Abstract
Aims. This paper proposes a new approach to detecting 𝒪(MeV) neutrino bursts such as those associated with supernovae. Methods. A novel ‘real-time test statistic’ (RTS) exploits the temporal structure of the expected signal, discriminating against the diffuse background, to allow detection of very weak signals that would elude standard clustering methods. Results. For a given background rate, the proposed method increases signal efficiency while keeping the same false alarm rate for a Poisson-distributed background. By adding a spatial penalty term to the definition of RTS, it is also possible to reject spatially correlated backgrounds such as those due to spallation events. Conclusions. The algorithm can be implemented in a real-time monitoring system for detectors of all sizes, allowing prompt alerts to be sent to the wider community, for example through the SNEWS 2.0 network.
Highlights
The detection of low-energy (Eν ∼ O(MeV)) neutrino bursts is relevant in the search for supernovae
We present the RTS2 method and apply it to the detection of neutrino bursts with a generic time spectrum that fits well with current supernova models
The RTS2 method allows for high flexibility as to the choice of the function characterising the contribution of each event and it may be further improved by considering the addition of new variables in the analysis, such as the event energy
Summary
The detection of low-energy (Eν ∼ O(MeV)) neutrino bursts is relevant in the search for supernovae. The usual method consists in watching for positive fluctuations from the typical expected background. In most of the neutrino detectors, the main background at energies relevant for supernovae can be modelled with Poisson statistics, which is a constant rate with simple statistical fluctuations. For a given background rate r, the number of background events in a time window w follows the Poisson distribution with λ = r × w: Poisson(m; r, w) = e−rw (rw)m · (1) m!. Let us denote the typical timescale of the physical phenomenon under study, τphys. If the detector has a relatively low background (rτphys 1), this method cannot be applied and the usual technique consists in dividing the data into clusters using time windows of size w, covering the physical timescale τphys. There are two different approaches to building these time windows, as illustrated in Fig. 1: sliding windows of width w, with each window starting in the middle of the previous one (Agafonova et al 2008); and dynamic windows of width w, one starting from each selected event (Ikeda et al 2007; Abe et al 2016; Novoseltsev et al 2020)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
