Measurement of Nonrandom Event-by-Event Fluctuations of Average Transverse Momentum in $\sqrt{s_{NN}}=200\mathrm{G}\mathrm{e}\mathrm{V}$ $\mathrm{A}\mathrm{u}+\mathrm{A}\mathrm{u}$ and $p+p$ Collisions

Event-by-event fluctuations of the average transverse momentum of produced particles near midrapidity have been measured by the PHENIX Collaboration in $\sqrt{s_{NN}}=200\mathrm{G}\mathrm{e}\mathrm{V}$ $\mathrm{A}\mathrm{u}+\mathrm{A}\mathrm{u}$, and $p+p$ collisions at the Relativistic Heavy Ion Collider. The fluctuations are observed to be in excess of the expectation for statistically independent particle emission for all centralities. The excess fluctuations exhibit a dependence on both the centrality of the collision and on the $p_T$ range over which the average is calculated. Both the centrality and $p_T$ dependence can be well reproduced by a simulation of random particle production with the addition of contributions from hard-scattering processes.

Figure 1

Comparisons between the data and mixed event $M_{p_T}$ distributions for the representative 0%–5% and 30%–35% centrality classes. Plots (a) and (c) show direct comparisons of the data (points) and normalized mixed event (solid line) $M_{p_T}$ distributions. Plots (b) and (d) show the residuals between the data and mixed events in units of standard deviations of the data points from the mixed event points.

Figure 2

$F_{p_T}$ (in percent, $0.2\mathrm{G}\mathrm{e}\mathrm{V}/c<p_T<2.0\mathrm{G}\mathrm{e}\mathrm{V}/c$) as a function of centrality, which is expressed in terms of the number of participants in the collision, $N_{\mathrm{p}\mathrm{a}\mathrm{r}\mathrm{t}}$. The solid squares represent the $\mathrm{A}\mathrm{u}+\mathrm{A}\mathrm{u}$ data. The solid triangle represents the minimum bias $p+p$ data point. The open triangle is the result from an analysis of PYTHIA minimum bias $p+p$ events within the PHENIX acceptance. The error bars include statistical and systematic errors and are dominated by the latter. The curves are the results of a Monte Carlo simulation with hard processes modeled using PYTHIA with a constant (dotted curve) and $R_{AA}$-scaled (dashed curve) hard-scattering probability factor, and include the estimated contribution due to elliptic flow.

Figure 3

$F_{p_T}$ (in percent) of nonrandom fluctuations as a function of the $p_T$ range over which $M_{p_T}$ is calculated, $0.2\mathrm{G}\mathrm{e}\mathrm{V}/c<p_T<p_T^{\max }$, for the 20%–25% centrality class ($N_{\mathrm{p}\mathrm{a}\mathrm{r}\mathrm{t}}=181.6$). The curve is the result of a Monte Carlo simulation with hard-scattering processes modeled using PYTHIA with $S_{\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{b}}(N_{\mathrm{p}\mathrm{a}\mathrm{r}\mathrm{t}})=0.075$ and $R_{AA}=0.41$ [23]. The error bars include statistical and systematic errors and are dominated by the latter. The contribution of elliptic flow is estimated to be negligible at this centrality.