Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Long-range azimuthal correlation
Document related concepts
ALICE experiment wikipedia , lookup
Elementary particle wikipedia , lookup
Grand Unified Theory wikipedia , lookup
Technicolor (physics) wikipedia , lookup
ATLAS experiment wikipedia , lookup
Electric charge wikipedia , lookup
Standard Model wikipedia , lookup
Transcript
Multi-particle azimuthal correlation with subevent method Mingliang Zhou for the ATLAS collaboration Quark Matter 2017, Feb. 5-11, Chicago arXiv: 1606.08170 ATLAS-CONF-2017-002 Long-range pseudo-rapidity correlation Nature of sources seeding the long-range collective behavior? ππ΅ ππΉ participant partons string/tubes β’ ππ ππ shape reflects asymmetry in the number of forward-backward sources; Event 1 π π π π Event 2 π π π π β 1 + π1 π π PRC 87, 024906 (2013) PRC 93, 044905 (2016) β’ Dominated by linear fluctuation! Not enough FB sources in models. β’ Independent of collision system! arXiv: 1606.08170 1 Long-range azimuthal correlation Nature of sources seeding the long-range collective behavior? β’ Azimuthal correlation: well understood in A+A. 2 Long-range azimuthal correlation PRL 116 (2016) 172301 β’ Near-side SRC: jet, HBTβ¦ β’ Away-side ridge: mainly from dijet. β’ Near-side ridge β’ Ridge structure also observed in small systems! Competing theories: β’ Initial stage interaction (CGC, β¦)? PRD 87 (2013) 094034 β’ Initial stage fluctuation + final stage interaction? PRC 88 (2013) 014903 β’ Fluctuation in proton can be constrained in pp and p+Pb collisions. β’ However, ridge hard to extract in data: larger non-flow in small system. From the experimental perspective, central question about the ridge is whether it involves all particles (collectivity?). 2 Standard cumulant β’ Cumulant method: natural to probe collectivity: πΆ2 4 β‘ 4 β 2 2 π£2 2 > π£2 4 π£2 4 β π£2 6 ππβ 2 PRL 115, 012301 (2015) However, does cumulant work in pp? is smaller, non-flow contribution is larger. 3 Limitations of standard cumulant arXiv: 1701.03830 Energy dependence? Large residual non-flow in PYTHIA. 6% flow Standard cumulant πΆ2 4 cannot effectively remove non-flow contribution in pp! ATLAS-CONF-2016-106 π£2 2 β π£2 4 ? Sign depends on event class definition? 4 Subevent cumulant 5 β’ In standard cumulant, non-flow sources contribute to four-particle correlation 4 ; 4 β‘ π 2π ππ +ππ βππ βππ π π πΆ2 4 β‘ 4 β 2 2 2 β‘ 4 nonβflow + π£4 β 2 π£2 2 Subevent cumulant 5 β’ In standard cumulant, non-flow sources contribute to four-particle correlation 4 ; β’ In the subevent method, particles are correlated across all subevents (long-range). β’ 3 subevent cumulant can further suppress away-side jet contribution; β’ New method validated in PYTHIA, but is there a data-driven way to check the residual non-flow in data? 4 β‘ π 2π ππ +ππ βππ βππ π π π π π 2π|π,π πΆ2 4 β‘ 4 2π|π,π β2 2 π|π 2 π|π arXiv: 1701.03830 Test of residual non-flow πΆ2 4 β‘ nonflow + flow Non-flow changes greatly EbyE 6 ππ£π‘ Flow changes little EbyE non-flow fluc. β multiplicity fluc. β how β¦ πΊππ is defined: π΅ πππ ππ πππ β’ ππβ defined with different ππ : very different non-flow fluctuation. πππ ππβ defined with 0.3 < ππ < 3.0 GeV πππ ππβ defined with ππ > 0.4 GeV β’ Non-flow fluctuation might mimic the flow signal (negative πΆ2 4 )! Standard cumulant v.s. Subevent cumulant 7 4% flow non-flow fluctuation π ππ Different for ππβ πππ Different ππβ ? ? ? Same π ππ for ππβ Different non-flow fluctuation Different πΆ2 4 Less non-flow Less non-flow fluctuation πππ Less dependence on ππβ 3 subevent cumulant is a more reliable method in pp! New method tested in 5.02 TeV p+Pb 4% flow ? β’ Consistent at large ππβ : non-flow is smaller; β’ Split observed at low ππβ : suppression of non-flow; 8 13 TeV pp: comparison of three methods π£2 4 β‘ βπΆ2 4 4% flow β’ Standard cumulant has positive πΆ2 4 : large residual non-flow; β’ 2 subevent cumulant already suppresses non-flow; β’ 3 subevent cumulant measures 4% flow down to 70 tracks. 9 13 TeV pp: higher ππ region 4% flow β’ Higher ππ range: higher fraction of non-flow; β’ ONLY 3 subevent method gives larger flow signal; 10 Comparison among three collision systems β’ β’ β’ β’ pp results from standard cumulant unable to compare with π+Pb; With 3 subevent, negative πΆ2 4 observed in 5.02 TeV pp; Weak energy dependence in pp: more data for 5.02 TeV pp in 2017? π+Pb has larger flow than pp in the comparable ππβ region; 11 Third harmonic πͺπ π β’ πΆ3 4 still has residual non-flow in standard method; β’ πΆ3 4 from 3 subevent is consistent with 0: β’ π£3 βͺ π£2 β’ Fluctuation kills πΆ3 4 ? 12 Comparison with 2-particle correlation method 13 π£2 2 from template fit π£2 4 from 3 subevent π£2 2 from peripheral sub. β’ π£2 4 < π£2 2 (template fit): flow fluctuation; β’ π£2 4 β π£2 2 (peripheral subtraction): underestimation of π£2 2 ; β’ Decreasing trend of πΆ2 4 for central p+Pb: β’ Many sources in central p+Pb? Number of sources in initial stage 14 β’ π£2 2 β π£2 4 : EbyE flow fluctuations associated with fluctuating initial conditions. PRL 112, 082301 (2014) β’ Fluctuation can be quantified to the number of sources ππ in the initial stage: π£2 4 4 = π£2 2 3 + ππ 1 4 π π π π β 1 + π1 π, π1 β 1 π΅π ? β’ ππ for π+Pb goes up to 20 at high multiplicity; β’ ππ for ππ approximately consistent with π+Pb at comparable ππβ value. ππ Summary 15 Weak energy dependence. Less non-flow in 3 subevent. Hope the new results can better constrain models in small systems. π£2 2 from template fit π£2 4 from 3 subevent π£2 4 < π£2 2 β ππ Less non-flow fluctuation in 3 subevent. ## Back Up Data sets β’ Studied in this analysis: β’ Energy dependence of πΆ2 4 in pp; β’ Comparison between pp and p+Pb; β’ 3 collision systems included: β’ 13 TeV pp (2015+2016) β’ 5.02 TeV pp (2015) β’ 5.02 TeV p+Pb (2013+2016) β’ Two πT ranges: 0.3 < πT < 3.0 GeV and 0.5 < πT < 5.0 GeV; β’ High-multiplicity track triggers developed to enhance statistics at large ππβ : HMT 5.02 TeV pp Formulas without particle weight non-flow fluctuation Multiplicity fluctuation ? ? ? 4 ππ£π‘ β’ In CMS 0.3 < πT < 3.0 GeV β’ 4 is calculated in 0.3 < ππ < 3.0 GeV; πππ β’ β¦ ππ£π‘ is binned with ππβ of ππ > 0.4 GeV; β’ In ATLAS β’ 4 is calculated in 0.3 < ππ < 3.0 GeV; πππ β’ β¦ ππ£π‘ is binned with ππβ of 0.3 < ππ < 3.0 GeV; β’ X-axis of πΆ2 4 is projected to ππβ with ππ > 0.4 GeV; Residual non-flow check in 2 subevent Residual non-flow check in higher ππ Residual non-flow check in higher ππ π΅π in π. π < ππ < π. π GeV Summary 16 Nature of sources seeding the long-range collective behavior? Longitudinal correlation ππ πΉ ππ π΅ PYTHIA Azimuthal correlation CGC? Hydro? PRC 94, 044918 (2016) Models underestimated data. β’ Fluctuation of saturation scale. Physics signal and background β’ 2-particle pseudo-rapidity correlation: πΆ π1 , π2 π π1 π π2 β‘ π π1 π π2 β’ Short-range correlation (SRC) reflects correlation in the same source: jet fragmentation, resonance decayβ¦ β’ Since SRC has a strong charge dependence, a data-driven way was developed to remove the SRC. β’ An alternative way is through multiparticle correlation (cumulant) in pseudo-rapidity. PRC.93.024903 β’ Long-range correlation (LRC) From the experimental perspective, the key is to extract the physics signal from the background. Charge dependence of πͺ πΌπ , πΌπ β’ Particles from the same source (SRC) have strong charge π+ dependence. Opposit e charges Same charge s β’ Ratio of opposite to same charges π π1 , π2 β’ Very strong Gaussian-like SRC; β’ Very weak LRC: charge-independent; β’ Amplitude of π π1 , π2 along π+ : π π+ , reflects the strength of SRC in the longitudinal direction; β’ Assumption: strength of SRC along π+ is same for same charge and opposite charge. π¨π©π©π¨ π¬ππ¦π β‘ π1 + π2 Estimation of short-range correlation β’ To estimate SRC, LRC pedestal is estimated first. Same charge s SRC β’ πΆ π1 , π2 from same charge used to estimate LRC pedestal because of small SRC; β’ LRC pedestal is fitted with quadratic function; β’ The additional structure upon LRC pedestal determines the shape of SRC; β’ The full πΏππ πΆ π1 , π2 is then populated using π π scaling. Why it is important to remove SRC? Before SRC removal After SRC removal Oppo pairs All Same pairs β’ Higher order coefficients observed; β’ Coefficients have charge dependent; β’ Results hard to interpret: due to SRC! β’ Simpler picture after SRC removal! β’ LRC dominated by linear fluctuation; β’ LRC is charge independent. Why it is important to remove SRC? Short-range correlation βππ πΆ β‘ Long-range correlation πΏππ πΆ π1 , π2 ππ1 ππ2 4π 2 β’ SRC increases towards peripheral; β’ SRC is stronger in small systems; β’ Dominated by linear fluctuation! β’ Independent of collision system! β’ Follow-up model studies β’ Viscous hydro: PLB 2015.11.063 β’ Length of sources fluctuation: PRC 93, 064910 (2016) β’ Saturation scale fluctuation: PRC 94, 044918 (2016)
