Constraining primordial non-Gaussianity with DESI 2024 LRG and QSO samples

作者：E. Chaussidon, Christophe Yèche, Arnaud de Mattia, Constantin Payerne, Patrick McDonald, Ashley J. Ross, S. Ahlen, Davide Bianchi, David H. Brooks, E. Burtin, T. Claybaugh, Axel de la Macorra, P. Doel, Simone Ferraro, Andreu Font-Ribera, J. E. Forero-Romero, E. Gaztañaga, Héctor Gil-Marín, Satya Gontcho A Gontcho, G. Gutiérrez, J. Guy, K. Honscheid, Cullan Howlett, Dragan Huterer, R. Kehoe, D. Kirkby, T. Kisner, Anthony Kremin, L. Le Guillou, M. E. Levi, Marc Manera, Aaron Meisner, R. Miquel, John Moustakas, J. A. Newman, Gustavo Niz, N. Palanque‐Delabrouille, Will J. Percival, Francisco Prada, Ignasi Pérez-Ràfols, C. Ravoux, Graziano Rossi, E. Sánchez, D. Schlegel, M. Schubnell, Hee‐Jong Seo, D. Sprayberry, G. Tarlé, M. Vargas-Magaña, Benjamin Alan Weaver, Cheng Zhao, Hu Zou · 发表于：Journal of Cosmology and Astroparticle Physics · 年份：2025 · DOI：10.1088/1475-7516/2025/06/029 · 被引用次数：38 · 研究领域：Cosmology and Gravitation Theories、Computational Physics and Python Applications、Geophysics and Gravity Measurements

Abstract We analyse the large-scale clustering of the Luminous Red Galaxy (LRG) and Quasar (QSO) sample from the first data release (DR1) of the Dark Energy Spectroscopic Instrument (DESI). In particular, we constrain the primordial non-Gaussianity (PNG) parameter f NL loc via the large-scale scale-dependent bias in the power spectrum using 1,631,716 LRGs (0.6 < z < 1.1) and 1,189,129 QSOs (0.8 < z < 3.1). This new measurement takes advantage of the enormous statistical power at large scales of DESI DR1 data, surpassing the latest data release (DR16) of the extended Baryon Oscillation Spectroscopic Survey (eBOSS). For the first time in this kind of analysis, we use a blinding procedure to mitigate the risk of confirmation bias in our results. We improve the model of the radial integral constraint proposing an innovative technique allowing the correction through the window matrix convolution. We also carefully test the mitigation of the dependence of the target selection on the photometry qualities by incorporating an angular integral constraint contribution to the window function, and validate our methodology with the blinded data. Finally, combining the two samples, we measure f NL loc = -3.6 -9.1 +9.0 at 68% confidence, where we assume the universality relation for the LRG sample and a recent merger model for the QSO sample about the response of bias to primordial non-Gaussianity. Adopting the universality relation for the PNG bias in the QSO analysis leads to f...