Jets and their substructure at the ATLAS detector : new tools for new physics

• Main Author: Ennis, Joseph Stanford
• Published: University of Warwick 2017
• Subjects: 530; QC Physics
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.742183

This thesis is primarily concerned with two topics: the trigger performance and calibration of large-radius (large-R) jets at the ATLAS detector, and the presentation of a search for new high-mass particles that decay to vector bosons in the fully hadronic decay channel. The performance of the ATLAS large-R jet trigger during 2015 was evaluated and found to be 100% efficient by 380 GeV for all large-R jets. The impact of jet substructure on this trigger was studied, and we discovered that a newly proposed set of variables, anti-subjettiness variables, could reduce trigger rates by approximately 30%, while still accepting 95% of large-R jets from heavy particles. Alongside this, a new calibration method for large-R jets is presented. A neural-network is used to perform a multivariate calibration for jet energy and mass using substructure variables. While there were no gains observed in the jet energy calibration, a 26% improvement in the jet mass resolution was found. A search for X →V V→ qqqq, using large-R jets tagged as boosted vector bosons, was conducted using 37.6 fb -1 of data collected at √s =13 TeV by ATLAS in 2015 and 2016. No significant deviation from the Standard Model was observed and limits were placed on a variety of generic signal models. The Heavy Vector Triplet model A (B) was excluded in the mass range 1:2(1:2) < mV < 3:1(3:5) TeV, the RS graviton with k/M planck = 1 was excluded for 1:3 < mG < 1:6 TeV, and limits on σ x BR were placed on a generic scalar model.