Using 1.9 fb^-1 of CDF data collected between 2002-2007, we have measured the mass of the top quark in the all hadronic decay channel of tt pairs. This channel is characterized by a 6 jet final state topology with little missing energy in the event. The reconstruction of the top quark mass in this channel is challenging for two main reasons: the large jet ambiguities to reconstruct two top quarks out of 6 jets and a large amount of QCD multi-jet background of which the cross sections are badly known.

The Ideogram method tackles this problem by combining both template and matrix element approaches to extract the maximum amount of mass information out of a tt candidate event. This is realized by construction a two-dimensional event-by-event likelihood in which the signal purity (or in other words, signal cross section) is left as a free parameter, in addition to the top quark mass itself. The event likelihood takes into account all possible independent jet combinations where the tagged jets are assigned as b jets. Its signal part is purely data driven and consists of the two dimensional convolution of natural Breit-Wigners with Gaussian resolution function, using fitted top and W masses and their variances obtained from a kinematical fit. The QCD multi-jet background and poorly reconstructed signal events are parametrized by a 2D mass template. Each of thejet configurations is weighted by it's goodness of fit.

The measured top quark mass in the all hadronic channel amounts to M_top = 165.2 ± 4.4 (stat+JES ) ± 1.9 (syst ) GeV/c². The statistical uncertainty is expected to scale with 1/sqrt(int lum.). The dominant source for the systematic uncertainties are the radiation, the Monte Carlo generator and the residual jet energy scale.

Top quark mass was measured in the all-hadronic channel by using a template method:
www-cdf.fnal.gov/physics/new/top/2007/mass/allhad_TMT_2D/bless_dec07.html
The template analysis uses different event selection and therefore the data sample does not completely overlap with the data sample used in Ideogram analysis. The consistency of these two analyses was evaluated by estimating the correlation between the analyses (between 0.42 and 0.67), giving an estimate that the two analyses are two to three standard deviations apart from each other.

Measurement using data sample corresponding to 310 pb ^-1:

Public note:

Blessed Result (1.9 fb^-1)

M_top = 165.2 ± 4.4 (stat+JES) ± 1.9 (syst) GeV/c²

Blessed Tables and Plots

Click on the thumbnail to get the full-size file.

Description of Plots	gif	eps
The two dimensional likelihood based on the measurement by using the 1.9 fb-1 dataset. The cross indicates the maximum likelihood point and countours show different confidence levels.		eps
The expected statistical uncertainties and the measured uncertainties from data.		eps
The top quark mass spectrum obtained by using the signal likelihood for the background, signal (Mt = 165 GeV/c2) and data events.		eps
The top quark spectra for the background, signal (Mt = 165 GeV/c2, JES=1.0 &sigmaJES) and data events, obtained by weighting the combinations with their respective compatibility to be a correct combination.		eps
The W boson mass spectra for the background, signal (Mt = 165 GeV/c2, JES=1.0 &sigmaJES) and data events, obtained by weighting the combinations with their respective compatibility to be a correct combination.		eps
The input and measured top quark mass and the JES values in a two dimensional plot.		eps
The width of pull distribution for different top masses.		eps
The top mass calibration constant as a function of the input jet energy scale.		eps
The top mass calibration slope as a function of the input jet energy scale.		eps
The jet energy scale calibration constant as a function of the generator level top quark mass.		eps
The jet energy scale calibration slope as a function of the generator level top quark mass.		eps
The two dimensional fitted parameterization of the matched events for JES=0 &sigmaJES and Mt = 175 GeV/c2 to illustrate difference between the non-matched and background events.		eps
The two dimensional fitted parameterization of the non-matched events for JES=0 &sigmaJES and Mt = 175 GeV/c2.		eps
The two dimensional mass distribution of the non-matched events (histogram) and the fitted parameterization (solid line) for JES=0 &sigmaJES and Mt = 175 GeV/c2 projected into mt plane.		eps
The two dimensional mass distribution of the non-matched events (histogram) and the fitted parameterization (solid line) for JES=0 &sigmaJES and Mt = 175 GeV/c2 projected into mW plane.		eps
The two dimensional fitted parameterization of the background events.		eps
The two dimensional mass distribution of the background events (histogram) and the fitted parameterization (solid line) projected into mt plane.		eps
The two dimensional mass distribution of the background events (histogram) and the fitted parameterization (solid line) projected into mW plane.		eps
The top mass obtained from kinematic fit.		eps
The W boson mass obtained from kinematic fit.		eps
The uncertainty on the top mass obtained from kinematic fit.		eps
The uncertainty on the W boson mass obtained from kinematic fit.		eps
The &chi2 obtained from the kinematic fit.		eps
The correlation coefficient between top quark and W boson mass obtained from kinematic fit.		eps
Table of systematic uncertainties.		eps

Last modified: Mar 19 2008