ETD

Archivio digitale delle tesi discusse presso l'Università di Pisa

Tesi etd-06232014-150706


Tipo di tesi
Tesi di laurea magistrale
Autore
MANFREDA, ALBERTO
URN
etd-06232014-150706
Titolo
Pass 8: development and science prospects for the new Fermi LAT event-level analysis
Dipartimento
FISICA
Corso di studi
FISICA
Relatori
relatore Dott. Baldini, Luca
Parole chiave
  • Pass 8
  • event reconstruction
  • cosmic ray
  • gamma ray
  • Fermi
Data inizio appello
15/07/2014
Consultabilità
Completa
Riassunto
The Large Area Telescope (LAT) onboard the Fermi Gamma-ray Space Telescope (Fermi) is a pair-conversion telescope for high-energy gamma rays operating on orbit since June 2008.
The experience and the great amount of data accumulated in this first years of mission,
have led to a deeper understanding and improved description of the LAT performance in the orbital environment. On that basis, a radical revision of the entire event-level analysis, which includes virtually every aspect of the data reduction process, is in course of development by
the Fermi collaboration. This new event analysis framework, going under the name of Pass 8,
is expected to give a significant improvement to the quality and quantity of the data collected
by the LAT.
The new event reconstruction being essentially frozen, the present work fits into the subsequent stage of the event-level analysis, namely the characterization of the topological information available on an event-by-event basis and its use for the background rejection. The connections with some relevant science topics and the corresponding prospectives will be described in details in the thesis.
First, I present the results of a multivariate analysis, made using the variables produced
by the various reconstruction algorithms, aimed at characterizing the quality of the energy
reconstruction. This is achieved by an algorithm that, based on the topology of the event in the detector, tries to estimate the probability that its energy measurement will be in the core of the energy dispersion distribution. The algorithm is created by training several Classification Trees on the simulated data. Various possible approaches to the task are examined, as well as a series of different design choices, and their results are compared to determine the one with the best performances.
The resultant estimator (PE) will become part of the standard selection algorithms developed by the Fermi collaboration as a variable onto which operate quality cuts finalized to remove the tails of the energy dispersion. Furthermore, it can be used to select event sample with a narrower energy dispersion, which can potentially enhance the sensitivity in several different science analyses. A natural application of this first stage of the work is the search of possible monochromatic lines in the gamma rays spectrum. Through a series of Monte Carlo simulations I show that, taking into account the information given by P(E), the sensitivity to a line can improve up to 15% in interesting regions of the phase space.
Furthermore I present my work on various stages of an ongoing analysis of the high-energy cosmic ray electrons measured by the LAT with the new Pass 8 event level analysis. My main contribution was the development of a custom event selection, made of both hand-made cuts and specifically trained Classification Trees, aimed at discriminating electrons from the other charged species. In the thesis this event selection will be characterized in terms of the acceptance of the detector and of the residual background after all the cuts. A preliminary electron spectrum is included, as final conclusion of this work.
File