BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Maximum-Likelihood Biases in PSF and Model-Fitting Photometry - St
 ephen Portillo (University of Washington)
DTSTART:20191216T113000Z
DTEND:20191216T123000Z
UID:TALK136192@talks.cam.ac.uk
CONTACT:88879
DESCRIPTION:Many surveys use maximum-likelihood (ML) methods to fit models
  when extracting photometry from images. We show these ML estimators syste
 matically overestimate the flux as a function of the signal-to-noise ratio
  and the number of model parameters involved in the fit. This bias is subs
 tantially worse for resolved: while a 1% bias is expected for a 10 sigma p
 oint source\, a 10 sigma galaxy with a simplified Gaussian profile suffers
  a 2.5% bias. This bias also behaves differently depending how multiple ba
 nds are used in the fit: simultaneously fitting all bands leads the flux b
 ias to become roughly evenly distributed between them\, while fixing the p
 osition in “non-detection” bands (i.e. forced photometry) gives flux e
 stimates in those bands that are biased low\, compounding a bias in derive
 d colors. We show that these effects are present in idealized simulations\
 , Hyper Suprime-Cam fake object pipeline (SynPipe)\, and observations from
  Sloan Digital Sky Survey Stripe 82.
LOCATION:Large Martin Ryle Seminar Room\, KICC
END:VEVENT
END:VCALENDAR