A experiência AMS
- 1st time: checkout svn co --username=user_name svn://fcomp.ist.utl.pt/PROJECT - commit svn add <filename> svn ci -m "primeiro programa e teste de svn" <file> - update do projecto svn update - remove a file svn rm <filename> svn ci -m "filename removida" - help command svn help
The Sun emits a continuous stream of highly conductive plasma that permeates the entire Solar system,
transporting Solar magnetic field lines with it.
The Solar magnetic field changes the direction and energy of cosmic-rays inside the Solar system,
creating an effect known as Solar modulation.
The cosmic ray flux is especially sensitive to this effect on the low energy range, up to 30 GV.
The goal of this work is to explore this phenomenon in 1 and 2 dimensions (radial and polar) under a stochastic resolution approach.
The students are proposed to work on the stochastic resolution of diffusion-like equations in order to get acquainted with the resolution technique and to gain knowledge in the area.
After this process we want to be able to develop a 2D stochastic solution in order to master the technique and compare to the 1D finite difference method previously developed by other students.
Detailed working plan:
About 10% of all cosmic particles are helium (Z=2), they are composed mainly by He3 and He4 isotopes and the
ratio He3/He4 varies from 10 to 20%, as a function of magnetic rigidity.
By studying AMS data, understanding the different measurements involved and their significance in the selection of cosmic ray events, one can develop a multivariate analysis framework in order to identify helium nuclei and accurately separate its isotopes.
Due to the sheer amount of data and variables involved, state-of-the-art data analysis techniques are growing in popularity due to their speed and discrimination capabilities. They usually require the user to choose appropriately significant variables and to be trained using either monte-carlo events or highly pure data samples.
The students are proposed to develop reduced datasets from AMS trees (known as miniDST's) and to study the data in order to identify the key observables in cosmic ray event selection (namely helium). These key variables would be used to train a neural network selection framework (based on the ROOT TMVA). This selection platform developed by the students would then serve to identify helium and its isotopes from AMS data. As an optional end-goal, students would try to estimate the time-variable helium flux from their selected events.
Detailed working plan: