Photometric Biases in Modern Surveys. (arXiv:1902.02374v1 [astro-ph.IM])
<a href="http://arxiv.org/find/astro-ph/1/au:+Portillo_S/0/1/0/all/0/1">Stephen K. N. Portillo</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Speagle_J/0/1/0/all/0/1">Joshua S. Speagle</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Finkbeiner_D/0/1/0/all/0/1">Douglas P. Finkbeiner</a>
Most surveys use maximum-likelihood (ML) methods to fit models when
extracting photometry from images. We show these ML estimators systematically
overestimate the flux as a function of the signal-to-noise ratio (SNR) and the
number of model parameters involved in the fit. This bias is substantially
worse for galaxies: while a 1% bias is expected for a 10-sigma point source, a
10-sigma galaxy with a simplified Gaussian profile suffers a 2.5% bias. This
bias also behaves differently depending how multiple bands are used in the fit:
simultaneously fitting all bands leads the flux bias to become roughly evenly
distributed between them, while fixing the position in `non-detection’ bands
(i.e. forced photometry) gives flux estimates in those bands that are biased
low, compounding a bias in derived colors. We show that these effects are
present in idealized simulations, outputs from the HSC fake object pipeline
(SynPipe), and observations from SDSS Stripe 82. Prescriptions to correct for
these biases are provided along with more detailed results related to biases in
ML error estimation.
Most surveys use maximum-likelihood (ML) methods to fit models when
extracting photometry from images. We show these ML estimators systematically
overestimate the flux as a function of the signal-to-noise ratio (SNR) and the
number of model parameters involved in the fit. This bias is substantially
worse for galaxies: while a 1% bias is expected for a 10-sigma point source, a
10-sigma galaxy with a simplified Gaussian profile suffers a 2.5% bias. This
bias also behaves differently depending how multiple bands are used in the fit:
simultaneously fitting all bands leads the flux bias to become roughly evenly
distributed between them, while fixing the position in `non-detection’ bands
(i.e. forced photometry) gives flux estimates in those bands that are biased
low, compounding a bias in derived colors. We show that these effects are
present in idealized simulations, outputs from the HSC fake object pipeline
(SynPipe), and observations from SDSS Stripe 82. Prescriptions to correct for
these biases are provided along with more detailed results related to biases in
ML error estimation.
http://arxiv.org/icons/sfx.gif
