TY - JOUR
T1 - Precision NNLO determination of alpha(s)(M-Z) using an unbiased global parton set
AU - Ball, Richard D.
AU - Bertone, Valerio
AU - Del Debbio, Luigi
AU - Forte, Stefano
AU - Guffanti, Alberto
AU - Latorre, Jose I.
AU - Lionetti, Simone
AU - Rojo, Juan
AU - Ubiali, Maria
N1 - 11 pages, 6 figures. Various small corrections and improvements: Chi2 values for PDF fits provided, discussion of pulls clarified. Final version, to be published in Phys.Lett.B
PY - 2012/1/16
Y1 - 2012/1/16
N2 - We determine the strong coupling alpha(s) at NNLO in perturbative QCD using the global dataset input to the NNPDF2.1 NNLO parton fit: data from neutral and charged current deep-inelastic scattering. Drell-Yan, vector boson production and inclusive jets. We find alpha(s)(M-Z) = 0.1173 +/- 0.0007(stat), where the statistical uncertainty comes from the underlying data and uncertainties due to the analysis procedure are negligible. We show that the distribution of alpha(s) values preferred by different experiments in the global fit is statistically consistent, without need for rescaling uncertainties by a "tolerance" factor. We show that if deep-inelastic data only are used, the best-fit value of alpha(s) is somewhat lower, but consistent within one sigma with the global determination. We find that the shift between the NLO and NNLO values of alpha(s) is Delta alpha(pert)(s) = 0.0018, and we estimate the uncertainty from higher-order corrections to be Delta alpha(NNLO)(s) similar to 0.0009. (C) 2011 Elsevier B.V. All rights reserved.
AB - We determine the strong coupling alpha(s) at NNLO in perturbative QCD using the global dataset input to the NNPDF2.1 NNLO parton fit: data from neutral and charged current deep-inelastic scattering. Drell-Yan, vector boson production and inclusive jets. We find alpha(s)(M-Z) = 0.1173 +/- 0.0007(stat), where the statistical uncertainty comes from the underlying data and uncertainties due to the analysis procedure are negligible. We show that the distribution of alpha(s) values preferred by different experiments in the global fit is statistically consistent, without need for rescaling uncertainties by a "tolerance" factor. We show that if deep-inelastic data only are used, the best-fit value of alpha(s) is somewhat lower, but consistent within one sigma with the global determination. We find that the shift between the NLO and NNLO values of alpha(s) is Delta alpha(pert)(s) = 0.0018, and we estimate the uncertainty from higher-order corrections to be Delta alpha(NNLO)(s) similar to 0.0009. (C) 2011 Elsevier B.V. All rights reserved.
KW - DISTRIBUTIONS
UR - http://www.scopus.com/inward/record.url?scp=84855195463&partnerID=8YFLogxK
U2 - 10.1016/j.physletb.2011.11.053
DO - 10.1016/j.physletb.2011.11.053
M3 - Article
VL - 707
SP - 66
EP - 71
JO - Physics Letters B
JF - Physics Letters B
SN - 0370-2693
IS - 1
ER -