The Fast Stochastic Matching Pursuit for Neutrino and Dark Matter Experiments

Yuyi Wang,Aiqiang Zhang,Yiyang Wu,Benda Xu,Jiajie Chen,Zhe Wang,Shaomin Chen

2024-03-06

Abstract:Photomultiplier tubes (PMT) are widely deployed at neutrino and dark matter experiments for photon counting. When multiple photons hit a PMT consecutively, their photo-electron (PE) pulses pile up to hinder the precise measurements of the count and timings. We introduce Fast Stochastic Matching Pursuit (FSMP) to analyze the PMT signal waveforms into individual PEs with the strategy of reversible-jump Markov-chain Monte Carlo. We demonstrate that FSMP improves the energy and time resolution of PMT-based experiments, gains acceleration on GPUs and is extensible to microchannel-plate (MCP) PMTs with jumbo-charge outputs. In the condition of our laboratory characterization of 8-inch MCP-PMTs, FSMP improves the energy resolution by up to 12% from the long-serving method of waveform integration.

High Energy Physics - Experiment,Data Analysis, Statistics and Probability,Instrumentation and Detectors

What problem does this paper attempt to address?

### Problems Addressed by the Paper

The paper aims to address the issue of photoelectron (PE) signal analysis in photon counting experiments, particularly in neutrino and dark matter experiments. When multiple photons consecutively hit a photomultiplier tube (PMT), their photoelectron pulses overlap, which hinders accurate counting and timing measurements. Specifically, the paper introduces a method called Fast Stochastic Matching Pursuit (FSMP) to analyze PMT signal waveforms, decompose them into individual photoelectron pulses, and improve energy and time resolution.

### Main Contributions

1. **Improved Energy and Time Resolution**: The FSMP method represents the uncertainty inferred from the waveform through a Bayesian posterior distribution, thereby improving the energy and time resolution in PMT experiments.
2. **GPU Acceleration**: The FSMP method is accelerated on GPUs, enabling efficient processing of large amounts of waveform data.
3. **Applicable to New MCP-PMT**: The FSMP method is not only applicable to traditional PMTs but also extends to microchannel plate (MCP) PMTs with massive charge output, demonstrating excellent performance.

### Experimental Validation

The paper validates the effectiveness of the FSMP method through simulated experiments and actual data:

- **Simulated Experiments**: In simulated 8-inch MCP-PMTs, the FSMP method improved energy resolution by up to 12%.
- **Actual Data**: Reanalysis of experimental data from Zhang et al.'s study on new 8-inch MCP-PMTs showed that the FSMP method outperformed existing methods in terms of transit time spread (TTS).

### Conclusion

The FSMP method is a flexible and general Bayesian RJMCMC algorithm that can be used to sample photoelectron sequences from the posterior distribution. The method fully utilizes pulse shape and amplitude information to improve the accuracy of photoelectron signal estimation. With GPU acceleration, the FSMP method performs excellently in processing large amounts of waveform data, making it particularly suitable for MeV neutrino experiments in liquid scintillator detectors, such as the Jinping Neutrino Experiment and the JUNO experiment. In the most optimistic scenario, the FSMP method can improve the resolution of visible energy by 12%.