A measurement of the transverse momentum of Z-bosons and the associated W-mass uncertainty
Welcome Page
Page content:
The project is being carried out at University College London as part of the final year project of a Physics MSci.
An investigation into how quantum chromodynamics, such as gluon emission, affects the transverse momentum of Z bosons produced in pp-bar collisions.
Project Aims
The aims of the project were as follows:
- Understand the effect the gluon radiation has upon the kinematics of W and Z-boson production.
- Derive a form for the transverse momentum of Z, by fitting experimental data with that of a detector simulation, and comparing it with the latest theoretical models.
- Estimate how the uncertainty in the W mass is affected by the transverse momentum of Z
Work conducted
A simulation of the CDF's Central Electromagnetic (CEM) calorimeter was written. HERWIG events were run through the simulation to investigate how the energy resolution of the CEM affect the production of Z bosons. It was found that with increased kappa the peak of the PT(Z) increased. The affect of kappa was found to be significant and the value of kappa was derived by means of fitting mass distributions to MZ data from the CDF. Kappa was calculated at 1.5065 +/- 0.2368; the chi-squared for the fit was 123.383 with 138 degrees of freedom, corresponding to passing a 80% significance test.
A functional form was chosen to represent PT(Z):
The functional form
This was run through the detector simulation, and then compared with PT(Z) data from the CDF. The parameters of the best fit were:
- P1 = 0.1311 +/- 0.0103
- P2 = 0.0302 +/- 0.0262
- P3 = 6.8299 +/- 0.4895
- P4 = 0.6070 +/- 0.0607
The chi-squared of the it was 38.9175 with 46 degrees of freedom, passing a 76% significance test.
The best fit smeared functional form
Conclusion
A estimate for the corresponding uncertainty on the measured transverse mass of the W was not completed; there was not enough time to complete the program, and integrating the Fortran subroutines into C++ was tricky. With more experience of Fortran, the analysis should have been completed, it was the implementation that cause the delay in working.
Is believed that whether the CEM simulation is valid, or not, could only be determined with more transverse momenta data from the CDF. The chi-squared fits of the smeared and original functional form both fit the data, but the error bars on the data are large enough to allow this.
Throughout the project work was hampered by bug and other programming related issues. Given that C++ was an unfamiliar language at the start of the project, progress was slower than liked when faced with difficult programming tasks. However, the language was picked up, with analysis programs were being written from scratch and could generate histograms that could be viewed in Root immediately.
It is hoped that the final section of the project could have been completed within a fortnight if it were to be continued.
Modelling some of the other detector, such as the Plug EM calorimeter, would improve the simulation; the effect of the PEM's kappa was significant when the PEM was modelled along with the CEM. Taking other detector effects into account would also be beneficial.