DESI Deep Spectroscopy for Photometric Redshift Training and Calibration for LSST


Biprateep Dey

The Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) will depend critically on estimates of redshifts based only on imaging -- i.e., photometric redshifts (photo-z's) -- to enable cosmological measurements. Secure spectroscopic measurements of redshifts for faint samples will enable photo-z improvements for Rubin, maximizing the ability of LSST to constrain fundamental physics (as highlighted in the Snowmass Cosmic Frontiers reports).   Deep spectroscopic samples can be used to increase the performance of photo-z algorithms, yielding correspondingly large gains in the cosmological constraining power of Rubin data, while simultaneously reducing uncertainties on the overall redshift distributions of LSST samples, mitigating what may in many cases be a limiting systematic effect. We have now conducted pilot observations with DESI and find that it can measure redshifts at a rate comparable to past surveys with 10m class telescopes in only 50% more observing time (>4x faster than would be expected from scaling aperture area alone) while offering much higher multiplexing and covering a broader redshift range. DESI could thus provide the definitive optical redshift sample for targets down to the depths that LSST will reach in its early years with only a modest total time investment.  Extrapolating from our results, a Stage V Spectroscopic Facility with DESI-like instrumental performance could provide a baseline training sample for the full Rubin LSST dataset in as few as 50 dark nights. This pilot study was done by the photo-topical group of the DESI collaboration. We shall also introduce the group to the Rubin community and use the PCW to foster future collaborations between the two communities.


This talk will be given in the Synergies Between LSST and New Spectroscopic Surveys session.


Career Stage: 
Grad Student