Underlying Event Studies for LHC Energies
aa r X i v : . [ h e p - ph ] J a n Underlying Event Studies for LHC Energies
Gergely Gábor Barnaföldi ∗ , András G. Agócs ∗ ,† and Péter Lévai ∗ ∗ KFKI Research Institute for Particle and Nuclear Physics of the HAS,29-33 Konkoly-Thege M. Str. H-1121 Budapest, Hungary † Eötvös University,1/A Pázmány Péter Sétány, H-1117 Budapest, Hungary
Abstract.
Underlying event was originally defined by the CDF collaboration decades ago. Here we improvethe original definition to extend our analysis for events with multiple-jets. We introduce a definitionfor surrounding rings/belts and based on this definition the jet- and surrounding-belt-excluded areaswill provide a good underlying event definition. We inverstigate our definition via the multiplicityin the defined geometry. In parallel, mean transverse momenta of these areas also studied in proton-proton collisions at √ s = Keywords: underlying event, jet physics, LHC, proton-proton collisions
PACS:
INTRODUCTION
Underlying event (UE) was originally defined by the CDF Collaboration [1] and usedto investigate properties of the remaining of the events, after jets were identified andremoved from there. The CDF definition of the underlying event is a simple tool inorder to work, however detailed structure or information on off-jet particles cannot beobtained. On the other hand the definition is not capable to analyze more than 2-jetstructures. This motivate us to develop a new definition for the underlying event.To enhance the information content to be extracted from underlying events, we mod-ified the above CDF’s definition introducing multiple surrounding belts (SB) around theidentified jets [2, 3]. This new definition is immediately leads a more detailed analysisof the underlying event, even in case of multiple jets. On the other hand, as a specificcase of our new method, one can get the originally extracted physical observables cor-responding to the analysis based on the CDF-definition.In this short contribution we present the basic properties of the two ways of definingunderlying event. We recall the original CDF-based and our new definition of theunderlying events, which will be compared. We used two physical quantities for ourcomparison: (i) the average hadron multiplicity within the defined areas and (ii) themean transverse momenta versus multiplicity in the given regions. Quantities wereinvestigated for both definition in parallel.Our analysis is based on jet production and identification in proton-proton collisionsat 7 TeV. We used the LHC10e14 jet-jet sample generated by PYTHIA6.2 [4] frameworkwith cone-based UA1 [5] jet finder.
ENERALIZED DEFINITION OF THE UNDERLYING EVENT
Any definition of underlying event should strongly depend on a jet-identification methodapplied in the analysis. There are various state-of-the-art development on this direc-tion [7, 8], which are very promising. On the other hand there are still a problematicsof these definitions – the strong process dependence. E.g. changing from proton-protonto nucleus-nucleus collisions need to re-tune the properties of the algorithms in order tofind and separate jets and the baseline/background of each event [9].The CDF-based underlying event definition corresponds to the jet identification incase of a one- (or two-) jet events. Near side jets easily define the towardand the oppositeawayregions of the event geometry. Our original concept was to improve the CDF-baseddefinition on a two-folded way: • to develope an underlying event definition which is capable to handle multijetevents. • to investigate the surrounding areas around identified jets, even without majorchanges of the jet-findig parameters in a case of nucleus-nucleus collision.These requirements are led us to the definition of surrounding belts on the basis of theevent-background such as ’underlying event’, which completely satisfy our requirementsabove.In our method, we are using jet-finding algorithms also. We define jets, than based onthe physical properties of the concentrical surrounding belts and the remaining particlemultiplicities, a better background or baseline can be provided. On the other hand theanalysis of the surrounding area around the identified jets, can even give feedback on thegoodness of the jet finding parameters.On Fig. 1 the visual comparison of the two definitions can be seen. Left side ofthe figure is for the CDF-based definition, the right side displays the SB-based one.The two definitions can be summarized in a following ways, using the azimuth, F and(pseudo)rapidity, ( h or) y plane: CDF-based definition of the underlying event is based on the subtraction of two areasof the whole measured acceptance: one around the identified near jet (towardregion) and another to the opposite (away) direction. Both regions are DF × D h -slices of the measured acceptance around the near jet and to the opposite, with thefull
D h range and DF = ± o in azimuth. SB-based definition uses all identified jets of the event to subtract them from the back-ground. Each jet can have an approximate dial-like area, around which concentricbands (or rings) can be defined. If a jet cone angle, R = p DF + D h is given, afirst ’ SB ’ and a second ’ SB ’ surrounding belt can be defined for any jet with thethicknesses of d R SB and d R SB , respectively. Generally, d R SBi = . R ≈ . − d R SBi values, similar (but not the same) area can be cov-ered as in the original CDF-based definition. In this way the two model can becomparable too.
IGURE 1.
The schematic view of the underlying event (UE) defined by the CDF (left panel) and thesurrounding belts (SB, rightpanel). Details are in the text. (Color online.)
Now we investigate the basic properties of the areas and parallel the physical quanti-ties for the selected regions.
COMPARISON OF UE DEFINITIONS
Here we compare the details of the CDF- and surrounding belt (SB) based underlyingevent definitions. For our test we used PYTHIA6-simulated [4] proton-proton collisions(Perugia-0 tune [6]), namely LHC10e14 jet-jet at 7 TeV center-of-mass energy with150 ,
000 events. This sample contains jets identified by UA1 method [5]. We restrictedour analysis to the settings of p T HardMin =
10 GeV/c and p T HardMax =
20 GeV/c.Primarily we investigated the multiplicities of various geometrical regions of thegenerated events based on the full sample. After applying UA1 jet finding algorithm toidentify jets, we compared the selected areas using both CDF- and SB-based definitionsof the underlying events. On left panel of Fig. 2 we plotted the multiplicities, N i of theCDF-selected areas versus the total event multiplicity. Here N i refers for followings: themultiplicities of the identified ’leading/near jet’ (bluesquares), the jet-excluded ’toward’area (green dots), the ’away’ side area to the opposite direction (purple dots), and theCDF-defined underlying is event, ’transverse’ (pink dots). The right side of the Fig. 2stands for the SB-based underlying event definition with more areas: the multiplicities ofthe identified leading jet bluesquares), the away side jet (bluedots), multiplicity for thesurrounding belts, SB lead , , SB lead , , SB away , , and SB away , are open red squares, openpurple triangles, open red circles, open purple diamonds respectively. Finally orange ot N0 20 40 60 80 100 120 i N leading jettowards w/o jetaway region (CDF)transverse tot N0 20 40 60 80 100 120 i N leading jetaway jet lead,1 SB lead,2 SB away,1 SB away,2 SB UE FIGURE 2.
The multiplicity, N i for the selected areas as the multiplicity of the total event, N tot .Underlying event regions are defined on the left panel for the CDF-based and on the right panel forthe surrounding belt based definitions. More details are in the text. (Color online.) crosses denote multiplicity for the newly defined underlying event U E outside all jets.(Note, all color in accordance with the areas of Fig 1 above.)Fig. 2 shows multiplicity in almost all regions: N i increases almost linearly with thetotal multiplicity, in the N tot <
120 region of the event for both cases. In case of the CDF-based definition, the away region gives the biggest contribution, and the jet belongs tothe smallest one. The transverse (underlying event) area lies between the two extremalcontribution. Moreover, it is interesting to see, after excluding the jet from the towardregion, the remaining area has almost the same multiplicity as the underlying event.This shows the goodness of the jet finding algorithms and the "safety" of the CDF-based underlying event definition (e.g. 1/3 of the whole acceptance far from any jet-contaminated areas).The multiplicity relations of the SB-based definition differs from the CDF-based. Thenear jet has the same contribution, away side jet and the SB i s have small fraction fromthe N tot – due to the small areas. On the other hand, the newly defined underlying event, U E dominates the event multiplicity since it has almost the whole acceptance.In general the multiplicity fraction of the defined areas are almost proportional to thegeometrical surface, only the jet-content part violates this dependence, as Fig. 2 displays.Thus, the SB-based U E has larger multiplicity comparing to the CDF-based one, whichmight gives better statistics for an underlying event analysis.Secondly, the mean transverse momentum h p T i of the selected areas is investigatedincluding especially the underlying event. We plotted the h p T i vs. the multiplicity of thetotal event, N tot and the h p T i vs. multiplicity of the CDF-based and SB-based underlyingevents.On Fig. 3 we display the h p T i vs. the multiplicity of the total event, N tot for bothCDF-based (left panel) and SB-based (right panel) underlying event. We use the samecolor and mark encoding for the selected regions of the event as on Fig. 2 above.We found the mean- p T distributions of the regions are similar in proton-proton col-lisions. The identified leading jet has the highest values h p T i leading jet ∼ − h p T i leading jet ∼ − N tot , for both ot N0 20 40 60 80 100 120 [ G e V / c ] i > T < p leading jettowardstowards w/o jetaway region (CDF)transverse tot N0 20 40 60 80 100 120 [ G e V / c ] i > T < p leading jetaway jet lead,1 SB lead,2 SB away,1 SB away,2 SB UE FIGURE 3.
The average transverse momenta versus the total multiplicity of the events, N tot . Regionsdefined on the left panel are for the CDF-based and on the right panel are for the SB-based definitions.More details are in the text. (Color online.) definition’s cases. The mean- p T for both underlying event cases are the same with theconstant value h p T i UEi ∼ . SB i have also similar, but a slightly higher h p T i SBi ∼ . h p T i towards ∼ − h p T i away ∼ . − h p T i away jet ∼ − UE/CDF
N0 20 40 60 80 100 120 [ G e V / c ] i > T < p leading jettowardstowards w/o jetaway region (CDF)transverse UE2
N0 20 40 60 80 100 120 [ G e V / c ] i > T < p leading jetaway jet lead,1 SB lead,2 SB away,1 SB away,2 SB UE FIGURE 4.
The average transverse momenta versus the multiplicity of the underlying evens. Regionsare defined on the left panel for the CDF UE and on the right panel for the surrounding belt based UEdefinitions. More details are in the text. (Color online.)
Finally on Fig. 4 we compared mean- p T values vs. to the self-definition-given under-lying events: CDF-based UE, N UE / CDF on leftpanel and SB-based N UE on rightpanel.Colors and marks are the same as above figures.) Here, the comparison shows slightdifference between the panels. A stronger decrease in the highest h p T i -content regionspresent compared to Fig. 3. Furthermore, changing from N tot to N UE / CDF and N UE theseparation of the curves are more clear in both cases, especially at the largest h p T i val-ues. In parallel the mean- p T values for the underlying event are almost the same for theaverage multiplicity events and slightly higher for the rare ones. CONCLUSIONS
We studied our new underlying event definition in √ s = p T vs. multiplicities for the CDF-based and our SB-based definition.We found the multiplicity fraction of the defined regions are almost proportional to thegeometrical surface, only the jet-content part differs, due to the separation (or inclusion)of the leading or away side jets. The SB-based underlying event, U E found to havelarger multiplicity comparing to the CDF-based, N UE / CDF one, which might gives betterstatistics for the underlying event analysis.The mean- p T vs. N tot analysis led us to compare both definition on the same level.We got the same dependence of the underlying event for both, CDF- and SB-basedcases. On the other hand, the above mentioned jet and near/away-area handling leads todifferences.Finally, we compared our definitions by the mean- p T vs. the self-defined underlyingevent multiplicities, namely N UE / CDF and N UE . Our results have shown both definitionis reliable, and – due to the generalized definition of the surrounding belt based underly-ing event – multiple jets and detailed analysis of the surrounding areas can be performedin the future. ACKNOWLEDGMENTS
This work was supported by Hungarian OTKA NK77816, PD73596 and Eötvös Univer-sity. Authors (GGB & PL) are appreciate the local support by the UNAM, Mexico andGGB thanks for the János Bolyai Research Scholarship of the HAS.
REFERENCES
1. A. A. Affolder et al. [CDF Collaboration], Phys. Rev.
D65 , 092002 (2002).2. P. Lévai and A. G. Agócs, PoS
EPS-HEP2009 , 472 (2009).3. A. G. Agócs, G. G. Barnaföldi, and P. Lévai, arXiv:1011.5363 [hep-ph], accepted for the Proc. ofHot Quarks 2010 Workshop .4. T. Sjostrand, S. Mrenna, and P. Z. Skands, JHEP , 026 (2006).5. G. Arnison etal. [UA1 Collaboration], CERN-EP/83-118, Phys.Lett.
B132 , 214 (1983).6. P. Z. Skands, MCNET-10-08, CERN-PH-TH-2010-113, Phys.Rev.
D82 , 074018 (2010).7. G. P. Salam, Eur. Phys. J.
C67 , 637-686 (2010).8. M. Cacciari, G. P. Salam, S. Sapeta, JHEP , 065 (2010).9. S. Salur, Nucl. Phys.