ìPresence of Non-dynamical Fluctuations in the Higher Moments of Net-proton Measurements
Prakhar Garg Department of Physics and Astronomy
Stony Brook University
OUTLINE
z BriefIntroduc.onz SourcesofNon-dynamicalFluctua.ons
z MCstudiesforstoppedprotonsandpile-upcontribu.ons
z Conclusion
2PrakharGarg:CPOD2017
BESMotivation:CEP
BaryonChemicalPoten.al(μB)
Tempe
rature
NuclearMaLerHadronGas
QuarkGluonPlasma
LHC
RHIC
CBM
The temperature driven transi.on at zeroµB indicate a rapid crossover from thehadronicphasetotheQGPphase.
TheµBdriventransi.onatfiniteTisafirst
orderphasetransi.on. Afirstorderlineorigina.ngatzeroT must
end somewhere in themidst of the phasediagram where the phase transi.on is acrossover
This end point of a first order phase transition line is a critical end point (CEP)
Limitedtheore.calguidance,needdata!!
3
?
PrakharGarg:CPOD2017
Fluctua(ons:TheorytoExperimentalobservablesareperformedusingcumulants
¤ Cumulantsoffluctua.onsofconservedquan..esarerelatedtothermodynamicsuscep.bili.es(La$ceQCDandHadronResonanceGas(HRG)model)
Cumulantra.osareIndependentofVolume
M.Chengetal,Phys.Rev.D79,074505(2009)F.KarschandK.Redlich,Phys.LeB.B695,136(2011)
v 1stmoment:meanµ=<x>v 2ndcumulant:varianceκ2=σ2=<(x-µ)2>v 3rdcumulant:κ3=μ3=<(x-µ)3>v 3rdstandardizedcumulant:skewness=S=κ3/κ2
3/2=<(x- µ)3>/σ3
v 4thcumulant:κ4=<(x-µ)4>-3κ22
v 4thstandardizedcumulant:kurtosis=κ=κ4 /κ2
2={<(x- µ)4>/σ4}-3
Calculate moments from theevent-by-event net-mul(plicitydistribu(onΔN=N+-N-
κ2κ1
=σµ
2
=χ2χ1
κ4κ2
=κσ 2 =χ4χ2
κ3κ2
= Sσ =χ3χ2
4PrakharGarg:CPOD2017
SeveralMeasurements
5
Net-Proton Fourth-Order Fluctuation
Ø Non-monotonic energy dependence is observed for4th order net-proton, protonfluctuations in most centralAu+Au collisions.
Ø UrQMD results showmonotonic decrease withdecreasing collision energy.
STARPreliminary
aB5 = -2-5
Roli Esha (UCLA)August7,2017 11
RAPID COMMUNICATIONS
MEASUREMENT OF HIGHER CUMULANTS OF NET-CHARGE . . . PHYSICAL REVIEW C 93, 011901(R) (2016)
collisions at different collision energies. These !Nch distri-butions are not corrected for reconstruction efficiency. Thecentrality classes associated with the average number ofparticipants, ⟨Npart⟩, are defined for each 5% centrality bin.These classes are determined using a Monte-Carlo simulationbased on Glauber model calculations with the BBC, RXNP,and EMCal detector responses taken into account [22,23].
The !Nch distributions are characterized by cumulants andrelated quantities, such as µ, σ , S, and κ , which are calculatedfrom the distributions. The statistical uncertainties for thecumulants are calculated using the bootstrap method [24].Corrections are then made for the reconstruction efficiency,which is estimated for each centrality and energy using theHIJING1.37 event generator [25] and then processed througha GEANT simulation with the PHENIX detector setup. Forall collision energies, the average efficiency for detectingthe particles within the acceptance varies between 65%–72%and 76%–85% for central (0%–5%) and peripheral (55%–60%) events, respectively, with a 4%–5% variation as afunction of energy. The efficiency correction applied to thecumulants is based on a binomial probability distribution forthe reconstruction efficiency [26]. The efficiency correctedµ, σ , S, and κ as a function of ⟨Npart⟩ are shown inFigs. 1(c)–1(f).
The µ and σ for net-charge distributions increase withincreasing ⟨Npart⟩, while S and κ decrease with increasing⟨Npart⟩ for all collision energies. At a given ⟨Npart⟩ value, µ,S, and κ of net-charge distributions decrease with increasingcollision energy. However, the width σ of net-charge distribu-tions increases with increasing collision energy indicating theincrease of fluctuations in the system at higher
√s
NN.
The systematic uncertainties are estimated by: (1) varyingthe Zvertex cut to less than ±10 cm; (2) varying the matchingparameters of PC3 hits and EMCal clusters with the projectedtracks to study the effect of background tracks originatingfrom secondary interactions or from ghost tracks; (3) varyingthe centrality bin width to study nondynamical contributionsto the net-charge fluctuations due to the finite width of thecentrality bins [27–29]; and (4) varying the lower cut. The totalsystematic uncertainties estimated for various cumulants forall energies are: 10%–24% for µ, 5%–10% for σ , 25%–30%for S, and 12%–19% for κ . The systematic uncertainties aresimilar for all centralities at a given energy and are treated asuncorrelated as a function of
√s
NN. For clarity of presentation,
the systematic uncertainties are only shown for central (0%–5%) collisions.
Figure 2 shows the ⟨Npart⟩ dependence of µ/σ 2, Sσ ,κσ 2, and Sσ 3/µ[= (Sσ )/(µ/σ 2)] extracted from the net-charge distributions in Au + Au collisions at different
√s
NN.
The results are corrected for the reconstruction efficiencies.Statistical uncertainties are shown along with the data points.The systematic uncertainties are constant fractional errorsfor all centralities at a particular energy; hence they arepresented for the central (0%–5%) collision data point only.The systematic uncertainties on these ratios across differentenergies varies as follows: 20%–30% for µ/σ 2, 15%–34%for Sσ , 12%–22% for κσ 2, and 17%–32% for Sσ 3/µ. It isobserved in Fig. 2 that the ratios of the cumulants are weaklydependent on ⟨Npart⟩ for each collision energy; the values of
2 σ/ µ
0.01
0.1
1 Net-charge Au+Au| < 0.35η|
(a)
σS
0.0
0.2
0.4
0.67.7 GeV19.6 GeV27 GeV
39 GeV62.4 GeV200 GeV
(b)
⟩ part
N⟨100 200 300
2 σ κ
0.5
1.0
1.5
2.0(c)
PHENIX
⟩ part
N⟨100 200 300
µ/3 σS
0.5
1.0
1.5
2.0(d)
FIG. 2. ⟨Npart⟩ dependence of efficiency corrected (a) µ/σ 2, (b)Sσ , (c) κσ 2, and (d) Sσ 3/µ of net-charge distributions for Au + Aucollisions at different collision energies. Statistical errors are shownalong with the data points while systematic uncertainties are shownfor (0%–5%) collisions.
µ/σ 2 and Sσ decrease from lower to higher collision energies,while the κσ 2 and Sσ 3/µ values are constant as a function of√
sNN
within systematic uncertainties.The collision energy dependence of µ/σ 2, Sσ , κσ 2 and
Sσ 3/µ of the net-charge distributions for central (0%–5%)Au + Au collisions are shown in Fig. 3. The statistical andsystematic uncertainties are shown along with the data points.The experimental data are compared with negative-binomial-distribution (NBD) expectations, which are calculated bycomputing the efficiency corrected cumulants for the measuredN+ and N− distributions fit with NBD’s respectively, whichalso describe total charge (N+ + N−) distributions very well
2 σ/µ
0.01
0.1
1 Net-chargeAu+Au (0-5)%
| < 0.35η|
(a)
σS
0.0
0.2
0.4
0.6PHENIX(b)
Data
NBD
(GeV)NNs10 100
2 σ κ
0.5
1.0
1.5
2.0(c)
(GeV)NNs10 100
µ/3 σS
0.5
1.0
1.5
2.0(d)
FIG. 3. The energy dependence of efficiency corrected (a) µ/σ 2,(b) Sσ , (c) κσ 2, and (d) Sσ 3/µ of net-charge distributions forcentral (0%–5%) Au + Au collisions. The error bars are statisticaland caps are systematic uncertainties. The triangle symbol shows thecorresponding efficiency corrected cumulant ratios for net charge,from NBD fits to the individual N+ and N− distributions.
011901-5
RoliEshatalkinCPOD2017(STAR)
Phys.Rev.LeB.112,032302(2014)(STAR)
Phys.Rev.C93,011901(R)(2016)(PHENIX)
values for the Poisson baselines are always unity. Forperipheral collisions the κσ2 values show almost novariation as a function of beam energy and lie above thePoisson baseline and below the NBD baseline. For centralcollisions, within the statistical and systematic errors of thedata, the κσ2 values at all energies are consistent with eachother, except for
ffiffiffiffiffiffiffiffisNN
p ¼ 7.7 GeV. The weighted mean ofκσ2 calculated for central collisions at all energies is2.4" 1.2. For central collisions, both of the baselinecalculations follow the data points except for the one atthe lowest energy. Deviations of the data points withrespect to the baseline calculations have been quantifiedin terms of the significance of deviation, defined as
ðjData − BaselinejÞ=ðffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffierr2stat þ err2sys
qÞ, where errstat and
errsys are the statistical and systematic errors, respectively.These deviations remain within 2 in the case of Sσ and κσ2
with respect to the corresponding Poisson and NBD base-lines. This implies that the products of moments do notshow nonmonotonic behavior as a function of beam energy.The fluctuations of conserved quantities can be used to
extract the thermodynamic information on chemical freeze-out by comparing experimentally measured highermoments with those from first-principle lattice QCDcalculations [23]. Traditionally, by using the integrated
hadron yields, the first moment of the fluctuations, thechemical freeze-out have been extracted from hadronresonance gas (HRG) models [25,41]. However, higher-order correlation functions should allow stricter tests on thethermal equilibrium in heavy-ion collisions. Calculations offreeze-out parameters based on preliminary experimentaldata on moments of net-charge distributions have beenobtained [42,43]. From the latest lattice [44] and HRGanalyses [45] using the STAR net-charge and net-protonresults for central Auþ Au collisions at 7.7 to 200 GeV, theextracted freeze-out temperatures range from 135 to151 MeV and μB values range from 326 to 23 MeV. Theerrors in these calculations increase from 2% to 10% as afunction of decreasing beam energy, which is mostly due tothe statistical uncertainty in the experimental measure-ments. More details can be found in Refs. [44,45]. Note thatthis is the first time that the experimentally measured highermoments have been used to determine the chemical freeze-out conditions in high-energy nuclear collisions. Thefreeze-out temperatures obtained from the higher momentsanalysis are lower with respect to the traditional method[25,46]. This difference could indicate a higher sensitivityto freeze-out in the higher moments, which warrants furtherinvestigation.In summary, the first results of the moments of net-
charge multiplicity distributions for jηj < 0.5 as a functionof centrality for Auþ Au collisions at seven collisionenergies from
ffiffiffiffiffiffiffiffisNN
p ¼ 7.7 to 200 GeV are presented.These data can be used to explore the nature of theQCD phase transition and to locate the QCD critical point.We observe that the σ2=M values increase monotonicallywith increasing beam energy. Weak centrality dependenceis observed for both Sσ and κσ2 at all energies. The Sσvalues increase with decreasing beam energy, whereas κσ2
values are uniform except at the lowest beam energy. Mostof the data points show deviations from the Poissonbaselines. The NBD baselines are closer to the data thanthe Poisson baselines, but do not quantitatively reproducethe data, implying the importance of intraevent correlationsof the multiplicities of positive and negative particles in thedata. Within the present uncertainties, no nonmonotonicbehavior has been observed in the products of moments as afunction of collision energy. The measured moments of net-charge multiplicity distributions provide unique informa-tion about the freeze-out parameters by directly comparingwith theoretical model calculations. Future measurementswith high statistics data will be needed for a precisedetermination of the freeze-out conditions and to makedefinitive conclusions regarding the critical point.
We thank M. Asakawa, R. Gavai, S. Gupta, F. Karsch,V. Koch, S. Mukherjee, K. Rajagopal, K. Redlich, andM. A. Stephanov for discussions related to this work. Wethank the RHIC Operations Group and RCF at BNL, theNERSC Center at LBNL, the KISTI Center in Korea, andthe Open Science Grid consortium for providing resources
M2 σ
10
210Net-charge
Au+Au| < 0.5η|
(a)
σS
0
0.2
0.4
0.6 (b)
(GeV)NNs√5 6 10 20 30 100 200
2 σκ
-15
-10
-5
0
5(c)
0-5%70-80%0-5% NBD70-80% NBD0-5% Poisson70-80% Poisson
FIG. 4 (color online). Beam-energy dependence of (a) σ2=M,(b) Sσ, and (c) κσ2, after all corrections, for most central(0%–5%) and peripheral (70–80%) bins. The error bars arestatistical and the caps represent systematic errors. Results fromthe Poisson and NBD baselines are superimposed. The values ofκσ2 for the Poisson baseline are always unity.
PRL 113, 092301 (2014) P HY S I CA L R EV I EW LE T T ER Sweek ending
29 AUGUST 2014
092301-6
Phys.Rev.LeB.113,092301(2014)(STAR)
Helen Caines - RHIC&AGS Users Meeting 17
Presence of Critical Point?
11
M. Stephanov. PRL 107:052301(2011)
Correlation lengths diverge → Net-p κσ2 diverge
Hints of Critical fluctuations
HADES data + upcoming FXT testing if mapping correct
New HADES data causing some tension
Comparison with STAR BES-I
red/black = unfolding (preferred method) + vol. flucs. corr.
green = evt-by-evt eff correction of factorial moments + vol. flucs. corr.
HADES preliminary
HADES preliminary
STAR analysis: Xiaofeng Luo et al., PoS (CPOD2014) 019 arXiv:1503.02558v2
QM2017 February 6 - 11, 2017 Chicago IL 20
3
and δ = 5, which are within few percent of their exactvalues in three dimensions. The result of Eq. (9) can thenbe simplified to
κ4(t,H) = −1281− 783θ2 + 105θ4 − 5θ6 + 2θ8
R14/3(3− θ2)3(3 + 2θ2)5. (10)
We represent κ4(t,H) graphically as a density plot inFig. 1. We see that the 4-th cumulant (and kurtosis)is negative in the sector bounded by two curved raysH/tβδ = ±const (corresponding to θ ≈ ±0.32).
(a)
!0.4 !0.2 0.0 0.2 0.4 0.6!20
020406080
100120
t
Κ4
(b)
FIG. 1: (color online) (a) – the density plot of the functionκ4(t,H) given by Eq. (10) obtained using Eq. (9) for the linearparametric model Eqs. (6), (7), (8) and β = 1/3, δ = 5. Theκ4 < 0 region is red, the κ4 > 0 – is blue. (b) – the dependenceof κ4 on t along the vertical dashed green line on the densityplot above. This line is the simplest example of a possiblemapping of the freezeout curve (see Fig. 2). The units of t,H and κ4 are arbitrary.
Also in Fig. 1 we show the dependence of κ4 along aline which could be thought of as representing a possiblemapping of the freezeout trajectory (Fig. 2) onto the tHplane. Although the absolute value of the peak in κ4
depends on the proximity of the freezeout curve to thecritical point, the ratio of the maximum to minimumalong such an H = const curve is a universal number,approximately equal to −28 from Eq. (10).
µB, GeV
, GeV
0
0.1
T
t
1
H
criticalpoint
freezeoutcurve
nuclearmatter
QGP
hadron gas
FIG. 2: A sketch of the phase diagram of QCD with the freeze-out curve and a possible mapping of the Ising coordinates tand H .
The negative minimum is small relative to the positivepeak, but given the large size of the latter, Ref.[7, 15],the negative contribution to kurtosis may be significant.In addition, the mapping of the freezeout curve certainlyneed not be H = const, and the relative size of the posi-tive and negative peaks depends sensitively on that.The trend described above appears to show in the re-
cent lattice data, Ref.[10], obtained using Pade resum-mation of the truncated Taylor expansion in µB. As thechemical potential is increased along the freezeout curve,the 4-th moment of the baryon number fluctuations be-gins to decrease, possibly turning negative, as the criticalpoint is approached (see Fig.2 in Ref.[10]).Another observation, which we shall return to at the
end of the next section, is that −κ4 grows as we approachthe crossover line, corresponding to H = 0, t > 0 on thediagram in Fig. 1(a). On the QCD phase diagram thefreezeout point will move in this direction if one reducesthe size of the colliding nuclei or selects more peripheralcollisions (the freezeout occurs earlier, i.e., at higher T ,in a smaller system).
EXPERIMENTAL OBSERVABLES
In this section we wish to connect the results for thefluctuations of the order parameter field σ to the fluctua-tions of the observable quantities. As an example we con-sider the fluctuations of the multiplicity of given chargedparticles, such as pions or protons.For completeness we shall briefly rederive the results of
Ref.[7] using a simple model of fluctuations. The modelcaptures the most singular term in the contribution of thecritical point to the fluctuation observables. Consider agiven species of particle interacting with fluctuating crit-ical mode field σ. The infinitesimal change of the field δσleads to a change of the effective mass of the particle bythe amount δm = gδσ. This could be considered a def-inition of the coupling g. For example, the coupling ofprotons in the sigma model is gσp̄p. The fluctuations δfp
NB: Different y and pT ranges R. Holzmann QM2017
N.B. non-monotonic behavior not observed by PHENIX - due to different acceptances? (PRC 93, 011901 (2016))
Nor in pp or light nuclei by NA61(PRC 93, 011901 (2016))
HelenCainestalkinRHIC&AGSUsersMeeUng-2017
10 210
10
20
30
40
0-5% STAR 70-80% STAR 0-5% Poisson
70-80% Poisson 0-5% UrQMD
0-5% NBD70-80% NBD
<1.6(GeV/c),|y|<0.5T
0.2<pNet-kaonAu+Au Collisions at RHIC(a)
10 210
0
1
2(b)
STAR Preliminary
7 8 910 20 30 40 100 200
2−
0
2(c)
(GeV)NNs2
σκ
/M2σ
/Ske
llam
σS
Figure 4. Energy dependence of cumulant ratios (�2/M, S�/Skellam, �
2) of net-kaonmultiplicity distributions for top 0-5% central and 70-80% peripheral collisions at Au+Aucollisions at
psNN = 7.7, 11.5, 14.5, 19.6, 27, 39, 62.4, and 200GeV. The Poisson and NBD
expectations are denoted as dotted and solid lines, UrQMD calculations are shown as bluebands. The error bars are statistical errors and the caps represent systematic errors.
phase-space to enhance the search for the QCD critical point in the high net-baryon densityregion 420 < µB < 250 MeV (7.7 <
psNN < 19.6 GeV) .
Acknowledgments
The work was supported in part by the MoST of China 973-Project No.2015CB856901, NSFCunder grant No. 11575069, 11221504.
References
[1] M. A. Stephanov, Phys. Rev. Lett. 102, 032301 (2009).[2] M. Asakawa et al., Phys. Rev. Lett. 103 262301 (2009).[3] M. A. Stephanov, Phys. Rev. Lett. 107 052301 (2011).[4] R. V. Gavai and S. Gupta, Phys. Rev. D 78, 114503 (2008).[5] M. Cheng et al., Phys. Rev. D 79, 074505 (2009); A. Bazavov et al., arXiv:1203.0784.[6] S. Borsanyi et al., JHEP01, 138 (2012).[7] F. Karsch and K. Redlich, Phys. Lett. B 695, 136 (2011).[8] P. Garg et al., Phys. Lett. B 726, 691 (2013).[9] Marlene Nahrgang et al., arXiv: 1402.1238.
[10] J. Fu, Phys. Lett. B 722, 144 (2013).[11] S. Gupta, X. Luo, B. Mohanty, H. G. Ritter and N. Xu, Science 332, 1525 (2011).[12] Aggarwal M M et al. (STAR Collaboration) 2010 Phys. Rev. Lett. 105 022302[13] X. Luo (STAR Collaboration), Nucl. Phys. A 904 911c (2013).[14] X. Luo, arXiv:1512.09215[15] X. Luo, J. Phys. G: Nucl. Part. Phys. 39 025008 (2012).[16] X. Luo, Phys. Rev. C 91, 034907 (2015).[17] L. Adamczyk et al., (STAR Collaboration), Phys. Rev. Lett.113, 092301 (2014).
4
JiXutalkinsQM-2016
PrakharGarg:CPOD2017
SourcesofNon-dynamicalFluctuations
6
¤ Effectofphase-spaceacceptance:P.Garg,D.K.Mishraetal.(Phys.LeB.B726(2013)691-696)FrithjofKarschelal.Phys.Rev.C93(2016),034907¤ Effectofe-by-eeficiencycorrec.ons:P.Garg,D.K.Mishraetal.(J.Phys.G40(2013)055103)A.BzdakandV.Koch,Phys.Rev.C86(2012),044904A.BzdakandV.Koch,Phys.Rev.C91(2015)027901¤ Effectofnon-extensivesta.s.cs:D.K.Mishra,P.Gargetal.(J.Phys.G42(2015),105105)¤ Ensembledependence:P.Garg,D.K.Mishraetal.(Eur.Phys.J.A52(2016),27)¤ Effectofcorrela.onsoncumulants:P.Garg,D.K.Mishraetal.(Phys.Rev.C93(2016),024918)¤ EffectofResonanceDecay:D.K.Mishra,P.Gargetal.(Phys.Rev.C94(2016),014905)MarleneNahrgangetal.Eur.Phys.J.C(2015)75:573¤ EffectofPar.cipantFluctua.ons:P.Braun-Munzingeretal.(NPA960(2017)114)¤ Globalbaryonnumberconserva.on:P.Braun-Munzingeretal.(PLB747(2015)292)¤ ........ANDSOON
z EffectofstoppedprotonFluctua.ons->D.K.Mishra&P.GargarXiv:1706.04012andD.Thakur,S.Jakhar,P.Gargetal.(Phys.Rev.C95(2017),044903)z EffectofeventPile-up->P.Garg&D.K.MishraarXiv:1705.01256
Inthistalk(fornet-protonfluctua.onmeasurements)
PrakharGarg:CPOD2017
BaryonStopping
Ionsabouttocollide Ioncollision Quarks,gluonsfreed Plasmacreated
v Largeamountofenergyisdepositedin a small region of space in a shortdura.onof.me.
v Occurrenceofhighenergydensityregions:->BaryonrichQuarkGluonPlasma(@low√sNN)->BaryonfreeQuarkGluonPlasma(@high√sNN)
AtRHICBES-IEnergies:Inclusiveprotonscontainproducedprotonsandstoppedprotons
Numberofstoppedprotonsfluctuatee-by-eleadingtoaddi.onalfluctua.ons
Needtodisentanglethecontribu.onofstoppedprotonsandproducedprotonsExperimentallydifficulttotagstoppedprotonandproducedproton
PartI:
7PrakharGarg:CPOD2017
Getprotonandanti-protondistributionsbytuningwithSTARcumulantsdataforAu+Aucollisions
8
p
5
10
15
20 1C
Expt. dataBinomial
Au + Au (0-5)%
p
0
1
2
3
4 1C
(GeV) NNs6 10 20 100200
pp
-
05
101520 1C
2C
2C
(GeV) NNs6 10 20 100 200
2C
3C
3C
(GeV) NNs6 10 20 100 200
3C
4C
4C
(GeV) NNs6 10 20 100 200
4C1
/C 2C
0.6
0.8
1.0
1.2 p
Expt. dataBinomial
Au + Au (0-5)%
2/C 3C
0.6
0.8
1.0
1.2 p
2/C 4C 0.5
1.0p
1/C 2C
0.8
1.0
1.2 p
2/C 3C
0.8
1.0
1.2 p
2/C 4C
0.8
1.0
1.2 p
(GeV) NNs6 10 20 100200
1/C 2C
2
4
6 pp -
(GeV) NNs6 10 20 100200
2/C 3C 0.5
1.0 pp -
(GeV) NNs6 10 20 100200
2/C 4C
0.5
1.0pp -
p
TheBinomialexpecta(onsaretunedforprotonandan(-protoncumulantsdataateach√sNN
Thecumulantsindataarealreadycorrectedforefficiencyandfinitebinwidtheffects
arXiv:1706.04012STARnet-protonDATA
MCPoints
hTps://drupal.star.bnl.gov/STAR/files/starpublica(ons/205/data.html
PrakharGarg:CPOD2017
ComparisonofStoppedProtonswith
STARresults
9
v Alargecontribu(onofstoppedprotonsatBESenergies.v AZer subtrac(ng the stopped protons from themean of STAR protons distribu(on, remaining
producedprotonsareconsistentwithmeanofan(-protondistribu(onmeasuredbySTAR.
Phys.Rev.C95(2017),044903
STARDATA:hTps://drupal.star.bnl.gov/STAR/files/starpublica(ons/205/data.html
Phys.LeB.B690,358(2010)
PrakharGarg:CPOD2017
Method:Extracttheweightfactors
10
C1ofpstop_____________=Frac(onofStoppedProtonsC1ofpinclu
Phys.Rev.C95(2017),044903
Now,“assuming”thatthisfrac(onisalsodistributedaccordingtothe
Binomialdistribu(on
Wegetacorrecteddistribu(onforproducedprotons
PrakharGarg:CPOD2017
Cumulantsandtheirratiosforindividuals
11
(GeV) NNs6 10 2030 100 200
1C
0
5
10
15
20 inclpprodpstopp
p
(GeV) NNs6 10 2030 100 200
2C0
5
10
15
20
(GeV) NNs6 10 2030 100 200
3C
0
5
10
15
20
(GeV) NNs6 10 2030 100 200
4C
0
5
10
15
20
21C
0.00.51.01.52.0 inclp
prodpstopp
p
32C
0.00.51.01.52.0
(GeV) NNs6 10 2030 100 200
42C
-0.50.00.51.01.52.0
arXiv:1706.04012
Atlowercollisionenergies,thepinclfluctua(onsaredominatedbypstopandathigherenergiestheyaredominatedbyproducedprotonfluctua(ons.
PrakharGarg:CPOD2017
Cumulantsandtheirratiosfornet-protons
12
(GeV) NNs6 10 2030 100 200
1C
0
5
10
15
20 p- inclpp- prodp
(GeV) NNs6 10 2030 100 200
2C
0
5
10
15
20
(GeV) NNs6 10 2030 100 200
3C
0
5
10
15
20
(GeV) NNs6 10 2030 100 200
4C
0
5
10
15
20
32C
-1.0
-0.5
0.0
0.5
1.0 p- inclp
p- prodp
(GeV) NNs67810 20 3040 100 200
42C
0.4
0.6
0.8
1.0
1.2
arXiv:1706.04012
Ø Correc.onsforstoppedprotonfluctua.onmayenhancethesignalascanbeseenforC32PrakharGarg:CPOD2017
EventPile-upEffects
13
PartII:
Ref:HarnarineIan,”AStudyofPile-upin200GeVAu+AuCollisionsatRHIC”,DoctoraldissertaOon,UniversityofIllinoisatChicago,2005.andreferencestherein.
Inhighluminosityheavy-ioncollisions,theremaybefollowingsourcesofthebackgroundeventsduringacollisionØ In-(mepile-up:Ifmorethanonecollisionsareoccurringinthesamebunch-crossingina
collisionofinterest;
Ø Out-of-(mepile-up:Ifaddi(onalcollisionsareoccurringinabunch-crossingbeforeandaZerthecollision.
Ø Cavernbackground:Mainlylowenergyneutronsandphotons
Ø Beamhaloevents:Thedispersioninthebeam
Ø Beamgasevents:Collisionsbetweenthebunchandtheresidualgasinsidethebeam-pipe.
PrakharGarg:CPOD2017
SimpleMCforeventpile-upstudy
14
Cou
nts
110
210
310410
510
610
710 p19.6 GeV
w/o pile-upw (0.05%) pile-up
pN0 10 20 30 40
Cou
nts
110
210310410510610 p
200 GeV
p19.6 GeV
pN0 5 10 15 20
p200 GeV
pp - 19.6 GeV
pp-N-10 0 10 20 30 40
pp - 200 GeV
arXiv:1705.01256
GeneratethePoissonorNBDDistribu.onforprotonsandan.-
protonsusingtheMeanvaluesfromSTARdata
Addextraprotonsandan.-protonscomingfrompile-upevent
Randomlyselecttheprotonmul.plicityfromMBprotonsampleandaddittotheoriginaldistribu.on
Studytheeffectoncumulantswithdifferentfrac.onofmixedevents
Ø Theprotonandan(-protonmul(plici(esfromminimumbiasevents as pile-up are added with the individual p and p ̄distribu(onsassumingeachofthedistribu(onasNBD.
PrakharGarg:CPOD2017
Central+MinimumBiaseventaspile-up
15
Individual cumulants for Central + min.bias assuming PoissonSimilar to Fig.2 of the draft.
Individual cumulants for Central + min.bias assuming NBDSimilar to Fig.5 of the draft
arXiv:1705.01256PoissonAssump(on NBDAssump(on
Ø Moreimportantforhighercumulants!!PrakharGarg:CPOD2017
Central+MinimumBiaseventaspile-up
16
(GeV) NNs6 10 20 100 200
21C
2
4
6% of pileup
0.00.010.050.100.501.002.00
(a)
NBD
(GeV) NNs6 10 20 100200
32C
0
1
2
3 (b)
(GeV) NNs6 10 20 100 200
42C
1
10
210(c)
(GeV) NNs6 10 20 100200
31C
0
1
2
3
4 (d)
arXiv:1705.01256PoissonAssump(on NBDAssump(on
PrakharGarg:CPOD2017
MostExtremeSituation(Central+Centraleventasapile-up)
17
(GeV) NNs6 10 20 100200
21C
2
4
6% of pileup
0.00.010.050.100.501.002.00
(a)
Poisson
(GeV) NNs6 10 20 100200
32C
02
4
6
8 (b)
(GeV) NNs6 10 20 100200
42C
1
10
210(c)
(GeV) NNs6 10 20 100200
31C
02468
10 (d)
(GeV) NNs6 10 20 100 200
21C
2
4
6% of pileup
0.00.010.050.100.501.002.00
(a)
NBD
(GeV) NNs6 10 20 100200
32C
0
2
4
6
8 (b)
(GeV) NNs6 10 20 100 200
42C
1
10
210(c)
(GeV) NNs6 10 20 100200
31C
02468
10 (d)
arXiv:1705.01256PoissonAssump(on NBDAssump(on
PrakharGarg:CPOD2017
Conclusion
18
Ø StoppedProtonFluctua(onsmayhavesignificanteffecton the net-proton measurements which needs to beaddressedcarefully.
Ø Event-pile up can also influence the Fluctua(onMeasurements and should be studied in eachexperimentalset-upforcumulantsobservable.
PrakharGarg:CPOD2017