References - Bogoliubov Laboratory of Theoretical Physics - JINR

POLARIZED PDFs: REMARKS ON METHODS OF QCD ANALYSIS OF
THE DATA
E. Leader 1 , A.V. Sidorov 2,† and D.B. Stamenov 3
(1) Imperial College London, Prince Consort Road, London SW7 2BW, England
(2) Theoretical Laboratory, Joint Institute for Nuclear Research, Dubna, Russia
(3) Institute for Nuclear Research and Nuclear Energy, Bulgarian Academy of Sciences,
Sofia, Bulgaria
† E-mail: sidorov@theor.jinr.ru
Abstract
The results on polarized parton densities (PDFs) obtained using different methods
of QCD analysis of the present polarized DIS data are discussed. It is pointed
out that the precise data in the preasymptotic region require a more careful matching
of the QCD predictions to the data in this region in order to determine the
polarized PDFs correctly.
In this talk we discuss how the results on polarized PDFs depend on the method used
in the QCD analysis, accounting or not for the kinematic and dynamic 1/Q 2 corrections
to the spin structure function g1.
One of the peculiarities of polarized DIS is that more than a half of the present data
are at moderate Q 2 and W 2 (Q 2 ∼ 1 − 4GeV 2 , 4GeV 2 < W 2 < 10 GeV 2 ), or in the socalled
preasymptotic region. So, in contrast to the unpolarized case, this region cannot be
excluded from the analysis, and the role of the 1/Q 2 terms (kinematic - γ 2 factor, target
mass corrections, and dynamic - higher twist corrections to the spin structure function
g1) in the determination of the polarized PDFs has to be investigated.
The best manner to determine the polarized PDFs is to perform a QCD fit to the
data on g1/F1, which can be obtained if both the A|| and A⊥ asymmetries are measured.
The data on the photon-nucleon asymmetry A1 are not suitable for the determination of
PDFs because the structure function g2 is not well known in QCD and the approximation
(A1) theor = g1/F1 − γ 2 g2/F1 ≈ (g1/F1) theor
used by some of the groups in the preasymptotic region, where γ 2 cannot be neglected, is
not reasonable.
In QCD, one can split g1 and F1 into leading (LT) and dynamical higher twist (HT)
pieces
g1 =(g1)LT,TMC +(g1)HT, F1 =(F1)LT,TMC +(F1)HT. (2)
In the LT pieces in Eq. (2) the calculable target mass corrections (TMC) are included
g1(x, Q 2 )LT,TMC = g1(x, Q 2 )LT + g1(x, Q 2 )TMC, (3)
F1(x, Q 2 )LT,TMC = F1(x, Q 2 )LT + F1(x, Q 2 )TMC.
126
(1)