BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Bayesian Component Separation for DESI LAE Automated Spectroscopic
  Redshifts and Photometric Targeting - Ana Sofía Uzsoy (Harvard)
DTSTART:20250721T150000Z
DTEND:20250721T160000Z
UID:TALK234427@talks.cam.ac.uk
CONTACT:65128
DESCRIPTION:Lyman Alpha Emitters (LAEs) are valuable high-redshift cosmolo
 gical probes traditionally targeted with specialized narrow-band photometr
 ic surveys. In ground-based spectroscopy\, it can be difficult to distingu
 ish the sharp LAE peak from residual sky emission lines\, leading to miscl
 assified redshifts. We present a Bayesian spectral component separation te
 chnique to automatically determine spectroscopic redshifts for LAEs while 
 marginalizing over sky residuals. We use visually inspected DESI (Dark Ene
 rgy Spectroscopic Instrument) LAE targets to create a data-driven prior an
 d can determine redshift by jointly inferring sky residual\, LAE\, and res
 idual components for each individual spectrum. We demonstrate this method 
 on 910 photometrically targeted z = 2-4 DESI LAE candidate spectra and det
 ermine their redshifts with >90% accuracy compared to visually inspected r
 edshifts. Using the chi-squared value from our pipeline as a proxy for det
 ection confidence\, we then explore potential survey design choices and im
 plications for targeting LAEs with medium-band photometry. This method all
 ows for scalability and accuracy in determining spectroscopic redshifts in
  DESI and the results provide recommendations for LAE targeting in anticip
 ation of future high-redshift spectroscopic surveys\, such as DESI-2.
LOCATION:Martin Ryle Seminar Room\, KICC
END:VEVENT
END:VCALENDAR