Tuning Monte Carlo generators: The Perugia tunes

We present 9 new tunes of the $p_{\perp }$-ordered shower and underlying-event model in Pythia 6.4. These “Perugia” tunes update and supersede the older “S0” family. The data sets used to constrain the models include hadronic $Z^0$ decays at LEP, Tevatron min-bias data at 630, 1800, and 1960 GeV, Tevatron Drell-Yan data at 1800 and 1960 GeV, and SPS min-bias data at 200, 546, and 900 GeV. In addition to the central parameter set, called “Perugia 0,” we introduce a set of 8 related “Perugia variations” that attempt to systematically explore soft, hard, parton density, and color structure variations in the theoretical parameters. Based on these variations, a best-guess prediction of the charged track multiplicity in inelastic, nondiffractive minimum-bias events at the LHC is made. Note that these tunes can only be used with Pythia 6, not with Pythia 8.

Figure 1

Comparisons to the CDF and D0 measurements of the $p_{\perp }$ of Drell-Yan pairs [51, 52]. Insets show the high-$p_{\perp }$ tails. Left panels: virtuality-ordered showers. Right panels: $p_{\perp }$-ordered showers. See 43 for other tunes and collider energies.

Figure 2

Comparison of new and old tunes for the $p_{\perp }$ of $t\overline{t}$ pairs at the Tevatron (left panel) and at the LHC at 7 TeV (right panel). See 43 for other tunes and collider energies.

Figure 3

Comparisons to the CDF measurements of the charged track multiplicity at 1800 (left panel) and 1960 GeV (right panel) in minimum-bias $p\overline{p}$ collisions. See 43 for other tunes and collider energies.

Figure 4

Comparisons to the CDF measurement of the charged particle $p_{\perp }$ spectrum in minimum-bias $p\overline{p}$ collisions at 1960 GeV for two sets of models. See 43 for other tunes and collider energies.

Figure 5

Comparisons to the CDF run II measurement of the average track $p_{\perp }$ as a function of track multiplicity in min-bias $p\overline{p}$ collisions. Left panel: The older generation of tunes. Right panel: the Perugia variations compared to two older tunes. See 43 for other tunes and collider energies.

Figure 6

Comparisons to the CDF measurements [72, 73] of the charged particle multiplicity (top row) and $p_{\perp }$ (bottom row) densities in the “TOWARDS” (left panels), “TRANSVERSE” (middle panels), and “AWAY” (right panels) regions of Drell-Yan production at 1960 GeV, as a function of the Drell-Yan $p_{\perp }$.

Figure 7

Double-logarithmic plot of the probability distribution of the number of parton-parton interactions in min-bias collisions at the Tevatron, showing that the Perugia tunes obtain the same multiplicity distribution, Fig. 3, with fewer MPI than Tune A. See 43 for other tunes and collider energies.

Figure 8

Forward-backward correlation strengths at the Tevatron in (left panel) charged particles and (right panel) charged plus neutral transverse-momentum sum at generator level. See 43 for other tunes and collider energies.

Figure 9

Comparisons to UA5 and CDF measurements of the charged track multiplicity in minimum-bias $p\overline{p}$ collisions at 200 GeV (left panels), 630 GeV (middle panels), and 900 GeV (right panels). Top row: Field’s tunes DW and DWT. Bottom row: the main Perugia variations compared to DW. See 43 for other tunes and collider energies.

Figure 10

Charged particle multiplicity and Drell-Yan $p_{\perp }$ spectra at the Tevatron (left and middle) and at the LHC at 14 TeV (right) for the Perugia 0, 3, LO*, and 6 tunes. In particular, the Perugia 3 curve on the lower right-hand plot illustrates the consequences of choosing a different regularization procedure for ISR in the infrared, which shifts the position of the infrared peak of the Drell-Yan $p_{\perp }$ spectrum without affecting the tail of the distribution.

Figure 11

Perugia predictions for the $p_{\perp }$ of Drell-Yan pairs (left panel), the charged track multiplicity in min-bias (center panel), and the average track $p_{\perp }$ vs $N_{\mathrm{ch}}$ in min-bias (right panel) at the LHC at 7 TeV. See 43 for other tunes and collider energies.

Figure 12

Perugia predictions for the charged multiplicity (left panel), $p_{\perp }$ (center panel), and $⟨p_{\perp }⟩(N_{\mathrm{ch}})$ (right panel) distributions in inelastic, nondiffractive $pp$ collisions at 900 GeV, compared with ATLAS data. See 43 for other tunes and collider energies.