1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %\format=latex
3: %\documentstyle[12pt]{article}
4: %\pagestyle{myheadings}
5: %\begin{document}
6: %------------------This is Section 3---------------------------------
7: \mysection{$d\Delta \sigma/dQ$ for the process
8: $p + p\rightarrow \gamma^* +'X'$}
9: %----------------------------------------------------------
10: In this section we will present total cross sections (see Eq. (\ref{eq2.2}))
11: for polarized Drell-Yan production in proton-proton collisions at the RHIC
12: and make a comparison with similar results in previous work. The cross
13: section can be rewritten as
14: \begin{eqnarray}
15: \label{eq4.1}
16: \frac{d\Delta \sigma}{dQ}=\frac{8\,\pi\,\alpha^2}{3\,N\,S\,Q}\,
17: \sum_{i,j=q,\bar q,g}\,\int_{\tau}^1\frac{dy}{y}\,
18: \Phi_{ij}(y,\mu^2)\,\Delta_{ij}\left (\frac{\tau}{y},\frac{Q^2}{\mu^2}\right )
19: \,,
20: \end{eqnarray}
21: where $\tau=Q^2/S$ and $\Phi_{ij}$ is the parton-parton flux defined by
22: \begin{eqnarray}
23: \label{eq4.2}
24: \Phi_{ij}(y,\mu^2)=\int_y^1\frac{du}{u}\,\Delta f_i(u,\mu^2)\,\Delta f_j
25: \left (\frac{y}{u}, \mu^2\right )\,.
26: \end{eqnarray}
27: In particular we study the dependence of the cross section on the input
28: parameters
29: such as the renormalization/factorization scale $\mu$, the virtuality of the
30: photon $Q$ and the input parton densities. At this moment LO and NLO
31: polarized parton densities are available but NNLO are not. Therefore an exact
32: NNLO polarized cross section cannot be determined yet. Even an approximated
33: polarized cross section cannot be given because a finite moment analysis
34: of the anomalous dimensions is not yet available. Only the non-singlet
35: anomalous dimension is known at least up to $N=14$ \cite{rv}. Because of a lack
36: of data the polarized parton density sets show a larger variation than in
37: the unpolarized density sets. This in particular
38: holds for the sea-quark and the gluon densities. For this reason polarized
39: Drell-Yan production is very useful.
40: The sea-quark contribution is dominant for the totally integrated cross
41: section in proton-proton collisions while the gluon contribution dominates
42: the differential distribution w.r.t. $p_T$ for $p_T>Q/2$.
43: We shall use BB set 1 (BB1) for our plots \cite{blbo} in the whole paper
44: except for the longitudinal asymmetry where we shall choose also other
45: density sets like BB set 2 (BB2) \cite{blbo}, the
46: GRSV01 (standard scenario) and GRSV01 (valence
47: scenario) \cite{grsv}. The gluon in BB1 is larger than in BB2. In its turn
48: the gluon in the standard scenario is smaller than in BB2 but larger than in
49: the valence scenario. For all these sets LO and NLO versions exist.
50: It is clear that in the case of LO we will use the one-loop parametrization
51: for the running coupling constant and for NLO the two-loop corrected running
52: coupling constant. However for an estimate of the NNLO corrections
53: we will only choose NLO polarized parton densities with the two-loop running
54: coupling constant.
55: \begin{table}
56: \begin{center}
57: \begin{tabular}{|c|c|c|}\hline
58: GRSV01 (LO, standard scenario)& $\Lambda_4^{\rm LO}=175~{\rm MeV}$ &
59: $\alpha_s^{\rm LO}(M_Z)=0.121$ \\
60: GRSV01 (NLO, standard scenario) & $\Lambda_4^{\rm NLO}=257~{\rm MeV}$ &
61: $\alpha_s^{\rm NLO}(M_Z)=0.109$ \\
62: GRSV01 (LO, valence scenario) & $\Lambda_4^{\rm LO}=175~{\rm MeV}$ &
63: $\alpha_s^{\rm LO}(M_Z)=0.121$ \\
64: GRSV01 (NLO, valence scenario) & $\Lambda_4^{\rm NLO}=257~{\rm MeV}$ &
65: $\alpha_s^{\rm NLO}(M_Z)=0.109$ \\
66: BB (LO, scenario 1) & $\Lambda_4^{\rm LO}=203~{\rm MeV}$ &
67: $\alpha_s^{\rm LO}(M_Z)=0.123$ \\
68: BB (NLO, scenario 1) & $\Lambda_4^{\rm NLO}=235~{\rm MeV}$ &
69: $\alpha_s^{\rm NLO}(M_Z)=0.107$ \\
70: BB (LO, scenario 2) & $\Lambda_4^{\rm LO}=195~{\rm MeV}$ &
71: $\alpha_s^{\rm LO}(M_Z)=0.123$ \\
72: BB (NLO, scenario 2) & $\Lambda_4^{\rm NLO}=240~{\rm MeV}$ &
73: $\alpha_s^{\rm NLO}(M_Z)=0.107$ \\
74: MRST02(LO, lo2002.dat) & $\Lambda_4^{\rm LO}=220~{\rm MeV}$ &
75: $\alpha_s^{\rm LO}(M_Z)=0.125$ \\
76: MRST01(NLO, alf119.dat) & $\Lambda_4^{\rm NLO}=323~{\rm MeV}$ &
77: $\alpha_s^{\rm NLO}(M_Z)=0.113$ \\
78: MRST02(NNLO, vnval1155.dat)& $\Lambda_4^{\rm NLO}=235~{\rm MeV}$ &
79: $\alpha_s^{\rm NNLO}(M_Z)=0.109$ \\
80: \hline
81: \end{tabular}
82: \end{center}
83: \caption{Polarized and unpolarized parton density sets with the values for
84: the QCD scale $\Lambda_4$ and the running coupling $\alpha_s(M_Z)$.}
85: \label{table1}
86: \end{table}
87: The details are given in Table \ref{table1}. Further we put $n_f=4$
88: in the coefficient functions and the running coupling constants and
89: the densities are only presented for the u,d,s and g partons. Finally we also
90: make a comparison with unpolarized Drell-Yan production. For this cross section
91: we adopt for LO the MRST02 \cite{maro1} densities, for NLO the
92: MRST01 \cite{maro2} densities and for NNLO the
93: approximate MRST02 \cite{maro1} densities, with one, two and three
94: loop running coupling constants respectively.
95:
96: In Fig. 4 we have plotted the polarized Drell-Yan cross section in Eq.
97: (\ref{eq4.1}) up to NLO. It is clear that the $q\bar q$ contribution
98: dominates the $qg$ contribution. The former is negative above $Q=$ 25 GeV,
99: while the latter is already negative in a region above $Q$=13 GeV.
100: This implies that
101: the total NLO contribution is actually negative above $Q$=27 GeV where
102: it is very small. For this reason we have plotted the absolute values
103: of the contributions in Fig. 4.
104:
105: The relative sizes are not changed if we go to NNLO as demonstrated in Fig.5.
106: Again the $q\bar q$ contribution is negative above $Q$= 25 GeV and the
107: $qg$ contribution is negative between $Q$= 4 GeV and $Q$= 13 GeV.
108: Here the $qq$ and $gg$ channels appear for the first time and they
109: are both negative. However their contribution is negligible
110: compared with the one given by $qg$ and certainly by the $q\bar q$ result.
111: Again we have plotted the absolute values of the contributions.
112: We conclude that via the $q\bar q$ subprocess the sea-quark contribution
113: dominates the whole cross section in proton-proton collisions
114: irrespective of the value taken by the gluon density.
115:
116: In Fig. 6 we have plotted the absolute values of the LO, NLO and NNLO
117: polarized cross sections for $2<Q<30$ GeV.
118: Notice that we are in the range of small $Q$-values
119: so the $K$-factors are pretty large. We have defined the $K$-factors as
120: \begin{eqnarray}
121: \label{eq4.3}
122: K^{\rm NLO}=\frac{\Delta \sigma^{\rm NLO}}{\Delta \sigma^{\rm LO}}\,,
123: \qquad
124: K^{\rm NNLO}=\frac{\Delta \sigma^{\rm NNLO}}{\Delta \sigma^{\rm LO}}\,,
125: \end{eqnarray}
126: and plotted them in Fig. 7. At $Q=7~{\rm GeV}$ the K-factors reach a minimum,
127: $K^{\rm NLO}\sim 1.2$ and $K^{\rm NNLO}\sim 1.3$, and at larger $Q$ they
128: rise again. The rapid rise
129: above $Q$= 18 GeV is due to the change in sign of the LO process
130: near $Q$=25 GeV where the ratio is infinite. However the magnitude of the
131: polarized cross section is extremely small in this region.
132:
133: Next we turn to the variation of the cross section w.r.t. scale $\mu$.
134: In Figs. 8a, 8b and 8c we show the LO, NLO and NNLO polarized cross sections
135: at the scales $\mu=Q/2$, $\mu=Q$ and $\mu =2Q$. In each figure we see
136: that if $\mu$ gets larger the cross sections increase at
137: small $Q$ but decrease it at larger $Q$. Therefore there are specific
138: values in $Q$ where the scale variations are small and as we go from
139: LO to NLO to NNLO these points are at larger values of $Q$.
140: The scale variation gets smaller as we go from
141: LO to NLO as expected. In spite of the fact that we do not have the
142: NNLO parton densities there is still an improvement in the scale
143: dependence if we go from NLO to NNLO.
144: The corresponding plots for the unpolarized cross
145: section are shown in Figs. 9a, 9b and 9c. These plots also show that when
146: $\mu$ becomes large the cross sections increase at small $Q$
147: and decrease it at larger $Q$. Further we see the expected overall
148: decrease in scale variation as we go from LO to NLO and then to NNLO.
149:
150: To show the scale variation from a different point of view we plot the
151: following
152: quantity
153: \begin{eqnarray}
154: \label{eq4.4}
155: N\left (\frac{\mu}{\mu_0}\right )=\frac{\Delta \sigma(\mu)}{\Delta
156: \sigma(\mu_0)}\,.
157: \end{eqnarray}
158: In Fig. 10a we plot the polarized quantity in Eq. (\ref{eq4.4})
159: for $0.4<\mu/\mu_0<2$
160: at $\mu_0=Q=$ 5 GeV. In this figure we see
161: an improvement in the scale variation while going from LO to NLO
162: and then to NNLO.
163: This feature also persists at higher $Q$-values i.e. $Q=$ 10 GeV in Fig. 10b
164: and $Q=$ 15 GeV in Fig. 10c.
165: For the unpolarized cross section the trend is the same. Here
166: we have the NNLO parton densities and
167: in Fig. 11a at $Q=$ 5 GeV we see an improvement in scale variation
168: while going to higher order. The same observation can be made in Fig. 11b at
169: $Q=$ 10 GeV. However for $Q=$ 15 GeV in Fig. 11c the NLO curve becomes
170: slightly worse than the LO one. While on the other hand the NNLO curve
171: is slightly better than the LO at least for larger scales $\mu$.
172:
173: Now we look at effect of the higher order corrections on the longitudinal
174: asymmetry defined by
175: \begin{eqnarray}
176: \label{eq4.5}
177: A_{LL}=\frac{\Delta \sigma}{\sigma}\,.
178: \end{eqnarray}
179: In Fig. 12 we have plotted $A_{LL}$ in percent
180: for BB1 in the range $2<Q<30$ GeV
181: in LO, NLO and NNLO. Since the polarized cross section changes sign
182: we see that the asymmetry is negative for large $Q$ (note the displaced zero).
183: It is obvious that the longitudinal
184: asymmetry is small reaching about one percent between $Q=$ 10 GeV
185: and $Q=$ 20 GeV.
186:
187: To show the predictions of the various polarized parton density sets
188: we compare the BB1 NLO contribution with the other three NLO
189: parton density sets i.e. BB2, GRSV01 (standard scenario or SS) and GRSV01
190: (valence scenario or VS) in Fig 13. Obviously the first three sets have
191: small positive longitudinal asymmetries while the GRSV01 valence scenario
192: set has a large negative asymmetry. Therefore the set with the smallest
193: gluon, here GRSV01 (valence scenario), has the largest asymmetry. On the other
194: hand the set with the largest gluon, here BB1, has the smallest asymmetry.
195: This feature is unchanged in NNLO as can be seen in Fig. 14.
196: Hopefully the RHIC experiments will get sufficient luminosities and
197: large enough proton polarizations to distinguish between the first
198: three predictions and the last one.
199:
200: To summarize we have computed the NNLO corrections to the production
201: of massive lepton-pairs in polarized proton-proton collisions.
202: Hence as far as coefficient functions as concerned we have the NNLO
203: corrections to both polarized and unpolarized processes. Unfortunately
204: the corresponding anomalous dimensions of the partons are not available
205: at NNLO. While fits have been made in the unpolarized case it is not possible
206: yet in the polarized case. Therefore we have convoluted the NLO polarized
207: densities with the NNLO coefficient functions to get an idea of the
208: stability of the perturbation series. First we have seen that the
209: dominance of the sea-quark initiated process, which holds at LO and NLO,
210: also prevails at NNLO. For the unpolarized cross section we have shown
211: that the addition of the NNLO terms produces a significant reduction in the
212: scale variation of the cross section. However the same phenomenon is
213: also shown for the polarized cross section in spite of the fact
214: that the NNLO parton densities are not known yet.
215: Finally
216: the longitudinal asymmetry is small and positive for the polarized
217: parton density sets BB1, BB2 and GRSV(SS). If one uses the GRSV(VS)
218: densities then the longitudinal polarization is negative and large.
219:
220:
221: %\end{document}
222: