Update of the global fit of PDFs including the low-Q DIS data

Abstract

We perform the next-to-next-leading-order (NNLO) QCD global fit of PDFs using inclusive charged-lepton and neutrino DIS data down to Q = 1 GeV. We also consider the data on neutrino-nucleon dimuon production, that allows us to disentangle the strange sea distribution. The fit results in χ 2 /NDP = 5150/4338 = 1.2 that demonstrates a good consistency of the data sets used in analysis. The resulting value of αs(MZ) = 0.1136 ± 0.0007(exp.) is in a good agreement with the previous version of the fit with a more stringent cut on Q. This analysis allows us to improve the accuracy of PDFs. The HT terms of the neutrino-nucleon structure functions F2 and xF3 are determined, the former is found to be consistent with one for the charged-leptons if the charge factor is taken into account.

Precision determination of electroweak parameters and the strange content of the proton from neutrino deep-inelastic scattering

“…The average value of the χ 2 we obtain is in good agreement with that expected on the basis of the "effective number of degrees of freedom" published in Ref. [25], and with other fits to the same data [14].…”

“…A simultaneous extraction of this parameter along with the determination of strangeness in Ref. [14] leads to a similar 8 result. In the determination of the dimuon cross section, the branching ratio will be set equal to the central value used in the NuTeV analysis [10].…”

“…We find that the shape of the strange and antistrange distributions which are compatible with data are rather more general than those obtained in other recent studies [11][12][13][14][15]17]. Our uncertainty on the ratio K S = [S + ] / Ū +D of strange to light sea momenta is rather more asymmetric than hitherto assumed: K S Q 2 = 20 GeV 2 = 0.71 +0.19 −0.31 stat .…”

“…(3) depends on the branching ratio Br (D → µ) . The uncertainty in this is actually rather significant: in previous analyses [10,14] of dimuon data this turned out to be one of the dominant sources of uncertainty. To take account of this uncertainty, the value of the branching ratio used in the fit has been randomized about its central value, analogously to the procedure used for the preprocessing exponents, with a Gaussian distribution of width equal to the stated uncertainty Br (D → µ) = 0.099 ± 0.012 [10].…”

“…An earlier measurement of the dimuon cross section using the same detector (but a different beam-line) was performed by the CCFR collaboration [28]. This previous measurement is significantly less accurate and its compatibility with the NuTeV data is debatable [8,14]; we will not include it in our fit. A recent measurement of the dimuon cross section has also been performed by the CHORUS collaboration [9]; unfortunately, however, only the results of a leading-order QCD analysis of this data have been published, and not the cross-section data themselves, which therefore cannot be used in our analysis.…”