High-precision measurement of the W boson mass with the CMS experiment

The measurement of the W boson mass (m_W) is a crucial test of the Standard Model of Particle Physics. This thesis presents the m_W measurement by CMS, announced in September 2024. The result agrees with both the Standard Model and the world average of direct measurements, which has an uncertainty of 13.3 MeV. CMS achieved an uncertainty of 9.9 MeV, close to the Standard Model indirect prediction precision (~10^−4). This matches the precision of the most precise single-experiment result from CDF, which, however, disagrees strongly with the Standard Model. The CMS measurement is the culmination of over a decade of work, gained special attention after the CDF result. CMS uses an innovative approach, fitting the two-dimensional p_T − η distribution of muons, simultaneously constraining theoretical uncertainties and m_W without relying on external Z boson data to determine the W production mechanism. This uses state-of-the-art theoretical predictions and nuisance parameter modeling. This method required the largest dataset ever used in an m_W analysis—1.2 × 10^8 events and 4 × 10^9 simulated events—necessitating a major software overhaul. Experimentally, the analysis depends on very precise muon reconstruction and identification. Achieving better than 10^−4 precision on muon transverse momentum required developing new track refitting software, reducing biases, and applying detailed physics-driven corrections validated for accuracy and robustness. Muon reconstruction and identification efficiencies, as functions of p_T and η, were measured at unprecedented granularity to avoid bias in m_W extraction. Biases in the standard CMS muon reconstruction were identified and corrected. Efficiency corrections, derived via a tag-and-probe method, were thoroughly validated to address differences in Z and W boson production and quality criteria recalculations. My main contribution, documented here, involved this aspect of the analysis. Uncertainties on muon reconstruction and identification contribute about 3.0 MeV to the overall m_W uncertainty. This research was supported by the European Research Council under the Horizon 2020 program (Grant 101001205). Beyond the m_W measurement, during my Ph.D. I contributed to the CMS Tracking Physics Object Group by studying tracking performance using Run 3 data. This work supports tuning reconstruction algorithms, debugging detector issues, commissioning software updates, and ensuring data quality for prompt problem identification.