Improved Event-mixing for Resonance Measurements
Author
Summary, in English
This report first analyzed the reason for the deviation of the event-mixing method in background estimation for resonance research, and one possible explanation is that the event-mixing method can not reproduce the specific angular distribution of produced particles caused by jet-effects. This assumption was checked by Monte-Carlo simulations with PYTHIA, and a certain method of correction called reweighing was proposed to improve the event-mixing method. The improved event-mixing method was then applied for simulation as well as resonance analysis for experimental data from the ALICE experiment, and the results proved that this reweighing method can improve the performance on resonance signal extraction by reducing the residual background.
Summary, in English
The journey of understanding our world of matter, why it exists, how it works, was once dominated by a series of searching for new particles. Before the Standard Model was set up, our knowledge was limited at the level of protons, neutrons, electrons, and perhaps muons and pions from cosmic rays in some occasions. We already had the evidence of new particles’ existence at that time, although the mechanism of their formations and their interactions were still in deep fog. We could learn some properties from their decays into more stable forms that we already knew, or in analogy to the oscillators, we regarded the decay products as they had been through a specific interaction — a resonance. With the analysis to the resonance processes of these relatively stable particles, many new particle forms of existence were discovered, from the lightest leptonic resonance for photons, to the heaviest top quark resonance. Even with the most delicate detector in modern science research, we still rely on the resonance researches for particles studies, and our efforts for improving this way of research never stops.
One crucial step in resonance research is to separate background reading, which may vary in many aspects depending on our method of resonance reconstruction, from experimental data. The backgrounds usually contributed from background tracks(particles) and combinatoric of uncorrelated track pairs(here implies direct correlation originates in the same decay). While the background tracks can be rejected sometimes with sufficient experimental data and processing power of hardware, the combinatorial background is the major section in both scale and in difficulty especially for heavy-ion collision experiments.
The reweighed event-mixing method introduced in this thesis is an improved way of estimation for combinatorial backgrounds. In general, the normal event-mixing combines arbitrary tracks in different events which estimates an ‘isotropic’ background. This mismatch the real situation in resonance since the final states of tracks are spatially concentrated in cones due to their evolving mechanism. The improved method mentioned before is an artificially reproduction of their angular preference, and tested on experimental data from the ALICE group. The test result proves a significant performance boost in background estimation.
One crucial step in resonance research is to separate background reading, which may vary in many aspects depending on our method of resonance reconstruction, from experimental data. The backgrounds usually contributed from background tracks(particles) and combinatoric of uncorrelated track pairs(here implies direct correlation originates in the same decay). While the background tracks can be rejected sometimes with sufficient experimental data and processing power of hardware, the combinatorial background is the major section in both scale and in difficulty especially for heavy-ion collision experiments.
The reweighed event-mixing method introduced in this thesis is an improved way of estimation for combinatorial backgrounds. In general, the normal event-mixing combines arbitrary tracks in different events which estimates an ‘isotropic’ background. This mismatch the real situation in resonance since the final states of tracks are spatially concentrated in cones due to their evolving mechanism. The improved method mentioned before is an artificially reproduction of their angular preference, and tested on experimental data from the ALICE group. The test result proves a significant performance boost in background estimation.
Publishing year
2016
Language
English
Full text
Document type
Student publication for Master's degree (two years)
Topic
- Physics and Astronomy
Keywords
- Event mixing
- Resonance
- phi meson
Supervisor
- Peter Christiansen (Associate Professor)