June 14Th-17Th, 2019 Sudbury, Canada
Multi-messenger Triggers with the Jiangmen Underground Neutrino Observatory DONGLIAN XU1,2, ZIPING YE1, and FEIYANG ZHANG2 for the JUNO Collaboration 1Tsung-Dao Lee Institute (TDLI), Shanghai, 200240, China 2School of Physics and Astronomy, Shanghai Jiao Tong University, 200240, China Abstract
A new era of multi-messenger astronomy has arrived with the detection of gravitational waves and high-energy astrophysical neutrinos. The successful coordination of near real- time follow-up campaigns by multi-wavelength and multi-messenger instruments of those events have largely extended our understanding of the most violent phenomena in the Universe. The Jiangmen Underground Neutrino Observatory (JUNO) is a 20-kiloton liquid scintillator neutrino detector under construction, it will be the world’s largest of its kind when it turns on in 2021. JUNO will have highly competitive sensitivity to MeV-scale neutrino detection, and will be able to contribute to the nascent field of multi- messenger astronomy, especially for the transient events where high radioactivity background can be easily bypassed. We will present a multi-messenger triggering and filtering system that aims to read out physics hits as much and as low-threshold as possible in JUNO, which can provide the widest broadband neutrino bursts real-time monitoring and possibly steady signal searches at the sub-MeV (as low as 20 keV) to sub-GeV energies.
Neutrino & Multi-messenger Astronomy JUNO Detector
• Core-collapse supernovae: burst & pre-burst [1,2] • 20 kiloton LS detector • (see also Huiling Li’s talk on 2019.6.15 on supernovae neutrinos in JUNO) • 34 m diameter acrylic sphere Inverse beta decay (IBD) • 700 m underground • • Neutron star mergers : ~one per million years in the Galaxy [3] 18,000 20’’ +25,000 3’’ PMTs • 53 km to the nuclear reactors • Elastic v-e scattering • Gamma ray bursts : cosmological origin, causes could be core- collapse of massive stars, merger of binary compact objects, …
• Fast radio bursts : evaporating black holes MeV neutrinos [4] • 3% energy resolution @ 1 MeV • Elastic v-p scattering • Light yield ~ 1200 PE / MeV Fig. 2 : Schematic illustration JUNO: central • Mass hierarchy: detector is an acrylic sphere filled with liquid @ (3~4) by 2026 scintillator. Outer water pool provides shielding, and tags cosmic ray muons. A muon tracker is placed on top of the water pool. Data Fig. 1.: Left to right: supernova; neutron star merger; gamma ray burst; magnetar. taking begins in 2021 [5].
Low-threshold Trigger System Fast Online Filtering and Efficiency
• Lower detector energy threshold, extend the energy window as broadband as • Basic idea: dark noise hits are uniformly distributed, while physics-like hits will possible for transient phenomena searches cluster in time & space, as illustrated in Fig.5 • Fast early alert on detection of clustering of neutrinos • !" and !# are calculated using global hit info collected on VFU of the MM Trigger system (see Fig. 3), filtering decision is made by the onboard FPGA (see Fig. 6) Fig. 3 : Schematic design of • Data volume (driven by 14C) will be ~ 4 PB/year. the trigger systems in JUNO. 5_6278 1 The global trigger based on 200*+ !# = 1 :⃗ − :⃗ !" = *_ℎ./+ 2;4 2 simple multiplicity is set at a *_ℎ./+ 234 higher threshold to avoid the 14C radioactivity which has an end point of 156 keV. The multi-messenger (MM) trigger will lower the readout energy threshold into the O(10) keV region.
0.024 a.u. 0.022 0.02 Fig. 5. Temporal (left) and spatial (right) distribution of hits for physics events and dark noise. GCU: Global Control Unit 0.018 BEC : BackEnd Card 0.016 RMU: Reorganize and Multiplex Unit 0.014 0.012
VFU: Vertex Forming Unit 0.01 FPGA: Field Programmable Gate Array 0.008 DAQ: Data Acquisition 0.006 0.004
0.002
0 0 20 40 60 80 100 120 140 160 180 200 energy [keV] Fig. 4 : Beta spectrum of 14C decay. • 20-inch PMT dark noise rate ~ 50 kHz per PMT • 14C radioactivity [6]: nominally < 400 kHz for 20 kiloton liquid scintillator 14 • Data volume for hypothetical trigger-less readout (driven by dark noise): Fig. 6: Distribution of physics events ( C) Fig. 7: The efficiency of trigger on physics !" vs !#. events as a function of nhits. In JUNO, a Save waveforms for all dark noise hits ~ 6100 PB/year and dark noise in Black dots are (1) dark noise, red dots are physics events. keV energy deposition produces ~1 hit. (2) Save T-Q pair for all dark noise hits ~ 250 PB/year The blue dash line cut rejects 99.84% dark This indicates that the trigger efficiency 14 (3) Save T-Q pair for physics-like hits ( C driven) ~ 4.3 PB/year noise, while keeping 67.5% physics events. above ~ 50 keV is close to 100%. Further improvement on the algorithms are under way, which includes adding Need to filter > 99% dark noise, while keeping physics events with high more observables and exploring dedicated machine learning techniques. efficiency, based on the hit pattern differences between dark noise and physics. Conclusion and Outlook References
Neutrino detection plays an important role in the new era of multi-messenger astronomy. [1] A. Mirizzi et al, Rivista del Nuovo Cimento Vol. 39, N. 1-2 (2016) JUNO, a 20 kiloton liquid scintillator detector, has great capability to observe astrophysical [2] J.S. Lu, J. Cao, Y.F. Li, S. Zhou, JCAP (2015) 2015. neutrinos from sub-MeV to sub-GeV energies. A preliminary multi-messenger trigger system [3] K. Kyutoku and K. Kashiyama, Phys. Rev. D 97, 103001 (2018) has been designed to lower the energy threshold of the detector down to O(10) keV. With this [4] F. Halzen, B. Keszthelyi and E. Zas, Phys. Rev. D52 (1995) 3239-3247 system, we are looking to do real-time monitoring on neutrino clustering and issue alerts to [5] JUNO collaboration, J of Phys. (2016) 43.3. multi-messenger networks, as well as do fast follow-up analyses in responses to other [6] A. Moljk and S. C. Curran, Phys. Rev. 96, 395 messengers such as gravitational waves and high energy neutrinos.
June 14th-17th, 2019 [email protected] [email protected] [email protected] Sudbury, Canada