Understanding Jet Physics at Modern Particle Colliders

Abstract

We explore the phenomenon of jets at particle colliders from several diverse vantage points, beginning with a non-technical introduction to the field in general. Jets are messy objects that are difficult to model precisely; so much so, that the best way to design a high-precision observable is to avoid them. In the second chapter of this thesis, we do just that by introducing several electroweak observables that can be measured and computed to high precision, by carefully controlling the contribution of jets. In the third chapter, we attack the problem of jets at high precision head on by computing a jet substructure observable, the mass, to unprecedented accuracy. This is made possible through a grooming technique that removes the most complicated radiation from the jet before its mass is measured. In the fourth chapter, we make qualitative progress in understanding one of the most fundamental problems in jet physics: quark/gluon tagging. To do so, we introduce a new counting observable for jet substructure, and we glean insights through its quantitative calculation. In the final chapter, we explore jets from an entirely different angle, using techniques from modern machine learning. While such methods are generally opaque to interpretation, we build a framework for users to intuitively extract physics that the machine learns.