(Color online) (b) Integrated pion $R_{AA}$ over the range $5<p_{T}<8$ GeV/c versus $N_{part}$. The bands represent the systematic uncertainty on ratios. An additional scale error due to $pp$ normalization (∼14%) is not shown.

(Color online) (a) Nuclear modification factor, RAA, for protons and anti-protons ($p + \overline{p}$) in Cu+Cu (filled symbols) and Au+Au (open symbols) collisions at $\sqrt{s_{NN}}$=200 GeV. Error band is shown for most central Cu+Cu data to represent characteristic systematic uncertainties for Cu+Cu data. Error boxes at $R_{AA}$=1 represent Cu+Cu scale uncertainties due to the number of collisions and from pp spectra normalization.

(Color online) (a) Ratio of protons and anti-protons to charged pions, versus transverse momentum, for Cu+Cu collisions at $\sqrt{s_{NN}}$=200 GeV. For clarity, only one systematic error band is shown for Cu+Cu data (most central events), uncertainties on data in other centrality bins are similar in magnitude. Systematic errors for Au+Au data are not shown, for details see [9].

(Color online) (b) Average $(p+\overline{p})/(\pi^{+}+\pi^{-})$ ratio for $3<p_{T}<4$ GeV/c is shown as function of centrality ($N_{part}$) for Cu+Cu and Au+Au data. Error band represents the $pp$ measurement with uncertainty.

Fig. 2. (Color online.) Event-by-event photon multiplicity distributions (solid circles) for $\mathrm{Au}+\mathrm{Au}$ and $\mathrm{Cu}+\mathrm{Cu}$ at $\sqrt{s_{\mathrm{NN}}}=62.4$ and $200 \mathrm{GeV} .$ The distributions for top $0-5 \%$ central $\mathrm{Au}+$ Au collisions and top $0-10 \%$ central $\mathrm{Cu}+\mathrm{Cu}$ collisions are also shown (open circles). The photon multiplicity distributions for central collisions are observed to be Gaussian (solid line). Only statistical errors are shown. NOTE: For points with invisible error bars, the point size was considered as an absolute upper limit for the uncertainty.

Fig. 2. (Color online.) Event-by-event photon multiplicity distributions (solid circles) for $\mathrm{Au}+\mathrm{Au}$ and $\mathrm{Cu}+\mathrm{Cu}$ at $\sqrt{s_{\mathrm{NN}}}=62.4$ and $200 \mathrm{GeV} .$ The distributions for top $0-5 \%$ central $\mathrm{Au}+$ Au collisions and top $0-10 \%$ central $\mathrm{Cu}+\mathrm{Cu}$ collisions are also shown (open circles). The photon multiplicity distributions for central collisions are observed to be Gaussian (solid line). Only statistical errors are shown. NOTE: For points with invisible error bars, the point size was considered as an absolute upper limit for the uncertainty.

Fig. 2. (Color online.) Event-by-event photon multiplicity distributions (solid circles) for $\mathrm{Au}+\mathrm{Au}$ and $\mathrm{Cu}+\mathrm{Cu}$ at $\sqrt{s_{\mathrm{NN}}}=62.4$ and $200 \mathrm{GeV} .$ The distributions for top $0-5 \%$ central $\mathrm{Au}+$ Au collisions and top $0-10 \%$ central $\mathrm{Cu}+\mathrm{Cu}$ collisions are also shown (open circles). The photon multiplicity distributions for central collisions are observed to be Gaussian (solid line). Only statistical errors are shown. NOTE: For points with invisible error bars, the point size was considered as an absolute upper limit for the uncertainty.

Fig. 3. (Color online.) Photon pseudorapidity distributions for $\mathrm{Au}+\mathrm{Au}$ and $\mathrm{Cu}+\mathrm{Cu}$ at $\sqrt{\mathrm{s}_{\mathrm{NN}}}=62.4$ and $200\mathrm{GeV}$. The results for several centrality classes are shown. The solid curves are results of HIJING simulations for central $(0-5 \%$ for $\mathrm{Au}+\mathrm{Au}$ and $0-10 \%$ for $\mathrm{Cu}+\mathrm{Cu})$ and $30-40 \%$ mid-central collisions. The errors shown are systematic, statistical errors are negligible in comparison. NOTE: For points with invisible error bars, the point size was considered as an absolute upper limit for the uncertainty.

Fig. 3. (Color online.) Photon pseudorapidity distributions for $\mathrm{Au}+\mathrm{Au}$ and $\mathrm{Cu}+\mathrm{Cu}$ at $\sqrt{\mathrm{s}_{\mathrm{NN}}}=62.4$ and $200\mathrm{GeV}$. The results for several centrality classes are shown. The solid curves are results of HIJING simulations for central $(0-5 \%$ for $\mathrm{Au}+\mathrm{Au}$ and $0-10 \%$ for $\mathrm{Cu}+\mathrm{Cu})$ and $30-40 \%$ mid-central collisions. The errors shown are systematic, statistical errors are negligible in comparison. NOTE: For points with invisible error bars, the point size was considered as an absolute upper limit for the uncertainty.

Fig. 3. (Color online.) Photon pseudorapidity distributions for $\mathrm{Au}+\mathrm{Au}$ and $\mathrm{Cu}+\mathrm{Cu}$ at $\sqrt{\mathrm{s}_{\mathrm{NN}}}=62.4$ and $200\mathrm{GeV}$. The results for several centrality classes are shown. The solid curves are results of HIJING simulations for central $(0-5 \%$ for $\mathrm{Au}+\mathrm{Au}$ and $0-10 \%$ for $\mathrm{Cu}+\mathrm{Cu})$ and $30-40 \%$ mid-central collisions. The errors shown are systematic, statistical errors are negligible in comparison. NOTE: For points with invisible error bars, the point size was considered as an absolute upper limit for the uncertainty.

Fig. 3. (Color online.) Photon pseudorapidity distributions for $\mathrm{Au}+\mathrm{Au}$ and $\mathrm{Cu}+\mathrm{Cu}$ at $\sqrt{\mathrm{s}_{\mathrm{NN}}}=62.4$ and $200\mathrm{GeV}$. The results for several classes are shown. The solid curves are results of HIJING simulations for central $(0-5 \%$ for $\mathrm{Au}+\mathrm{Au}$ and $0-10 \%$ for $\mathrm{Cu}+\mathrm{Cu})$ and $30-40 \%$ mid-central collisions. The errors shown are systematic, statistical errors are negligible in comparison. NOTE: For points with invisible error bars, the point size was considered as an absolute upper limit for the uncertainty.

Fig. 4. (Color online.) Top panel: The number of photons divided by $\left\langle N_{\text{part}}\right\rangle / 2$ as a function of average number of participating nucleons for $\mathrm{Au}+\mathrm{Au}$ and $\mathrm{Cu}+\mathrm{Cu}$ at $\sqrt{\mathrm{s} _{\mathrm{NN}}}=62.4$ and $200 \mathrm{GeV}$ for $-3.7<\eta<-2.3 .$ Errors shown are systematic only and include those for $\left\langle N_{\text {part }}\right\rangle .$ Results from HIJING are shown as lines (solid for $\mathrm{Au}+\mathrm{Au}$ and dashed for $\mathrm{Cu}+\mathrm{Cu}$ ). Bottom panel: Same as above, for both photons and charged particles from PHOBOS [10]. NOTE: For points with invisible error bars, the point size was considered as an absolute upper limit for the uncertainty.

Fig. 4. (Color online.) Top panel: The number of photons divided by $\left\langle N_{\text{part}}\right\rangle / 2$ as a function of average number of participating nucleons for $\mathrm{Au}+\mathrm{Au}$ and $\mathrm{Cu}+\mathrm{Cu}$ at $\sqrt{\mathrm{s} _{\mathrm{NN}}}=62.4$ and $200 \mathrm{GeV}$ for $-3.7<\eta<-2.3 .$ Errors shown are systematic only and include those for $\left\langle N_{\text {part }}\right\rangle .$ Results from HIJING are shown as lines (solid for $\mathrm{Au}+\mathrm{Au}$ and dashed for $\mathrm{Cu}+\mathrm{Cu}$ ). Bottom panel: Same as above, for both photons and charged particles from PHOBOS [10]. NOTE: For points with invisible error bars, the point size was considered as an absolute upper limit for the uncertainty.

Fig. 5. (Color online.) Photon pseudorapidity distributions normalized by the average number of participating nucleon pairs for different collision centralities are plotted as a function of pseudorapidity shifted by the beam rapidity (-5.36 for $200 \mathrm{GeV}$ and -4.19 for $62.4 \mathrm{GeV}$ ) for $\mathrm{Au}+\mathrm{Au}$ and $\mathrm{Cu}+\mathrm{Cu}$ collisions at $\sqrt{s_{\mathrm{NN}}}=62.4$ and $200 \mathrm{GeV}$. Errors are systematic only, statistical errors are negligible in comparison. For clarity of presentation, results for only four centralities are shown. The $\mathrm{Cu}+$ Cu data are shifted by 0.1 unit in $\eta-y_{\text {beam }} .$ The solid line is a second order polynomial fit to the data (see text for details).

Fig. 5. (Color online.) Photon pseudorapidity distributions normalized by the average number of participating nucleon pairs for different collision centralities are plotted as a function of pseudorapidity shifted by the beam rapidity (-5.36 for $200 \mathrm{GeV}$ and -4.19 for $62.4 \mathrm{GeV}$ ) for $\mathrm{Au}+\mathrm{Au}$ and $\mathrm{Cu}+\mathrm{Cu}$ collisions at $\sqrt{s_{\mathrm{NN}}}=62.4$ and $200 \mathrm{GeV}$. Errors are systematic only, statistical errors are negligible in comparison. For clarity of presentation, results for only four centralities are shown. The $\mathrm{Cu}+$ Cu data are shifted by 0.1 unit in $\eta-y_{\text {beam }} .$ The solid line is a second order polynomial fit to the data (see text for details).

$R_{AA}$ as a function of $p_{T}$ from $0-5\%$ and $60-80\%$ central Au+Au events for $\Lambda$. Stat. and syst. errors added in quadrature.The band at unity shows the systematic uncertainty on $⟨N_{bin}⟩$. The dashed line below unity shows the expected value of RAA should the yields scale with $⟨N_{part}⟩$, and the band around it shows the systematic uncertainty on ⟨Npart ⟩. $N_{bin}$ Scaling 1 $\pm$ 0.16 and $N_{part}$ Scaling 0.168 $\pm$ 0.02 .

$R_{AA}$ as a function of $p_{T}$ from $0-5\%$ and $60-80\%$ central Au+Au events for $\Xi^{-} + \bar{\Xi}^{+}$. Stat. and syst. errors added in quadrature. The band at unity shows the systematic uncertainty on $⟨N_{bin}⟩$. The dashed line below unity shows the expected value of RAA should the yields scale with $⟨N_{part}⟩$, and the band around it shows the systematic uncertainty on ⟨Npart ⟩. $N_{bin}$ Scaling 1 $\pm$ 0.16 and $N_{part}$ Scaling 0.168 $\pm$ 0.02 .

$R_{AA}$ as a function of $p_{T}$ from $0-5\%$ and $60-80\%$ central Au+Au events for inclusive $p+\bar{p}$. Stat. and syst. errors added in quadrature. The band at unity shows the systematic uncertainty on $⟨N_{bin}⟩$. The dashed line below unity shows the expected value of RAA should the yields scale with $⟨N_{part}⟩$, and the band around it shows the systematic uncertainty on ⟨Npart ⟩. $N_{bin}$ Scaling 1 $\pm$ 0.16 and $N_{part}$ Scaling 0.168 $\pm$ 0.02 .

$R_{AA}$ from HIJING as a function of $p_{T}$ from $0-5\%$ central Au+Au events for inclusive $p+\bar{p}$.

$R_{AA}$ from HIJING as a function of $p_{T}$ from $0-5\%$ central Au+Au events for $\Lambda$.

$R_{AA}$ from HIJING as a function of $p_{T}$ from $0-5\%$ central Au+Au events for $\Xi^{-}+\bar{\Xi}^{+}$.

$R_{AA}$ from EPOS as a function of $p_{T}$ from $0-5\%$ central Au+Au events for inclusive $p+\bar{p}$.

$R_{AA}$ from EPOS as a function of $p_{T}$ from $0-5\%$ central Au+Au events for $\Lambda$.

$R_{AA}$ from EPOS as a function of $p_{T}$ from $0-5\%$ central Au+Au events for $\Xi^{-}+\bar{\Xi}^{+}$.

$R_{AA}$ from HIJING as a function of $p_{T}$ from $60-80\%$ central Au+Au events for inclusive $p+\bar{p}$.

$R_{AA}$ from HIJING as a function of $p_{T}$ from $60-80\%$ central Au+Au events for $\Lambda$.

$R_{AA}$ from HIJING as a function of $p_{T}$ from $60-80\%$ central Au+Au events for $\Xi^{-}+\bar{\Xi}^{+}$.

$R_{AA}$ from EPOS as a function of $p_{T}$ from $60-80\%$ central Au+Au events for inclusive $p+\bar{p}$.

$R_{AA}$ from EPOS as a function of $p_{T}$ from $60-80\%$ central Au+Au events for $\Lambda$.

$R_{AA}$ from EPOS as a function of $p_{T}$ from $60-80\%$ central Au+Au events for $\Xi^{-}+\bar{\Xi}^{+}$.

Efficiency corrected $t_{\perp}$ spectrum for $\rho^{0}$ from (top) minium bias and (bottom) topology data, for mid-rapidity (left) and larger rapidity (right) samples. The points are the data, while the solid lines are the results of fits to Eq. (3).

Efficiency corrected $t_{\perp}$ spectrum for $\rho^{0}$ from (top) minium bias and (bottom) topology data, for mid-rapidity (left) and larger rapidity (right) samples. The points are the data, while the solid lines are the results of fits to Eq. (3).

Efficiency corrected $t_{\perp}$ spectrum for $\rho^{0}$ from (top) minium bias and (bottom) topology data, for mid-rapidity (left) and larger rapidity (right) samples. The points are the data, while the solid lines are the results of fits to Eq. (3).

Efficiency corrected $t_{\perp}$ spectrum for $\rho^{0}$ from (top) minium bias and (bottom) topology data, for mid-rapidity (left) and larger rapidity (right) samples. The points are the data, while the solid lines are the results of fits to Eq. (3).

xT scaling (with xT.

J/psi nuclear modification factor RAA from sqrt(s).

J/psi nuclear modification factor RAA from sqrt(s).

J/psi-hadron azimuthal correlations in sqrt(s).

Contribution from B-hadron feed-down to inclusive J/psi production in sqrt(s).

$\langle cos(\phi_{\alpha}+\phi_{\beta}−2\phi_{c})\rangle$ as a function of reference multiplicity for different charge combinations, after corrections for acceptance effects. In the legend the signs indicate the charge of particles $\alpha$, $\beta$, and c. The results shown are for Au+Au collisions at 200 GeV obtained in the Full Field.

$\langle cos(\phi_{\alpha}-\phi_{\beta})\rangle$ as a function of centrality for different charge combinations and FF and RFF configurations. The data points corresponding to different charge and field configurations are slightly shifted in the horizontal direction with respect to each other for clarity. The error bars are statistical. Also shown are model predictions described in Section VII.

A comparison of the correlations obtained by selecting the third particle from the main TPC or from the Forward TPCs. The TPC and FTPC points are shifted horizontally relative to one another for clarity purposes. The error bars are statistical.

A comparison of the correlations obtained by selecting the third particle from the main TPC or from the Forward TPCs after scaling by the flow of the third particle. The shaded areas represent the uncertainty from $v_{2,c}$ scaling (see text for details). The TPC and FTPC points are shifted horizontally relative to one another for clarity purposes. The error bars are statistical.

$\langle cos(\phi_{\alpha}+\phi_{\beta}−2\Psi_{RP})\rangle$ in Au+Au and Cu+Cu collisions at $\sqrt{s_{NN}}$ = 200 GeV calculated using Eq. 7. The error-bars show the statistical errors. The shaded area reflects the uncertainty in the elliptic flow values used in calculations, with lower (in magnitude) limit obtained with elliptic flow from two-particle correlations and upper limit from four-particle cumulants. For details, see Section IV. Thick solid (Au+Au) and dashed (Cu+Cu) lines represent possible non-reaction-plane dependent contribution from many-particle clusters as estimated by HIJING, see Section VII A.

$\langle cos(\phi_{\alpha}+\phi_{\beta}−2\Psi_{RP})\rangle$ in Au+Au and Cu+Cu collisions at $\sqrt{s_{NN}}$ = 62 GeV calculated using Eq. 7. The error-bars show the statistical errors. The shaded area reflects the uncertainty in the elliptic flow values used in calculations. For details, see Section IV. Thick solid (Au+Au) and dashed (Cu+Cu) lines represent possible non-reaction-plane dependent contribution from many-particle clusters as estimated by HIJING, see Section VII A.

Au+Au and Cu+Cu collisions at $\sqrt{s_{NN}}$ = 200 GeV. The correlations are scaled with the number of participants and are plotted as function of centrality. The error-bars indicate the statistical errors. The shaded area reflects the uncertainty in the elliptic flow values used in calculations. For details, see Section IV. Thick solid (Au+Au) and dashed (Cu+Cu) lines represent possible non-reaction-plane dependent contribution from many-particle clusters as estimated by HIJING, see Section VII A.

Au+Au and Cu+Cu collisions at $\sqrt{s_{NN}}$ = 200 GeV. The correlations are scaled with the number of participants and are plotted as function of number of participants. The error-bars indicate the statistical errors. The shaded area reflects the uncertainty in the elliptic flow values used in calculations. For details, see Section IV. Thick solid (Au+Au) and dashed (Cu+Cu) lines represent possible non-reaction-plane dependent contribution from many-particle clusters as estimated by HIJING, see Section VII A.

Au+Au at 200 GeV. The correlations dependence on pseudorapidity separation $\Delta\eta$ = |$\eta_{\alpha}$ − $\eta_{\beta}$| for centrality 30-50%. The shaded band indicates uncertainty associated with $v_{2}$ measurements and has been calculated using two- and four-particle cumulant results as the limits.

Au+Au at 200 GeV. The correlations dependence on pseudorapidity separation $\Delta\eta$ = |$\eta_{\alpha}$ − $\eta_{\beta}$| for centrality 10-30%. The shaded band indicates uncertainty associated with $v_{2}$ measurements and has been calculated using two- and four-particle cumulant results as the limits.

Au+Au at 200 GeV. The correlations dependence on (p$_{t,\alpha}$ + p$_{t,\beta}$)/2 for centrality 30-50%. The shaded band indicates uncertainty associated with $v_{2}$ measurements and has been calculated using two- and four-particle cumulant results as the limits.

Au+Au at 200 GeV. The correlations dependence on (p$_{t,\alpha}$ + p$_{t,\beta}$)/2 for centrality 10-30%. The shaded band indicates uncertainty associated with $v_{2}$ measurements and has been calculated using two- and four-particle cumulant results as the limits.

Au+Au at 200 GeV. The correlations dependence on |p$_{t,\alpha}$ - p$_{t,\beta}$| for centrality 30-50%. The shaded band indicates uncertainty associated with $v_{2}$ measurements and has been calculated using two- and four-particle cumulant results as the limits.

Au+Au at 200 GeV. The correlations dependence on |p$_{t,\alpha}$ - p$_{t,\beta}$| for centrality 10-30%. The shaded band indicates uncertainty associated with $v_{2}$ measurements and has been calculated using two- and four-particle cumulant results as the limits.

Three-particle correlator in Au+Au and Cu+Cu collisions compared to HIJING calculations shown as thick lines. All three particles are taken in the main TPC region, |$\eta$| < 1.0. The correlator has been scaled with number of participants and are plotted versus number of participants. In this representation HIJING results for Au+Au and Cu+Cu collisions coincide in the region of overlap.

$\langle cos(\phi_{\alpha} + \phi_{\beta} − 2\Psi_{RP})\rangle$ calculated for 200 GeV Au+Au events with event generators HIJING (with and without an “elliptic flow afterburner”), UrQMD, and MEVSIM. Blue symbols mark opposite-charge correlations, and red are same-charge. Solid stars represent the values from the data to facilitate comparison. Acceptance cuts of 0.15 < p$_{t}$ < 2 GeV/c and |$\eta$| < 1.0 were used in all cases. For MEVSIM, HIJING, and UrQMD points the true reaction plane from the generated event was used for $\Psi_{RP}$ . Thick solid lighter colored lines represent possible non-reactionplane dependent contribution from many-particle clusters as estimated by HIJING and discussed in Section VII A. Corresponding estimates from UrQMD are about factor of two smaller.

Upper panels. $N_{\scriptsize{\mbox{part}}}$ scaled ($R^{N_{\scriptsize{\mbox{part}}}}_{AA}$) nuclear modification factors as a function of $p_{T}$ of $\phi$ mesons for $0-10\%$ and $20-30\%$ $Cu+Cu$ and $Au+Au$ collisions at $\sqrt{s_{NN}}=200$ GeV. Lower panel. Same as above for $N_{\scriptsize{\mbox{bin}}}$ scaled ($R^{N_{\scriptsize{\mbox{bin}}}}_{AA}$) nuclear modification factor. The error bars represent the statistical and systematic errors added in quadrature. The shaded band in upper panel around 1 at $p_{T}=4.5-5.5$ GeV/$c$ in the right side reflects the uncertainty in $N_{\scriptsize{\mbox{part}}}$ and that on the lower panel for $N_{\scriptsize{\mbox{bin}}}$ calculation for central $Au+Au$ collisions. The respective uncertainties for central $Cu+Cu$ collisions are of similar order.

Upper panels. $N_{\scriptsize{\mbox{part}}}$ scaled ($R^{N_{\scriptsize{\mbox{part}}}}_{AA}$) nuclear modification factors as a function of $p_{T}$ of $\phi$ mesons for $0-10\%$ and $20-30\%$ $Cu+Cu$ and $Au+Au$ collisions at $\sqrt{s_{NN}}=200$ GeV. Lower panel. Same as above for $N_{\scriptsize{\mbox{bin}}}$ scaled ($R^{N_{\scriptsize{\mbox{bin}}}}_{AA}$) nuclear modification factor. The error bars represent the statistical and systematic errors added in quadrature. The shaded band in upper panel around 1 at $p_{T}=4.5-5.5$ GeV/$c$ in the right side reflects the uncertainty in $N_{\scriptsize{\mbox{part}}}$ and that on the lower panel for $N_{\scriptsize{\mbox{bin}}}$ calculation for central $Au+Au$ collisions. The respective uncertainties for central $Cu+Cu$ collisions are of similar order.

Upper panels. $N_{\scriptsize{\mbox{part}}}$ scaled ($R^{N_{\scriptsize{\mbox{part}}}}_{AA}$) nuclear modification factors as a function of $p_{T}$ of $\phi$ mesons for $0-10\%$ and $20-30\%$ $Cu+Cu$ and $Au+Au$ collisions at $\sqrt{s_{NN}}=200$ GeV. Lower panel. Same as above for $N_{\scriptsize{\mbox{bin}}}$ scaled ($R^{N_{\scriptsize{\mbox{bin}}}}_{AA}$) nuclear modification factor. The error bars represent the statistical and systematic errors added in quadrature. The shaded band in upper panel around 1 at $p_{T}=4.5-5.5$ GeV/$c$ in the right side reflects the uncertainty in $N_{\scriptsize{\mbox{part}}}$ and that on the lower panel for $N_{\scriptsize{\mbox{bin}}}$ calculation for central $Au+Au$ collisions. The respective uncertainties for central $Cu+Cu$ collisions are of similar order.

Upper panel. The ratio of the yields of $K^{−}$, $\phi$, $\bar{\Lambda}$ and $\Xi+\bar{\Xi}$ normalized to $N_{\scriptsize{\mbox{part}}}$ in nucleus-nucleus collisions to corresponding yields in inelastic $p+p$ collisions as a function of $N_{\scriptsize{\mbox{part}}}$ at 200 GeV. Lower panel. Same as above for $\phi$ mesons in $Cu+Cu$ collisions at 200 and 62.4 GeV. The $p+p$ collision data at 200 GeV are from STAR [5] and at 62.4 GeV from ISR [29]. The error bars shown represent the statistical and systematic errors added in quadrature.

Upper panel. The ratio of the yields of $K^{−}$, $\phi$, $\bar{\Lambda}$ and $\Xi+\bar{\Xi}$ normalized to $N_{\scriptsize{\mbox{part}}}$ in nucleus-nucleus collisions to corresponding yields in inelastic $p+p$ collisions as a function of $N_{\scriptsize{\mbox{part}}}$ at 200 GeV. Lower panel. Same as above for $\phi$ mesons in $Cu+Cu$ collisions at 200 and 62.4 GeV. The $p+p$ collision data at 200 GeV are from STAR [5] and at 62.4 GeV from ISR [29]. The error bars shown represent the statistical and systematic errors added in quadrature.