Improved W boson mass measurement with the D0 detector

TL;DR

This study broadens the DØ W boson mass measurement by including central calorimeter edge electrons, increasing the W event sample and enabling refined energy-scale calibrations. Using a template-fitting approach with a dedicated edge-electron energy-response model calibrated via Z boson data, the authors extract MW from multiple observables and carefully account for systematic uncertainties and correlations. The edge-electron analysis yields MW = 80.483 ± 0.084 GeV when combined with previous DØ measurements, contributing to a more precise hadron-collider determination and reinforcing the consistency between direct measurements and the SM. When integrated with results from CDF and LEP, the world average MW remains at the forefront of precision tests of electroweak theory and Higgs sector constraints.

Abstract

We have measured the W boson mass using the D0 detector and a data sample of 82 pb^-1 from the Tevatron collider. This measurement used W -> e nu decays, where the electron is close to a boundary of a central electromagnetic calorimeter module. Such 'edge' electrons have not been used in any previous D0 analysis, and represent a 14% increase in the W boson sample size. For these electrons, new response and resolution parameters are determined, and revised backgrounds and underlying event energy flow measurements are made. When the current measurement is combined with previous D0 W boson mass measurements, we obtain M_W = 80.483 +/- 0.084 GeV. The 8% improvement from the previous D0 measurement is primarily due to the improved determination of the response parameters for non-edge electrons using the sample of Z bosons with non-edge and edge electrons.

Figures (18)

• Figure 1: End view of the central calorimeter showing the arrangement for electromagnetic (EM), fine hadronic (FH) and coarse hadronic (CH) modules. The Tevatron Main Ring passes through the circular hole near the top of the CH ring.
• Figure 2: Construction of central calorimeter EM modules in the region near module boundaries. Signal boards have the electrode pads for signal collection; readout boards carry traces bringing the signals to the module ends.
• Figure 3: Comparison of transverse mass distributions for background events to $W$ bosons for C (points with error bars) and $\widetilde{\rm C}$ (solid histogram). The two distributions are normalized to the same number of events.
• Figure 4: Distributions for $\widetilde{\rm C}$ samples as a function of the ratio of the electron impact distance $d_{\rm edge}$ from the module edge to the total module width, $d_{\rm mod}$: (a) the fitted scale factor $\alpha$, and (b) the fitted $W$ boson mass using the appropriate scale factor for each $d_{\rm edge}$ bin. The errors are statistical only.
• Figure 5: (a) Dielectron mass distributions for CC and $\widetilde{\rm C}$C samples, with the CC distribution scaled to give the same peak value as for the $\widetilde{\rm C}$C distribution. The solid histogram is for the CC $Z$ bosons and the points are for the $\widetilde{\rm C}$C $Z$ bosons. (b) The difference between $\widetilde{\rm C}$C and normalized CC samples. The curve is a Gaussian fit; no backgrounds are included in the fit to the difference.