Wu-Ki Tung, Michigan State University
Global analyses of parton distribution functions involve fitting a variety of hard scattering data in order to determine the x dependence of the parton distributions at some input scale Q_0. The pdf's at other values of Q are determined by the appropriate DGLAP evolution equations (see the CTEQ Handbook of Perturbative QCD). For the energy ranges typically available, the proton distributions involve two valence terms (u and d valence), five antiquark terms (ubar, dbar, sbar, cbar, and bbar) and the gluon. The challenge is to find data types that will allow for the (nearly) independent determination of each of these distributions.
In the distant past, one had to assume an SU(3) symmetric sea with the
charm distribution determined by evolution above some threshold. As time
went on, it became possible to disentangle the sbar distribution from the
ubar and dbar terms. Three types of data are currently employed:
1. Inclusive charged current DIS
F_3^{nu}-F_3^{nubar} on an isoscalar target ~ (s-c)
(assuming s=sbar, c=cbar)
2. Charged current/Neutral current DIS comparison
(5/6) F_2^{CC} - 3 F_2^{NC} ~ (s-c)
3. Dimuon production in nu DIS
F_2c^{nu} ~ s
Recent talks by the CCFR/NuTev Collaboration at Fermilab include (Bodek et al., 2000) and (Adams et al., 1999). The challenge for those performing global fits is to include these new data and to seek a common solution for the strange sea.
Traditionally, information on the dbar and ubar distributions has come primarily from low-x DIS and from high mass lepton pair production in hadronic collisions. The latter have mostly involved isoscalar or nearly isoscalar targets, so that differentialtion between ubar and dbar has not been possible. Thus, an SU(2) symmetric sea was often assumed. Now, data from NA-51 at CERN and E-866 at Fermilab have nade it possible to separate the ubar abd dbar distributions. Details can be found in the description of the CTEQ5 global fits (H.L. Lai et al., 2000).
Issues involving the separation of the d and u distributions in the large-x region are addressed in large-x behavior of parton distributions.
Another recent development due to the advent of high statistics neutrino data is the possibility of separately measuring the x-dependence of the s and sbar distributions. The quantum numbers of the proton require that the net number of strange quarks and antiquarks be zero. Nevertheless, the x-dependence of the two distributions need not be the same. See (Adams et al., 1999) for an experimental limit and also the global fit of (Barone, Pascaud, and Zomer, 2000).
Comments on this entry may be sent to Wu-Ki Tung.