Development Of The Daya Bay Neutrino Offline Software And Studies On The Neutron Background

Neutrino flavor mixing and oscillation is the only evidence of physics beyond the standard model in particle physics supported by various experimental results and con-sidered as the key to search for new physics. After the development in more than half a century, neutrino experiments has come to a era of precise measurements of the oscilla-tion parameters from the discovery of neutrino oscillation phenomena. Among the six parameters of the neutrino mixing matrix, the value of θ13is important to the future neu-trino experiments such as the measurement the mixing phase angle δ or neutrino mass hierarchy. The physics goal of Day a Bay reactor neutrino experiment is to measure sin22θ13to a precision of0.01or better. To reach this physics goal, there are totally8identical detector modules with systematic error less than0.38%and target mass lager than20tons at three experiment sites in Daya Bay reactor neutrino experiment. The energy resolution of each module is better than15%/(?).As the first reactor neutrino experiments in China, in order to adapt to the proper-ties of the physics data, the Daya Bay online data acquisition and offline data analysis software system are implemented by new technology, such as design patterns in object-oriented software, advanced Gaudi framework for particle physics experiments, and the python language in configuration scripts, which brought new client experiences in cal-1of software modules, modular design and development, data quality monitoring and maintenance of software system.As the first step of the off-line data processing, the raw data access is the basis of the entire offline physics data analysis. We have developed raw data conversion service in Gaudi, which hid the trivial details about the raw data format to clients and provided uniform interfaces for data accessing. This service was mainly used to decode the binary raw data from the DAQ system, restored the compressed information into time and charge for the hits in sensors. And these information are stored as the standard data object in offline system for subsequent analysis. Using the raw data conversion service module, we analyzed the data from the anti-neutrino detector commissioning, calibrated the pedestal and threshold of each channel in FEEs for the AD modules.Lots of events are time correlated in the data of reactor neutrino experiment. Anal-ysis of time correlation between different events is an important task in physics data analysis, which is different from data analysis in collision physics experiments. To get the MC samples suitable for physics analysis, time correlations from different MC sam- pies should be reconstructed according to the time distributions of the physics events. With regards to this, we have upgraded the standard data I/O module in Gaudi frame-work by adding interfaces for the input of multi-format, multi-stream data in parallel and redesigned new data cycling mechanism based on event backtracking. We finally developed the "Interleaving" algorithm with high performance for MC events mixing. To enhance the efficiency and reliability of the MC simulation, we improved the archi-tecture of the digitization simulation by reorganizing the hits from sensors and changing the transmission way for events between different modules. Based on the above works, we mixed different MC samples at both SimHit level and Readout level, and generat-ed-102TB MC data for physics analysis. Based on the data, the preliminary events selection and background estimation were finished.All possible backgrounds should be investigated carefully in Daya Bay reactor neutrino experiment for such a low background, low energy, high precision experimen-t. Among all the backgrounds, series related backgrounds caused by cosmic ray are particularly important. In this thesis, we have used linear extrapolation method to esti-mate the fast neutron background from cosmic μ based on three month data. We have checked the energy spectrum of fast neutron events to confirm the linear extrapolation method. We also have calculated the spallation neutron yielded from cosmic ray μ, as the function of μ energy at the three different experimental site, and compared the results with that of other underground experiments and the expectation of phenomenological models.