Blessed plots from Dijet Angular Distribution analysis

This is a first preliminary measurement of the dijet angular distribution in Run 1A+1B data. The mass distribution, for |eta|<2, and |cosine theta*|<2/3 is shown for reference in fig. 1 and 2: it has an excess at high mass when compared to QCD from PYTHIA plus a CDF detector simulation. The angular variables we use are motivated in fig. 3. The measured angular distribution in fig. 4 is consistent with NLO QCD, in all mass regions. The angular distribution is plotted using the variable chi=|exp(eta1-eta2)|=(1+|cosine theta*|)/(1-|cosine theta*|) in 5 bins of dijet mass: 241< M <300, 300< M< 400, 400< M <517, 517< M <625 and M >625 GeV. We also compare the angular distributions to LO QCD in Fig. 5 and 6. For each mass bin the shape of the angular distribution can be summarized by a single sensitive variable: R_chi the ratio of the number of events with chi<2.5 to the number of events with 2.5< chi <5. The angular ratio is in good agreement with NLO QCD, as shown in fig. 7. The inner error bars are statistical, the outer error bars are statistical and systematic uncertainties added in quadrature. The uncertainties in theory for R_chi from QCD are explored in Fig. 8, the largest uncertainty being from the choice of renormalization scale mu: the angular distribution is insensitive to choice of parton distribution function, cone size, or source of LO calculation. A cross check on the experimental systematics is performed by comparing the angular distribution in the central region alone (jet 0.1< |eta_d| <1.0) to the default case (jet |eta|<2): Fig 9 shows that the chi distributions are identical and Fig 10 shows that R_chi is the same within experimental uncertainties. Fig. 11 discusses that if quarks are composite particles then the dijet angular distribution can be affected. In fig. 12 we compare our measurement of R_chi with the predictions of a left-handed contact interaction in which only u and d quarks are composite for different values of the compositeness scale lambda. This is the kind of contact interaction that was compared to the inclusive jet cross section, where the best fit was around 1.6 TeV. In contrast, this value of the compositeness scale is not favored by the angular distribution. In fig. 13 we compare with a flavor symmetric contact interaction, in which all quarks are composite. Limits on the compositeness scale are in progress.

• fig. 1: dijet_mass_pythia_log.ps--- Mass Distribution and Pythia+QFL (log scale)
• fig. 2: dijet_mass_pythia_lin.ps--- Mass Distribution and Pythia+QFL (linear scale)
• fig. 3: dijet_angle_intro.ps--- Intro to variables (print it if you can't view it)
• fig. 4: dijet_angle_nlo_qcd.ps--- Angular Distribution compared to NLO QCD
• fig. 5: dijet_angle_lo_qcd.ps--- Angular Distribution compared to LO QCD
• fig. 6: dijet_angle_lo_nlo_qcd.ps--- Angular Distribution compared to LO & NLO QCD
• fig. 7: dijet_angle_ratio_qcd.ps--- Angular Ratio (chi<2.5/chi<5) compared to QCD
• fig. 8: dijet_angle_theory.ps--- Theoretical Uncertainties in Angular Ratio.
• fig. 9: dijet_angle_check.ps--- Cross Check on Angular Distribution using Central
• fig. 10: dijet_angle_ratio_check.ps--- Cross Check on Angular Ratio using Central
• fig.11: dijet_composite_intro.ps--- Introduction to Quark Compositeness
• fig. 12: dijet_angle_ratio_comp_ud.ps--- Angular Ratio and Compositeness (only u and d composite)
• fig. 13: dijet_angle_ratio_comp.ps--- Angular Ratio and Compositeness (flavor symmetric)

Last updated April 30, 1996