HitSpooling: an improvement for the supernova neutrino detection system in icecube

which monitor one cubic kilometer of deep Antarctic ice at the geographic South Pole.

IceCube was primarily designed to detect neutrinos of energies greater than O(100 GeV).

Due to subfreezing ice temperatures, the photomultipliers' dark noise rates are particularly

low which enables IceCube to search for neutrinos from galactic supernovae by detecting

bursts of MeV neutrinos emitted during the core collapse and for several seconds following.

For that purpose, a dedicated online supernova DAQ system records the total number of hits

in the detector, without any further information from the PMTs, and generates supernova

candidate triggers in case of a significant detector rate enhancement. A new feature to the

standard DAQ, called HitSpooling, was implemented in IceCube during this thesis. The

HitSpooling system is implemented in the standard DAQ system and buffers the complete

raw data stream of the photomultipliers for several hours or days. By reading out time periods

of HitSpool data around supernova candidate triggers, generated by the online supernova

DAQ system, we overcome the limitations of the latter and have access to the entire information

of the detector in case of a supernova. Furthermore, HitSpool data is a powerful

source for studying and understanding the noise behavior of the detector as well as background

processes coming from atmospheric muons. The idea of HitSpooling was developed in the

scope of this thesis and is the basis of the work at hand. The developed interface between the

standard DAQ and the supernova DAQ system is presented. The correlated dark noise component

in optical modules of IceCube is quantified for the first time and possible explanations

are discussed. The possibility of identifying triggering and subthreshold atmospheric muons

in HitSpool data and subtracting them from a possible supernova signal is analyzed. Furthermore,

the conversion from HitSpool data to supernova DAQ type data was developed

which allows for a comparison of both data types with respect to lightcurves and significances

of selected supernova candidate triggers.