CERN Accelerating science

Tracklet candidate $\Delta\eta$ (left) and $\Delta\phi$ (right) distributions from data (histogram) and reweighted MC (shaded region) for \PbPb\ collisions at $\sqrts=2.76$~TeV. The top panels correspond to $|\eta|<1$ and the bottom panels correspond to $1<|\eta|<2$. Data and MC distributions are normalized to the same area.

{\bf Left:} Top: uncorrected track/tracklet $dN_{\mathrm{raw}}/d\eta$ distribution from tracklet Method~1 (points), tracklet Method~2 (squares) and pixel tracking (blue triangles) for 0-10$\%$ centrality events. Middle: corrected tracklet and track \dndeta\ distributions. Bottom: ratio of \dndeta\ from the tracklet Method~2 (squares) and pixel tracking (triangles) to tracklet Method~1. {\bf Right:} \dndeta\ distributions from tracklet Method~1 for eight 10\% centrality intervals. The statistical errors are shown as bars and the systematic errors are shown as shaded bands.

noimgSummary of the various sources of systematic uncertainties and their estimated impact on the \dndeta\ measurement in central (0-10\%) and peripheral (70-80\%) \PbPb\ collisions. Only the uncertainty due to the choice of the event generator is $\eta$-dependent.

{\bf Top:} \sqrts\ dependence of the charged particle \dndeta\ per colliding nucleon pair $\dndetamid/(\Npart/2)$ from a variety of measurements in \pp\ and \mbox{$\bar{\mathrm{p}}$+p} (inelastic and non-single diffractive results from \cite{Alver:2010ck} and references therein, as well as ~\cite{:2010ir}-\cite{:2009dt}) and central \nucnuc\ collisions, including the ATLAS 0-6\% centrality measurement reported here for $|\eta|<0.5$ and the previous 0-5\% centrality ALICE~\cite{Collaboration:2010cz} and CMS~\cite{CMS:2011mult} measurements (points shifted horizontally for clarity). The curves show different expectations for the \sqrts\ dependence in \nucnuc\ collisions: results of a Landau hydrodynamics calculation~\cite{Carruthers:1973ws} (dotted line) , an $s^{0.15}$ extrapolation of RHIC and SPS data proposed by ALICE \cite{Collaboration:2010cz} (dashed line), a logarithmic extrapolation of RHIC and SPS data from \cite{Busza:2007ke} (solid line). {\bf Bottom:} $\dndetamid/(\Npart/2)$ vs \Npart\ for 2\% centrality intervals over 0-20\% and 5\% centrality intervals over 20-80\%. Error bars represent combined statistical and systematic uncertainties on the \dndetamid\ measurements, whereas the shaded band indicates the total systematic uncertainty including \Npart\ uncertainties. The RHIC measurements (see text) have been multiplied by 2.15 to allow comparison with the $\sqrts = 2.76$~\TeV\ results. The inset shows the $\langle N_{\mathrm{part}} \rangle<60$ region in more detail.

{\bf Top:} \dndeta\ distributions from tracklet Method~1, scaled by \dndetamid, as a function of the pseudorapidity for the 70-80\% centrality interval. The statistical errors are shown as error bars. {\bf Bottom:} Ratio of $\dndeta/(\Npart/2)$ measured in different centrality intervals: 0-10\% (squares), 20-30\% (triangles), 40-50\% (inverted triangles) and 60-70\% (crosses) to that measured in peripheral collisions (70-80\%). Statistical uncertainties are shown as bars while $\eta$-dependent systematic uncertainties are shown as shaded bands.