Zhen Gao University of Science and Technology of China November 21th, 2013

Transcription

1 Measurement of luminosity at 2.23GeV,2.4GeV, 2.8GeV and 3.4GeV Zhen Gao University of Science and Technology of China November 21th,

2 Outline Introduction Data Sets Measurement of Luminosity at 2.23GeV bhabha digamma dimuon Results at other energy points Summary Backup 2

3 Introduction 3

4 In e + e - collider experiment, the integrated luminosity is a basic parameter, which relates to the production cross section for e + e - >X. In principle, any process can be used to determine the integrated luminosity of the data set. The integrated luminosity could be measured by: L = N obs σ ε ε trig N obs : observed number of the events for the final state σ : production cross section for the same final states ε : detection efficiency ε trig : trigger efficiency 4

5 The QED processes of e + e γ e + e, e + e γ γγ and e + e γ μ + μ are most commonly used for determination of the luminosity because of their large production cross sections, the large detection efficiencies and the simplest final states topology. In this study,we use Bhabha process to calculate the luminosity and use Digamma process and Dimu process to check it 5

6 Data Set 6

7 Boss BESⅢ experiment data Data for R-scan at 2.23GeV, 2,4GeV, 2.8GeV and 3.4GeV (reconstructed at Boss 6.6.4) MC sample Babayaga3.5 :bhabha process digamma process dimuon process dipion process Lundlw: hadronic process KKMC: K + K - process 7

8 Measurement of luminosity at 2.23GeV 8

9 The setup of babayaga3.5 generator ApplicationMgr.DLLs += {"Babayaga"}; ApplicationMgr.TopAlg += {"Babayaga"}; Babayaga.Channel=1; // 1: e+e-->e+e-;2:e+e_->mu+mu-;3:e+e-->gamma gamma;4:e+e--->pi+pi- Babayaga.Ebeam=1.1162; // Ebeam = ECM/2 [GeV] Babayaga.MinThetaAngle=20.00; // minimum angle(deg.) Babayaga.MaxThetaAngle=160.00; //maximum angle(deg.) Babayaga.MinimumEnergy=0.01; //minimum energy (GeV) Babayaga.MaximumAcollinearity=180; //maximum acollinearity (deg.) Babayaga.RunningAlpha=1; //running alpha (0 = off, 1 = on) Babayaga.HadronicResonance=1; //hadronic resonances for ICH = 1 or 2 Babayaga.FSR_swich=1; //FSR switch for ICH = 2 (0 = off, 1 = on) Babayaga.MinEnerCutG=0.01; //minimum energy for CUTG = Y (GeV) Babayaga.MinAngCutG=5; //minimum angle for CUTG = Y (deg.) Babayaga.MaxAngCutG=21; //maximum angle for CUTG = Y (deg.) Babayaga.HBOOK = 0; // close babayaga.ntuple hbook output Babayaga.PHCUT = 0; Babayaga.CUTG = 0; 9

10 Events Selection(bhabha) MDC Track Vr < 1.0 cm Vz < 10.0 cm N + = N - = 1 Δθ < 10 Δϕ < 5 EMC Shower N cluster == 2 cosθ < 0.8 E shower > 0.65*E beam 10

11 Deposited energy in EMC Scale to the same luminosity of bhabha! Deposited energy of cluster + Deposited energy of cluster -

12 Back to back information After the energy cut,there are few background events left!to remove the bhabha events with FSR gamma which carry big energies, we need the two good candidate tracks are back-to back. 12

13 Cosθ distribution cosθ of e+ in EMC cosθ of e- in EMC 13

14 Compare data and MC(1) E(e+) after selection E(e-) after selection 14

15 Compare data and MC(2) Momentum of e+ in MDC Momentum of e- in MDC Black plots with error bars are data, red histograms are MC. All of them are in the laboratory frame. 15

16 Check the MC Momenta of e+ and e- in laboratory frame Momenta of e+ and e- in center-of-mass frame They are different! 16

17 Systematic error Uncertainty in the event selection criteria Uncertainty in the detection efficiency Uncertainty in the background estimation Uncertainty in the electron tracking Uncertainty from the generator 17

18 Uncertainty in the event selection criteria The difference of the ratios between the data and the Monte Carlo was defined by: sys = R data R MC 1,in which R = N cut N total Cut on cosθ: change cut range of cosθ Cut on Δθ : change cut range of Δθ Cut on Δφ: change cut range of Δφ Cut on deposited energy : change ratio factor (e+ and e- are separated) 18

19 Uncertainty in the detection efficiency The statistical uncertainty in the Monte Carlo efficiency is estimated by: 1 N (1 ε) ε where N is the number of Monte Carlo events for e + e (γ)e + e which are used in the determination of detection efficiency and ε is the detection efficiency. 19

20 Uncertainty in the background estimation Scale to the same luminosity of bhabha(2.638pb -1 ): s=2.23gev : Bkg N total N sur σ (nb) Scale N scale γγ 1,000, μ + μ - 1,000, π + π - 495, K + K - 500, qqbar 1,000, background level is very low! So we regard the uncertainty of background estimation as 0! 20

21 Uncertainty from the generator the uncertainty of babayaga3.5 is 0.5% 21

22 Uncertainty in the electron tracking We use two different methods : 1.use both MDC and EMC information 2.only use EMC information so, sys = R data R MC - 1, in which R = N method2 N method1 22

23 Method2(without MDC) N cluster == 2 E 1,2 > 0.65*E beam cosθ < < Δφ < 40 23

24 Δϕ Distribution ϕ1- ϕ2-180 in MC ϕ1- ϕ2-180 (data and MC) 24

25 Results of systematic error Source Δ sys (%) cosθ < Δθ < Δφ < Deposited energy of e Deposited energy of e MC statistics 0.09 Background estimation 0 Generator 0.50 Tracking efficiency 0.40 total

26 Events Selection(digam) MDC Track Vr < 1.0 cm Vz < 10.0 cm N + = N - = 0 EMC Shower N gamma = 2 cosθ < 0.8 E gamma > 0.65*E beam -4 < Δϕ < 2 Δθ < 10 26

27 Events Selection(add gamma conversion) MDC Track Vr < 1.0 cm Vz < 10.0 cm N + = N - = 1 EMC Shower N gamma == 1 cosθ < 0.8 E gamma > 0.65*E beam Combine MDC and EMC: E/P > 0.8 to ensure they are electron angle(e+ e-) < 15 M(e+ e-) < 0.1 GeV

28 Data Analysis(1) ϕ1- ϕ2-180 in EMC θ1+ θ2-180 in EMC Black plots with error bars are data, red histograms are babayaga MC. 28

29 Data Analysis(2) Cosθ distribution in EMC 29

30 Gamma conversion Angle(e+e-) < 15 and M(e+e-) < 0.1GeV 30

31 Background estimation Scale to the same luminosity of bhabha(2.638pb -1 ): s=2.23gev : Bkg N total N sur σ (nb) Scale N scale e + e - 1,000, μ + μ - 1,000, π + π - 495, K + K - 500, qqbar 1,000,

32 Results of systematic error Digamma : Gamma conversion : Source Δ sys (%) Deposited energy 0.1 Δθ < <Δφ < cosθ < MC statistics 0.07 Generator 1 Background estimation 0.01 Gamma detection efficiency 2 total 2.24 Source Δ sys (%) Deposited energy 3.30 cosθ < M(e+e-) < Angle(e+e-) < Gamma detection efficiency 1 Tracking efficiency 0.40 Background estimation 0 total 4.35 The weight of two process is 100:2.6, so the total systematic err is 2.32% 32

33 Events Selection(dimu) MDC Track Vr < 1.0 cm Vz < 10.0 cm N + = N - = 1 Δθ < 10 Δϕ < 5 EMC Shower cosθ < 0.8 E/P < 0.4 Isolated Nγ <= 2 If Nγ == 2,M(γγ) < 0.1GeV TOF tof + -tof - < 1ns 33

34 E/P ratio E/P of cluster+ Scale to the same luminosity of bhabha! E/P of cluster- Set E/P < 0.4 to remove almost all the bhabha and digamma events and many hadronic events. 34

35 Back to back information We need the two muon candidates are back-toback, so the angle cut is the same as bhabha 35

36 Tof information We cut the tof + -tof - < 1ns to remove the cosmicray 36

37 Potential background MC As the muon counter is closed at that moment, in order to remove the potential hadronic events, we require the number of isolated photons is not larger than 2 and the M(γ1γ2) is smaller than 0.1 GeV if Nγ ==2. 37

38 π 0 signal in Data 38

39 M(γ1γ2) cut Data and MC As η is heavier than π 0,so the hadronic tracks with η can not be back-to-back strictly. In Data, we can not find obviously η signal. But the cut for gamma is still needed. 39

40 Momentum distribution Momentum of u+ Momentum of u- 40

41 K + K - background We cut the P(u + ) range from 0.96GeV to 1.04GeV and show the P(u - ) distribution as followed. There are bins bumping around 1GeV, it s may be k + k - background. 41

42 π + π - and K + K - background Born cross section given by CLEO-C: Babar s result of σ(π + π - ) Babar s result of σ(k + K - ) σ(π + π - ) = 180pb σ(k + K - ) = 127pb 42

43 Background estimation Scale to the same luminosity of bhabha(2.638pb -1 ): s=2.23gev : Bkg N total N sur σ (nb) Scale N scale bhabha 1,000, γγ 1,000, π + π - 495, K + K - 500, qqbar 1,000, But in MC of qqbar, σ(π + π - ) = 1.27nb, σ(k + K - ) = nb (by learning the MC truth information in qqbar MC),so we can not use qqbar MC to estimate the background of dimu! 43

44 Results of systematic error Source Δ sys (%) cosθ < Δθ < Δφ < E/P of u E/P of u Δtof < M(γγ) < MC statistics 0.06 Background estimation 0.04 Generator 1 Tracking efficiency 2 total

45 Check the MC Momenta of u+u- in MC (Lab ) Momenta of u+u- in MC (cms ) The same as bhabha! 45

46 Results at 2.23GeV 2*Ebeam Process Bhabha Digamma Dimu MC generator Babayaga Babayaga Babayaga Events Number Cross-Section(nb) ± ± ±0.02 Efficiency(%) Luminosity(pb -1 ) 2.638±0.007± ±0.010± ±0.016±

47 Results at other energy points 47

48 @2.4GeV 2*Ebeam Process Bhabha Digamma Dimu MC generator Babayaga Babayaga Babayaga Events Number Cross-Section(nb) ± ± ±0.02 Efficiency(%) Luminosity(pb -1 ) 3.417± ± ±

49 @2.8GeV 2*Ebeam Process Bhabha Digamma Dimu MC generator Babayaga Babayaga Babayaga Events Number Cross-Section(nb) ± ± ±0.01 Efficiency(%) Luminosity(pb -1 ) 3.757± ± ±

50 @3.4GeV 2*Ebeam Process Bhabha Digamma Dimu MC generator Babayaga Babayaga Babayaga Events Number Cross-Section(nb) ± ± ±0.01 Efficiency(%) Luminosity(pb -1 ) 1.732± ± ±

51 P(u+)+P(u-) at 3.4GeV In data, there is bumping bins around 3.1GeV. This may be background from γj/ψ. So we add a cut to remove it.(in bhabha process, we can not see obviously J/Ψ signal ) 51

52 Summary 52

53 By analyzing the Bhabha scattering events, Digamma scattering events and Dimu scattering events, we measure the integrated luminosities of the data taken at s = 2.23GeV, 2.4GeV, 2.8GeV and 3.4GeV. These luminosities can be used in studies of R- scan. 53

54 Backup 54

55 Study of the difference in MC Distribution of P(e±) in MC versus ϕ from -180 to 0.The range of each histogram is 20 55

56 Distribution of P 3 (rec)/p 3 (truth) in MC versus ϕ(-180 to 0 ) 56

57 Distribution of Px(rec)/Px(truth) in MC versus ϕ(-180 to 0 ) 57

58 Py(rec)/Py(truth) 58

59 Pz(rec)/Pz(truth) 59