Test background estimates for resonance measurements using a combination of mixed events and like sign pairs

Abstract

Two approaches for extracting properties of neutral resonances out of invariant mass distributions are presented: like-sign and event-mixing signal. Additionally a correction function based on event-mixing, aimed at revising detection differences of opposite charged particles, which is applied to the like-sign signal is tested. The analysis is conducted using pp collisions at 7 TeV and concludes an improved description of the residual background by the correction function. Furthermore no change of the resonance's extracted parameters is observed. The event-mixing method exhibits a problematic description of the background which is concluded in this thesis to be caused by the necessity of an azimuthal angle rotation for event-mixing.

Popular Abstract

The question whether matter can be divided infinite times has occupied scientists for more than two hundred years now. The aim to describe matter, their elementary constituents and interactions is the definition of particle physics. In order to examine these constituents matter is smashed together at large energy and the fragments analyzed. This is done, among other places, at the CERN organization in Geneva, Switzerland, using the Large Hadron Collider. Some of these basic constituents, called quarks, are only observed as hadrons: particles that consist of two or three quarks (at least to prevalent awareness). In experiments such as ALICE at CERN the particles created are measured using elaborate detectors. There exist hadrons which have a very short lifetime. They are produced in the collision and decay already before reaching the first detectors. They are called resonances and their study is naturally difficult.

In order to still be able to detect these particles one looks at their decay products. The particles might though be rare and in general millions of particles are arriving at the detectors. A link between the decay products needs to be established in order to examine the mother particle. Confronted with the huge amount of data this can not be done by hand. Statistical methods are used to find the link, called correlation, between the decay product particles.

The name “resonance” comes from the signal it produces: as in the figure, a bump indicates that a particle is found with a certain mass (depending on the position of the bump). But how high is the bump, how wide is it? These questions are for particle physicists essential when they try to determine the properties of a resonance.

Here the description of the background plays a major role. In this thesis a new approach has been taken to describe this background: to the previous description a correction has been applied. And results show that this is in fact an improvement. Although not very large it can help to open the window into the subatomic world a bit wider and enrich our knowledge of particles. At the same time, as well done in this thesis, comparisons of established methods are conducted and results indicate that sometimes simpler methods can be highly effective. So the work of physicists continues explaining this world, step by step. And sometimes relying on simple methods can blend out almost all the noise, isn't that something?

• author: Thiele, Fabian Alexander Jürgen
• supervisor: Peter Christiansen
• organization: Particle and nuclear physics; Department of Physics
• course: FYSM31 20142
• year: 2015
• type: H2 - Master's Degree (Two Years)
• subject: Physics and Astronomy
• language: English
• id: 5050853
• date added to LUP: 2015-02-20 07:30:03
• date last changed: 2015-02-20 07:30:03