Muon density in extensive air showers measured with IceTop

Cosmic Rays are particles that are accelerated from astrophysical sources and being propagated through the universe. When Cosmic Rays strike particles of the Earth's atmosphere, they undergo interactions which trigger a cascade of particles travelling to the Earth's surface. This air shower carries constituents of different types, such as hadrons, electrons, and muons. Measurements of these constituents can be compared with air shower simulations, in order to draw conclusions on the properties of the initial Cosmic Ray particle, such as nuclear mass, energy, and direction. In particular, the number of muons shows a dependance on the nuclear mass. In this work, the detector signatures provided by the IceTop detector are analyzed. IceTop is the surface part of the IceCube detector situated at the South Pole. It consists of 81 stations ordered in a grid-like structure. Each station comprises two tanks filled with clear ice. Charged particles passing through ice produce Cherenkov light which is detected by photomultiplier tubes. In this analysis a method is provided which allows the determination of muon number densities based on measured detector signatures. Initially, this method is developed on air showers simulated with the hadronic interaction model SIBYLL2.1, and having zenith angles between 0° and 36.9°. These air showers are subject to detector simulation and reconstruction, and undergo a selection of quality cuts. The selected set of simulated air showers is separated into bins in zenith angle and estimated energy. For each remaining air shower, the detector signatures are subclassified into signal and background, depending whether they contain muons or not. The signal content of air showers is carved out by introducing energy and zenith angle dependent cuts on the tank charge and its distance to the shower axis. After the application of these cuts, two muon number estimators are defined basing on lateral charge and hit distributions. Systematic studies show that the latter is more robust against changes in simulation. Using a conversion, the muon number estimator is transformed into a muon number density estimator, which is related to the true muon density known from air shower simulations. Finally, the muon number density in experimental data, recorded during the 2012/2013 season, is calculated by applying the conversion derived from simulations, on muon number density estimators derived from data. The result is the muon number density as function of reconstructed primary energy. Since the conversion is derived from air shower simulations, systematic uncertainties such as the hadronic interaction model or the primary mass propagate onto the final result.