The Dark Energy Survey Supernova Program: Corrections on Photometry Due to Wavelength-dependent Atmospheric Effects

Jaemyoung Lee, Maria Acevedo, Masao Sako, Maria Vincenzi, Dillon Brout, Bruno Sanchez

Abstract: Wavelength-dependent atmospheric effects impact photometric supernova flux measurements for ground-based observations. We present corrections on supernova flux measurements from the Dark Energy Survey Supernova Program’s 5YR sample (DES-SN5YR) for differential chromatic refraction (DCR) and wavelength-dependent seeing, and we show their impact on the cosmological parameters $w$ and $Ω_m$. We use $g − i$ colors of Type Ia supernovae to quantify astrometric offsets caused by DCR and simulate point-spread functions (PSFs) using the GalSIM package to predict the shapes of the PSFs with DCR and wavelength-dependent seeing. We calculate the magnitude corrections and apply them to the magnitudes computed by the DES-SN5YR photometric pipeline. We find that for the DES-SN5YR analysis, not accounting for the astrometric offsets and changes in the PSF shape cause an average bias of +0.2 mmag and −0.3 mmag, respectively, with standard deviations of 0.7 mmag and 2.7 mmag across all DES observing bands ($griz$) throughout all redshifts. When the DCR and seeing effects are not accounted for, we find that $w$ and $Ω_m$ are lower by less than $0.004 \pm 0.02$ and $0.001 \pm 0.01$, respectively, with 0.02 and 0.01 being the $1\sigma$ statistical uncertainties. Although we find that these biases do not limit the constraints of the DES-SN5YR sample, future surveys with much higher statistics, lower systematics, and especially those that observe in the $u$ band will require these corrections as wavelength-dependent atmospheric effects are larger at shorter wavelengths. We also discuss limitations of our method and how they can be better accounted for in future surveys.