ebook img

Resummed differential cross sections for top-quark pairs at the LHC PDF

0.38 MB·
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Resummed differential cross sections for top-quark pairs at the LHC

IPPP/16/04 Resummed differential cross sections for top-quark pairs at the LHC Benjamin D. Pecjakb, Darren J. Scottb, Xing Wanga, Li Lin Yanga,c,d aSchool of Physics and State Key Laboratory of Nuclear Physics and Technology, Peking University, Beijing 100871, China bInstitute for Particle Physics Phenomenology, University of Durham, DH1 3LE Durham, UK cCollaborative Innovation Center of Quantum Matter, Beijing, China dCenter for High Energy Physics, Peking University, Beijing 100871, China We present state of the art resummation predictions for differential cross sections in top-quark pair production at the LHC. They are derived from a formalism which allows the simultaneous resummation of both soft and small-mass logarithms, which endanger the convergence of fixed- order perturbative series in the boosted regime, where the partonic center-of-mass energy is much largerthanthemasstothetopquark. WecombinesuchadoubleresummationatNNLL(cid:48) accuracy with standard soft-gluon resummation at NNLL accuracy and with NLO calculations, so that our results are applicable throughout the whole phase space. We find that the resummation effects on the differential distributions are significant, bringing theoretical predictions into better agreement with experimental data compared to fixed-order calculations. Moreover, such effects are not well described by the NNLO approximation of the resummation formula, especially in the high-energy tailsofthedistributions,highlightingtheimportanceofall-ordersresummationindedicatedstudies of boosted top production. 6 1 0 INTRODUCTION ItisthereforeimportanttoassesstheeffectsofQCDcor- 2 rections even beyond NNLO, in order to see whether the b gap between theory and data at high p can be bridged. e The 8TeV run of the LHC delivered about 20fb−1 of T F integrated luminosity to both the ATLAS and CMS ex- 5 periments. Among the many important results coming Forboostedtop-quarkpairswithhighpT therearetwo from these data, the properties of the top-quark have classes of potentially large contributions. The first is the ] been measured with unprecedented precision. At the Sudakov-type double logarithms arising from soft gluon h same time, theoretical calculations of top-quark related emissions. Thesecondcomesfromgluonsemittednearly p - observableshaveseensignificantadvancementsinthelast paralleltothetopquarks,resultinginlargelogarithmsof p few years. In particular, very recently the next-to-next- the form lnn(mt/mT), where mt is the top quark mass, he to-leadingorder(NNLO)QCDcorrectionstodifferential and mT ≡ (cid:112)m2t +p2T is the transverse mass of the top [ crosssectionsintop-quarkpair(tt¯)productionhavebeen quark. In [5], some of the authors of the current work calculated[1]. In[2],theCMScollaborationperformeda developed a formalism for the simultaneous resumma- 2 comprehensive comparison between their measurements tionofbothtypeoflogarithmstoallordersinthestrong v 0 [3] of the differential cross sections and various theoreti- coupling constant αs. In this Letter, we report the first 2 cal predictions, including those from the NNLO calcula- phenomenological applications of that formalism, giving 0 tion and those from Monte Carlo event generators with predictionsforthetop-quarkpT andthett¯invariantmass 7 next-to-leading order (NLO) accuracy matched to par- distributionsatthe8TeVLHC,andcomparingwithex- 0 ton showers. The overall agreement between theory and perimental measurements as well as the NNLO calcula- . 1 data is truly remarkable, which adds to the success of tions when possible. With an eye to the future, we also 0 the Standard Model (SM) as an effective description of present predictions for the 13TeV LHC, where NNLO 6 Nature. results are not yet available. 1 : However,apersistentissueinthe8TeVresultsisthat v i the transverse momentum (pT) distribution of the top Our main finding is that the higher-order effects con- X quark is softer in the data than in theoretical predic- tained in our resummation formalism significantly alter r tions, i.e., the experimentally measured differential cross the high-energy tails of the p and tt¯invariant mass dis- a T section at high p is lower than predictions from event tributions, softening that of the p distribution but en- T T generators or from NLO fixed-order calculations [3, 4]. hancingthatofthett¯invariantmassdistribution. These While the NNLO corrections bring the fixed-order pre- effects bring our results into better agreement with the dictions into better agreement with the CMS data, as experimental data compared to pure NLO fixed-order notedin [1]and [2], there isstill somediscrepancy inthe calculations. Interestingly, for the case of the p distri- T high-p binswherep >200GeV. Giventheimportance bution,thissofteningofthespectrumisslightlystronger T T ofthett¯productionprocessasastandardcandleforval- than the similar effect displayed in recent NNLO results, idating the SM and as an essential background for new andleadstoabettermodelingofthep >200GeVpor- T physics searches, it would be disconcerting ifthis feature tion of the CMS data [3]. We comment further on this were to persist at higher p values and with more data. fact in the conclusions. T 2 FORMALISM Mellin space) as well as the NLO fixed-order result cal- culated in [10] and implemented in MCFM [11]. The Our predictions are based on the factorization and precise matching formula can be found in [6]. After such resummation formula derived in [5]. The technical de- a matching procedure, we denote the final accuracy of tails will be given in a forthcoming article, although the our predictions, which are valid throughout phase space, main elements have already been sketched out in [6]. In as NLO+NNLL(cid:48). the kinematic situation where the top quarks are highly It would be desirable to match with the recent NNLO boostedandtheeventsaredominatedbysoftgluonemis- results in [1] to achieve NNLO+NNLL(cid:48) accuracy. How- sions, theresummedpartonicdifferentialcrosssectionin ever, at the moment NNLO results are only available Mellin space can be written as for fixed (i.e., kinematics-independent) factorization and renormalization scales µ ∼ µ ∼ m , whereas for the f r t (cid:34) study of differential distributions over large ranges of (cid:101)cij(N,Mtt¯,mt,µf)=Tr U(cid:101)ij(µf,µh,µs)Hij(Mtt¯,µh) phase space we consider it important to follow common practice and use dynamical (i.e., kinematics-dependent) (cid:18) M2 (cid:19)(cid:35) scale choices. Therefore, such an improvement over our ×U(cid:101)i†j(µf,µh,µs)×s(cid:101)ij lnN¯2µtt¯2,µs (1) result is not currently possible, and we leave it for the s future. (cid:32) (cid:33) m ×U(cid:101)D2(µf,µdh,µds)CD2(mt,µdh)s(cid:101)2D lnN¯µt ,µds , ds PHENOMENOLOGY where for simplicity, we have suppressed some variables in the functional arguments which are unnecessary for In the following we present NLO+NNLL(cid:48) predictions the explanations below. In the above formula, Mtt¯ is for the Mtt¯ and pT distributions at the LHC. In all the invariant mass of the tt¯pair (which can be related our numerics we choose m = 173.2GeV and use t to the pT of the top quark in the soft limit through a MSTW2008NNLO PDFs [12]. For pT distributions, the change of variables), N is the Mellin moment variable, default values for the factorization scale and the four and N¯ ≡ NeγE with γE the Euler constant. The soft renormalizationscalesarechosenasµf =mT,µh =Mtt¯, limit corresponds to N → ∞ in Mellin space. The four µs = Mtt¯/N¯, µdh = mt and µds = mt/N¯. For Mtt¯ dis- coefficient functions Hij, s(cid:101)ij, CD and s(cid:101)D encode con- tributions, the only difference is µf =Mtt¯. We estimate tributions from four widely-separated energy scales Mtt¯, scaleuncertaintiesbyvaryingthefivescalesaroundtheir Mtt¯/N¯, mt and mt/N¯, respectively. The presence of the defaultvaluesbyfactorsoftwoandcombiningtheresult- fourscalesleadstothetwotypesoflargelogarithmsdis- ingvariationsofdifferentialcrosssectionsinquadrature; cussedintheintroduction. Incorrespondencewiththese we do not consider uncertainties from PDFs and α in s four physical scales, there are four unphysical renormal- this Letter. The hadronic differential cross sections are izationscalesµh,µs,µdh andµds,oneforeachcoefficient first evaluated in Mellin space at a given point in phase function. Thephilosophyofresummationistochoosethe space, and we then perform the inverse Mellin transform four unphysical scales to be around their corresponding numerically using the Minimal Prescription [13]. This physical scales, so that the four coefficient functions are procedure relies on an efficient construction of Mellin- free of large logarithms and are well-behaved in fixed- transformed parton luminosities, for which we use meth- order perturbation theory. One can then use renormal- ods outlined in [14, 15]. ization group (RG) equations to evolve these functions The differential cross sections considered below span to the factorization scale µ in order to convolute with f severalordersofmagnitudewhengoingfromlowtohigh the parton distribution functions (PDFs) and obtain the valuesofpT orMtt¯. Inordertobetterdisplaytherelative hadronic cross sections. The effects of the RG running sizes of various results, we show in the lower panel of areencodedinthetwoevolutionfactorsU(cid:101)ij (forHij and eachplotthedifferentialcrosssectionsnormalizedtoour s(cid:101)ij) and U(cid:101)D (for CD and s(cid:101)D), which resum all the large default prediction, i.e., the ratio defined by logarithms to all orders in α in an exponential form. s At the moment, the four coefficient functions are dσ Ratio≡ . (2) known to NNLO [5, 7, 8], while the two evolution fac- dσNLO+NNLL(cid:48)(cid:0)µ =µdefault(cid:1) i i tors are known to next-to-next-to-leading logarithmic (NNLL) accuracy [5]. Such a level of accuracy is usu- Fig. 1 compares our NLO+NNLL(cid:48) resummed predic- ally referredto asNNLL(cid:48) inthe literature, andweadopt tion for the normalized top-quark p distribution to T that nomenclature here. While the formula (1) is only the CMS measurement [3] in the lepton+jet channel at √ applicable in the boosted soft limit, we can extend its the LHC with a center-of-mass energy s = 8TeV. domainofvaliditybycombiningitwithinformationfrom Also shown is the NNLO result from [1], which adopted NNLL soft gluon resummation derived in [9] (recast into by default the renormalization and factorization scales 3 ss-3/GeV) (10/dp) d(1/T 10 NNCLNMOLSmLmO+ Hf(t N =l=C+(N m j( )1d18Le7/ Lf2T 3'=,e.1 2Vm, 2Gt ) =em V1T73.3 GeV) s (pb/GeV)/dMdtt 101 mLmHft ==C ( 1187/ 2T3,e.12NNAV, 2GTLL)OOL eMAV+tStN N(l+LjL)' 1 10-1 10-2 10-1 10-3 Ratio 11..21 Ratio 11..24 1.0 1.0 0.8 0.9 0.6 0.8 0.4 0 50 100 150 200 250 300 350 400 400 600 800 1000 1200 1400 1600 p (GeV) M (GeV) T tt FIG.1. Resummedprediction(blueband)forthenormalized FIG. 3. Resummed prediction (blue band) for the absolute top-quark pT distribution at the 8 TeV LHC compared with Mtt¯ distribution at the 8TeV LHC compared with ATLAS CMS data (red crosses) [3] and the NNLO result (magenta data (red crosses) [16] and the NLO result (magenta band). band) [1]. The lower panel shows results normalized to the default NLO+NNLL(cid:48) prediction. taluncertaintyisratherlargeduetolimitedstatistics, it s (pb/GeV)/dpdT1100--21 mLHt =C 187 T3e.2NNAV GTLLOOLeAV+SNNLL' itmIsnioainFsnsstigelho.reeg2rsaetwri,intesghisnmhtcoosewbcioetscmuioscpmhaeexraepmceiootcrmtewepdriteahltrehivstaoahtnne.tthtaThetehhoseoirgefNthtNieacrLaneldOnpesrrrmeegsdiaueilcsllt--. mf = (1/2,1,2) mT for such high pT values is not yet available, so we com- 10-3 pareinsteadwiththeNLOresultcomputedusingMCFM withMSTW2008NLOPDFsanddynamicalrenormaliza- tion and factorization scales, whose default values are 10-4 µ = µ = m . Scale uncertainties of the NLO results r f T areestimatedthroughvariationsofµ =µ byafactorof r f 10-5 two around the default value. From the plot one can see Ratio 11..48 that the NLO result calculated in this way does a good job in estimating the residual uncertainty from higher 1.0 order corrections, as the resummed band lies almost in- 0.6 300 400 500 600 700 800 900 1000 1100 1200 side the NLO one up to p = 1.2TeV. On the other p (GeV) T T FIG. 2. Resummed prediction (blue band) for the absolute hand, the inclusion of the higher-order logarithms in the p distributionatthe8TeVLHCintheboostedregioncom- NLO+NNLL(cid:48) result significantly reduces the theoretical T pared with the ATLAS data (red crosses) [4] and the NLO uncertainty, which is crucial for future high precision ex- result (magenta band). periments at the LHC. Our formalism is flexible and can be applied to other differential distributions as well. To demonstrate this µr = µf = mt, and also used a slightly different top- fact, in Fig. 3 we show the NLO+NNLL(cid:48) resummed pre- quark mass, mt = 173.3GeV. At low pT, it is clear dictionforthetop-quarkpairinvariantmassdistribution that both the NLO+NNLL(cid:48) and the NNLO results de- alongwithameasurementfromtheATLAScollaboration scribe the data fairly well. With the increase of pT, it [16] at the 8TeV LHC. Since the NNLO result in [1] for appearsthattheNNLOpredictionsystematicallyoveres- this distribution has an incompatible binning, it is cur- timatesthedata,althoughthereisstillagreementwithin rently not possible to include it in the plot, so we show errors. Ontheotherhand,withthesimultaneousresum- instead the NLO result computed with the same input mation of the soft gluon logarithms and the mass log- as in Fig. 2, but this time with the default scale choice arithms and also with the dynamical scale choices, our µr =µf =Mtt¯. OnecanseefromtheplotthattheNLO NLO+NNLL(cid:48) resummed formula produces a softer spec- result with this scale choice is consistently lower than trum which agrees well with the data. the experimental data. The resummation effects signif- In[4],theATLAScollaborationcarriedoutameasure- icantly enhance the differential cross sections, especially mentofthetop-quarkpT spectruminthehighly-boosted at high Mtt¯. As a result, the NLO+NNLL(cid:48) prediction regionusingfat-jettechniques. Althoughtheexperimen- agrees with data quite well. We have found that choos- 4 s (pb/GeV)/dpdT11001--0211 mLmHft ==C 1(1173/32 T.,21NNe ,VGLL2OO)e mV+NTNLL' Relative corrections 0001....4680 mLmHft ==C 1M173tt3 /T2.2e VGeV aBNeyNoLnOd cNoNrrLeOction 0.2 10-3 0.0 10-4 500 1000 1500 2000 2500 3000 3500 4000 M (GeV) tt Ratio11100011---..76524 Relative corrections-000...011 mLmHft ==C 1m173T3 T.2e VGeV aBNeyNoLnOd cNoNrrLeOction 1.0 0.80 200 400 600 800 1000 1200 1400 1600 1800 2000 -0.2 400 600 800 1000 1200 1400 1600 1800 2000 p (GeV) p (GeV) T T FIG. 5. Relative sizes of the corrections at approximate (pb/GeV)Mtt 101 NNLLOO+NNLL' NEqN.L(O3)(abnludet)haenedxpbleaynoantdion(bslathcker)e, wfoirthprreecsipseecdtetfionNitiLoOns..See s/dd 10-1 mLHt =C 11733 T.2e VGeV 10-2 mf = (1/4,1/2,1) Mtt 10-3 up to pT =2TeV and Mtt¯=4.34TeV, contrasted with 10-4 the NLO results. Note that for the Mtt¯distribution, we 10-5 have changed the default µf to a lower value Mtt¯/2 for 10-6 the reasons explained above. The plots exhibit similar patterns as observed at 8TeV, namely that the higher- 10-7 Ratio 0111....8024 oprTdedrisrterisbuumtimonatbiounteeffnehcatnscseetrhveattoofstohfteeMn ttt¯heditsatriliboufttiohne 0.6 compared to a pure NLO calculation. 0.4 0.2 500 1000 1500 2000 2500 3000 3500 4M00 0(GeV) Asmentionedbefore,wewouldliketomatchourcalcu- tt FIG. 4. Resummed predictions (blue bands) for the p and lations with the NNLO results when they become avail- T Mtt¯distributionsatthe13TeVLHCcomparedwiththeNLO able in the future. We end this section by discussing results (magenta bands). the expected effects of such a matching, by estimating the size of resummation corrections beyond NNLO. We do this in Fig. 5, where the relative sizes of the beyond- ing the default renormalization and factorization scales NNLO corrections generated through the resummation to be half the invariant mass increases the fixed-order formula are displayed as a function of Mtt¯ or pT with cross section and therefore mimics to some extent the the default scale choices. The exact NNLO results for resummationeffects. Infact,thisprocedurehasbeenex- these scale choices are not yet available, so we show in tensively employed in the literature for processes such as comparison the relative sizes of the approximate NNLO Higgsproduction[17],wherehigher-ordercorrectionsare (aNNLO) corrections obtained by expanding and trun- also large. Consequently, it may be advisable to employ cating our resummation formula to that order. More a renormalization and factorization scale of the order of precisely, the blue and black curves in Fig. 5 correspond Mtt¯/2infixed-ordercalculations(andMonteCarloevent to generators), and we shall use this choice when studying the Mtt¯distribution at the 13 TeV LHC below. The LHC has started the 13TeV run in 2015. So far dσaNNLO−dσNLO aNNLO correction≡ , (3) there are only two CMS measurements [18, 19] of dif- dσNLO ferential cross sections for tt¯production, based on just dσNLO+NNLL(cid:48) −dσaNNLO 42pb−1 of data. The resulting experimental uncertain- Beyond NNLO≡ dσNLO , ties are therefore quite large and it is not yet possible to probe higher pT or Mtt¯values. Nevertheless, in the near future there will be a large amount of high-energy data, where dσaNNLO refers to the approximate NNLO result. whichwillenablehigh-precisionmeasurementsoftt¯kine- The figure clearly shows that corrections beyond NNLO maticdistributions,alsointheboostedregime. InFig.4 are significant in the tails of the distributions, especially we show our predictions for the pT and Mtt¯ spectrum in the case of the Mtt¯distribution. 5 CONCLUSIONS AND OUTLOOK for collaboration on many related works. X. Wang and L.L.YangaresupportedinpartbytheNationalNatural ScienceFoundationofChinaunderGrantNo. 11575004. In this Letter we have presented new resummation D. J. Scott is supported by an STFC Postgraduate Stu- predictions for differential cross sections in tt¯ produc- dentship. tion at the LHC. The predictions include the simul- taneous resummation to NNLL(cid:48) accuracy of both soft and small-mass logarithms, which endanger the conver- genceofthefixed-orderperturbativeseriesintheboosted regimewherethepartoniccenter-of-massenergyismuch [1] M. Czakon, D. Heymes and A. Mitov, arXiv:1511.00549 largerthanthemassofthetopquark. Thisresummation [hep-ph]. is matched with both standard soft-gluon resummation [2] The CMS Collaboration, CMS-PAS-TOP-15-011. at NNLL accuracy and fixed-order NLO calculations, so [3] V. Khachatryan et al. [CMS Collaboration], Eur. Phys. that our results are applicable in the whole phase space. J. C 75, 542 (2015) [arXiv:1505.04480 [hep-ex]]. Such predictions for tt¯ differential distributions at the [4] G. Aad et al. [ATLAS Collaboration], arXiv:1510.03818 [hep-ex]. LHC are not only the first to be calculated in Mellin [5] A.Ferroglia,B.D.PecjakandL.L.Yang,Phys.Rev.D space, but also represent the highest resummation accu- 86, 034010 (2012) [arXiv:1205.3662 [hep-ph]]. racyachievedtodate,namelyNLO+NNLL(cid:48). Ourresults [6] A. Ferroglia, B. D. Pecjak, D. J. Scott and L. L. Yang, are thus a major step forward in the modeling of high- arXiv:1512.02535 [hep-ph]. energytailsofdistributions,whichisofgreatimportance [7] A. Broggio, A. Ferroglia, B. D. Pecjak and Z. Zhang, for new physics searches. JHEP 1412, 005 (2014) [arXiv:1409.5294 [hep-ph]]. [8] A. Ferroglia, B. D. Pecjak and L. L. Yang, JHEP 1210, The agreement of NLO+NNLL(cid:48) predictions with data 180 (2012) [arXiv:1207.4798 [hep-ph]]. indicates the value of including resummation effects and [9] V. Ahrens, A. Ferroglia, M. Neubert, B. D. Pecjak, using dynamical scale settings correlated with p or L. L. Yang, JHEP 1009, 097 (2010) [arXiv:1003.5827 T Mtt¯ when studying differential distributions. Interest- [hep-ph]]. ingly,inthecaseofnormalizedp distributionmeasured [10] P.Nason,S.DawsonandR.K.Ellis,Nucl.Phys.B327, T by the CMS collaboration [3], the NLO+NNLL(cid:48) calcu- 49(1989);M.L.Mangano,P.NasonandG.Ridolfi,Nucl. Phys. B 373, 295 (1992); S. Frixione, M. L. Mangano, lation produces a slightly softer spectrum than recent P. Nason and G. Ridolfi, Phys. Lett. B 351, 555 (1995) NNLO predictions (which use a fixed scale setting where [hep-ph/9503213]. µf =µr =mt bydefault), thusachievingabetteragree- [11] J.M.CampbellandR.K.Ellis,Nucl.Phys.Proc.Suppl. ment with the data. However, we emphasize that the 205-206, 10 (2010) [arXiv:1007.3492 [hep-ph]]. optimaluseofresummationistosupplementNNLOcal- [12] A.D.Martin,W.J.Stirling,R.S.ThorneandG.Watt, culations, not to replace them. With this in mind, we Eur. Phys. J. C 63, 189 (2009) [arXiv:0901.0002 [hep- ph]]. have studied the size of corrections beyond NNLO en- [13] S. Catani, M. L. Mangano, P. Nason and L. Trentadue, coded in our resummation formula, and found that their Nucl. Phys. B 478, 273 (1996) [hep-ph/9604351]. effects are significant in the high-energy tails of distri- [14] M. Bonvini and S. Marzani, JHEP 1409, 007 (2014) butions, especially for the tt¯invariant mass distribution [arXiv:1405.3654 [hep-ph]]. where they enhance the differential cross section. It will [15] M. Bonvini, arXiv:1212.0480 [hep-ph]. thereforebeanessentialandinformativeexercisetopro- [16] G. Aad et al. [ATLAS Collaboration], arXiv:1511.04716 duceNNLO+NNLL(cid:48)predictionsonceNNLOcalculations [hep-ex]. [17] C. Anastasiou, C. Duhr, F. Dulat, F. Herzog and are available with dynamical scale settings. B. Mistlberger, Phys. Rev. Lett. 114, 212001 (2015) Acknowledgments: We would like to thank Alexander [arXiv:1503.06056 [hep-ph]]. Mitov for providing us the results of the NNLO cal- [18] The CMS Collaboration, CMS-PAS-TOP-15-005. culations in [1]. We are grateful to Andrea Ferroglia [19] The CMS Collaboration, CMS-PAS-TOP-15-010.

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.