Observations of the Moon shadow in cosmic-ray-induced muons with the IceCube Neutrino Observatory

One of the main goals of the IceCube Neutrino Observatory is to measure cosmological neutrinos and discover their sources. Neutrinos are not directly measurable, but through their resulting charged leptons in charged-current interactions with matter, i.e. electrons, muons, and tauons. For the determination of the directions, muon neutrinos are most suitable, because the resulting muons create a track signature in the detector, while electrons create electromagnetic cascades and most tauons much shorter tracks as a muons of the same energy with cascades along the track, so their directional reconstructions give large uncertainties. An important task for a detector is the verification of the pointing and the determination of the pointing uncertainty, as well as the proof that the standard methods of neutrino source analyses work as expected. Usually, these calibration tasks are done with Monte-Carlo studies. The disadvantage of these studies is that they depend on how well the simulations reflect the reality. Another possibility for these calibration tasks is the usage of a standard candle. The issue here is that one can not use a neutrino source, as their neutrino fluxes are small, and the assignment of a neutrino event to its source is possible just very rarely. However, IceCube also measures a lot of muons from particle air showers created by cosmic rays interacting with matter in the atmosphere. Cosmic rays get blocked by massive objects like the Moon and the sun, resulting in a deficit in cosmic-ray-induced muons from their directions. The Moon is well suited for this task as its position is well know and it has a negligible magnetic field. Besides the pointing verification, the observation of the Moon can also be used to compare different reconstruction algorithms, as well as for testing source analysis methods. In this thesis some source analysis methods are improved to more precise descriptions of the reality and tested using the observation of the Moon. One new development is the usage of the full information about the asymmetric Gaussian directional uncertainty estimation of events provided by the tool paraboloid instead of using its symmetric approximation as it is the current default in IceCube analyses (see section 4.3.2). In section 4.3, the background description is improved by taking the uncertainty estimation into account (similar to a Kernel Density Estimation) instead of using a histogram. An extended disc source hypothesis is developed in section 4.4.2 to describe extended sources in a correct manner. The analysis methods are verified using simulation studies. This newly developed and improved methods are compared in section 5.2.1, using the Moon as a standard candle, to the approximations which are currently used in point-source analyses in IceCube. The uncertainty estimation provided by paraboloid is often underestimated because it neglects the correlations between the size of the uncertainties and the direction and energy. To scale the uncertainty one compares the deviations of the true and reconstructed directions to the uncertainty, using Monte-Carlo simulations. In section 4.2, a method to apply this scaling to the asymmetric Gaussian is developed, the impact of directional and energy dependent scaling is shown. It is compared in section 5.2.2. Further, a Monte- Carlo simulation is done to verify the accuracy of the uncertainty estimation by paraboloid before and after the uncertainty scaling. The Moon is also a useful tool for physics applications. In section 6.1, the time evolution of the Moon shadow is shown. In section 6.2, simplified tests for the influence of Earth's magnetic field are done by performing fits of the radius of the Moon and by testing for the shift of the shadow in dependence of the azimuth angle of the Moon.