instability of encircled energy for small-aperture photometry in WFC3/UVIS due to telescope breathing and detector-wide PSF variation. It introduces 2D aperture-correction maps derived from focus-diverse PSFs (phylograms) across five UVIS filters, enabling correction of 5–10 pixel photometry for temporal and spatial PSF changes. Validation in both uncrowded (47 Tucanae) and crowded (Omega Centauri) fields shows reduced scatter and improved cross-CCD agreement relative to conventional constant corrections, while EE tables can overestimate corrections for small radii. The study provides practical recommendations to compute EE corrections from science data or PSF cutouts (Methods 1 and 2) and warns against relying on EE tables for small-aperture photometry, with quantified gains in photometric accuracy and precision. Overall, the phylo-based corrections enhance the reliability of precise, small-aperture photometry in UVIS crowded and uncrowded fields.

Abstract

In crowded fields, small-aperture photometry can reduce contamination errors from neighboring sources compared to larger aperture photometry. However, the UVIS encircled energy (EE) varies with detector position and focus variations on orbital timescales for aperture radii less than 10 pixels ($\sim$0.4 arcseconds). Using a set of focus-diverse empirical PSFs by Anderson (2018), we compute 2D spatial maps of the aperture correction between 5-10 pixels and find a maximum change of $\sim$0.01 mag over all focus levels for a given detector position. The upper-left and lower-right corners of the UVIS detector are more focus-sensitive than the rest of the field of view, where the mean correction is systematically $\sim$0.01 mag higher in Amp A for bluer filters (F275W, F336W, F438W) and $\sim$0.01 mag higher in Amp D for redder filters (F606W, F814W) at all focus levels. We test the new aperture correction maps in globular clusters, and we find reduced scatter, better agreement between the two CCDs, and a small shift in the absolute photometry when compared to a single (constant) aperture correction per image. These improvements are specific to photometry with apertures $<$ 10 pixels in radius; results from larger apertures are not affected. Using published EE tables can introduce systematic uncertainties in absolute photometry due to its tendency to vary with detector position and focus level, with larger errors for smaller apertures. Users requiring photometric accuracy better than $\sim$1% for small apertures can use isolated stars in the individual FLT/FLC frames (or PSF cutouts at a similar detector position and focus level) to compute encircled energy corrections and accurately account for the amount of flux at radii larger than their photometric apertures.

Figures (23)

• Figure 1: Adapted from 18Anderson, phylogram plots for F275W, F336W, F438W, F606W, and F814W, along with the phylo (focus) zones identified for each filter. The axes for all filters are arbitrary. The shape of the phylogram and the number of zones varies per filter. In all filters, the nominal phylo level (at best focus where the PSF is the sharpest) is located between levels 5--7.
• Figure 2: 2D spatial maps of the UVIS F438W aperture correction from 5--10 pixels (in magnitudes) derived from PSFs at different phylo (focus) levels. The aperture correction is greatest in the upper-left corner, where Amp A is located. The corrections are the smallest and the most flat across the detector at nominal phylo levels (5--7).
• Figure 3: F438W aperture correction (r = 5--10 pixels) versus phylo level for different regions of the UVIS detector. Colored triangles show a 512$\times$512 pixel cutout at the corner of amplifiers A, B, C, and D, corresponding to subarray apertures 'C512$\ast$-SUB'. Black triangles show the mean at the center of the UVIS FoV in a 1024$\times$1024 pixel cutout, while black circles show the mean over the entire FoV. Dashed lines are plotted for reference only, to compare with the aperture corrections from the published EE tables for UVIS1, Amp A at 0.085 mag (blue) and for UVIS2, Amp C at 0.070 mag (green). For F438W, the aperture correction for UVIS1 (UVIS2) is approximately equal to the mean of the blue (green) triangles over all phylo levels.
• Figure 4: 2D spatial maps of the UVIS aperture correction from 5--10 pixels (in magnitudes) derived from focus-diverse PSF models at different phylo levels for bluer filters: F275W, F336W, and F438W. F336W was divided into only 9 focus groups, so phylo 10 map is blank for this filter.
• Figure 5: Same as Figure \ref{['fig:bluefilts']}, except for redder filters: F606W and F814W. F606W was divided into only 9 focus groups, so phylo 10 map is blank for this filter.