1 / 14

Outline: (1) The data sample 2001 + 2002 (2) Some news on the analysis method

f  hp 0 g with h  p + p - p 0. 2000 data  197 candidates / 16 pb -1  4  4 estimated background  19% efficiency C.Bini D.Leone KLOE Memo 250 04/02 KLOE Collab. Phys.Lett.B536 (2002). spectrum + combined fit. Outline:

Download Presentation

Outline: (1) The data sample 2001 + 2002 (2) Some news on the analysis method

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. f  hp0g with h  p+p-p0 2000 data  197 candidates / 16 pb-1  4  4 estimated background  19% efficiency C.Bini D.Leone KLOE Memo 250 04/02 KLOE Collab. Phys.Lett.B536 (2002) spectrum + combined fit Outline: (1) The data sample 2001 + 2002 (2) Some news on the analysis method (3) Efficiency revised (4) Background revised (5) Data: spectrum + “phi-curve” (6) Data-MC comparison

  2. The data sample • “good” runs: • luminosity value ok • good s value  (a) used in kin.fits •  (b) for “phi-curve” • removed trigger problems (KLOE Memo 281) • “peak” runs • 1018< s <1021 MeV 2001: pb-1 full sample 140.4 “good” runs 137.0 “peak” runs 136.4 100 evts 1 evt 2002: pb-1 full sample 264.9 “good” runs 260.8 “peak” runs 245.2 100 evts 1 evt Lum (nb-1 / 0.2 MeV) vss Full data sample  397.8 pb-1 “good”  381.6 pb-1 “peak”

  3. (2) Some news on the analysis method Kinematical fits are done numerically using MINUIT (“penalty function method”) N = number of measurements per event = 3X2 + 5X5 = 31 Xkmeas = measured quantities (momenta, energies, positions, times) Xkfit = parameters of the fit NC = number of constraints = 4 + 5 + (3) Ci = constraints (functions of the parameters) li = arbitrary parameters (in principle   ) The result has not to depend on l • On data and Montecarlo samples • Studied the l dependence: • Large “plateau” observed for data • and Montecarlo: • Small l  more events enter (mostly background) • Large l  loss of events (MINUIT “crisis”) MC data l values at “plateau center” 1/l (MeV)

  4. (3) Efficiency revised • Used MC with accele default (based on 2000) • Corrections on data / MC for photons and tracks (based on 2000) • Weighted M(hp) distribution using the curve obtained from 2000 data • Cuts: • 2T from vertex ( R < X cm |Z| < Y cm) BPOS used • 5 photons ( > 10 MeV ) • kin.fit 1 p(c2) > 5% • at least 1 “good” combination • kin.fit 2 (on all “good” combinations) p(c2) > 5% • E(rad) > 20 MeV M(hp) (MeV)

  5. (4) Background revised • Expected background ~ few % from MC but checked with data • Main sources: • final state s(equiv.) MC available Leq • f  hp0g  p+p-g p0 9.6 pb 580 • e+e- wp0  p+p-p0 p0 4.7 nb70 • f  hg  p+p-p0 g 8.4 nb 30 • f  KSKL  p+p-p0 p0p0 50 nb 4.2 • f  KSKL  p0p0 p+p-p0  p0p0 pmn  p0p0 pen 20 nb 9.3 • The KSKL final state are considered for KL decaying R < 25 cm Results of selection chain application: 2 wp0 events  11 events on the “peak” sample 1 KSKL  p0p0 p+p-p0 event  41 events on the “peak” sample No events from other channels  < 100 events (notice: 1 wp0 enters for an accidental; 1 wp0 for a splitting; the KSKL for a low energy photon lost)

  6. Distribution of M(p+p-p0 ) after kin.fit-1 (10 entries per event): MC expectations for signal and background Same distribution from 2002 data sample after kin.fit-1: 15358 events (only ~3000 of them are “good” signal events)

  7. Try to describe the data distribution with Sum of: MC (signal + wp background + Ksn background). (solid) data (dashed) MC sum It works but: wp = wp x 4 Ksn = Ksn x 1.5 Why ? Accidentals and splittings not at work in old MC ? Try with new MC Conclusion: Estimated background between 51 and 105 events / 4200 candidates In the worst case < 3%

  8. (5) The data: Number of events Events / L 2001 2002 2001 2002 “good” sample 1424 2856 10.39 10.95 “peak” sample 1422 275910.4211.25 Assuming the same efficiency 2001 - 2002 10.42  0.28 vs. 11.25  0.21 Difference = 0.83  0.35 f scan results:

  9. Comparison 2001-2002 (normalized to luminosity) Spectrum 2001+2002 [4181 evts] compared to 2000 [197 evts] (normalized to luminosity and bin size) Raw spectra: only “peak” samples

  10. Is it a spectrum compatible with a resonance ? Take away the signature of the radiative decay, plotting not N(Mhp) but Mhp (MeV) Simple fit with Breit-Wigner MR = 985  1 MeV (PDG  984.7  1.2 MeV) GR = 33  1 MeV (PDG  50 100 MeV)

  11. Dalitz plot density distribution: M(hg) vs. M(hp) Expected signals from rp and wp r (w)  hg a0 region Distribution of M(hg) (5 MeV bins) : signal of wp ?

  12. (6) Data – MC comparison. (a) tracks and photon distributions (b) c2 probability distributions: Fit-1 and Fit-2

  13. (d) cosqrad distribution: comparison with (1+ cosqrad2): try fit with: A(1+x2)+B(1-x2) If dist ~ (1+ cosqrad2) B=0 data need B0  deviation from (1+ cosqrad2) Solid = MC Points = data 2002 Curve = A(1+x2)

  14. Conclusions: • (0) Some improvement to the data sample • (1) Work on new Montecarlo with: • improved statistics • realistic background •  Understand 2001-2002 discrepancy • 1% estimate of background (2) Track and photon data/MC efficiency • (3) Estimate of BR with more stable efficiency • (4) Fit as 1 year ago • Compare with 5 photons analysis

More Related