Impact and mitigation of spectroscopic systematics on DESI DR1 clustering measurements

• Main Authors: Krolewski, A, Yu, J, Ross, A. J, Penmetsa, S, Percival, W. J, Zhou, R, Hou, J, Aguilar, J, Ahlen, S, Brooks, D, Chaussidon, E, Claybaugh, T, de la Macorra, A, Dey, Biprateep, Forero-Romero, J. E, Gontcho, S. Gontcho A, Guy, J, Honscheid, K, Juneau, S, Kirkby, D, Kisner, T, Kremin, A, Lambert, A, Le-Guillou, L, Levi, M. E, Martini, P, Meisner, A, Miquel, R, Moustakas, J, Myers, A. D, Newman, J. A, Niz, G, Palanque-Delabrouille, N, Rossi, G, Sanchez, E, Schlafly, E. F, Schlegel, D, Schubnell, M, Seo, H, Sprayberry, D, Tarlé, G, Weaver, B. A, Zhao, C
• Format: Journal Article
• Language: English
• Published: 27-05-2024
• Subjects: Physics - Cosmology and Nongalactic Astrophysics
• DOI: 10.48550/arxiv.2405.17208

The large scale structure catalogs within DESI Data Release 1 (DR1) use
nearly 6 million galaxies and quasars as tracers of the large-scale structure
of the universe to measure the expansion history with baryon acoustic
oscillations and the growth of structure with redshift-space distortions. In
order to take advantage of DESI's unprecedented statistical power, we must
ensure that the galaxy clustering measurements are unaffected by
non-cosmological density fluctuations. One source of spurious fluctuations
comes from variation in galaxy density with spectroscopic observing conditions,
lowering the redshift efficiency (and thus galaxy density) in certain areas of
the sky. We measure the uniformity of the redshift success rate for DESI
luminous red galaxies (LRG), bright galaxies (BGS) and quasars (QSO),
complementing the detailed discussion of emission line galaxy (ELG) systematics
in a companion paper (Yu et al., 2024). We find small but significant
fluctuations of up to 3% in redshift success rate with the effective
spectroscopic signal-to-noise, and create and describe weights that remove
these fluctuations. We also describe the process to identify and remove data
from certain poorly performing fibers from DESI DR1, and measure the stability
of the redshift success rate with time. Finally, we find small but significant
correlations of redshift success rate with position on the focal plane, survey
speed, and number of exposures required, and show the impact of weights
correcting these trends on the power spectrum multipoles and on cosmological
parameters from BAO and RSD fits. These corrections change the best-fit
parameters by $<15\%$ of their statistical errors, and thus contribute
negligibly to the overall DESI error budget.