1 The MicroBooNE Experiment

MicroBooNE and its Cross Section Measurement

NuPhys2016-Tsai April 13, 2018

MicroBooNE and its Cross Section Measurement

Yun-Tse Tsai, for the MicroBooNE collaboration

SLAC National Accelerator Laboratory, Menlo Park, CA, USA

MicroBooNE (the Micro Booster Neutrino Experiment) is a short-baseline neutrino experiment based on the technology of a liquid-argon time-projection chamber (LArTPC), and has recently completed its first year of data-taking in the Fermilab Booster Neutrino Beam. It aims to address the anomalous excess of events with an electromagnetic final state in MiniBooNE, to measure neutrino-argon interaction cross sections, and to provide relevant R&D for the future LArTPC experiments, such as DUNE. In these proceedings, we present the first reconstructed energy spectrum of Michel electrons from cosmic muon decays, the first kinematic distributions of the candidate muon tracks from -argon charged-current interactions, and a demonstration of an electromagnetic shower reconstruction from s produced by -argon charged-current interactions. The results demonstrate the first fully automated reconstruction and selection algorithms in a large LArTPC and serve as foundations for future measurements.


NuPhys2016, Prospects in Neutrino Physics

Barbican Centre, London, UK, December 12–14, 2016

1 The MicroBooNE Experiment

The MicroBooNE experiment is a neutrino experiment aiming to measure oscillation of neutrino flavors and neutrino-nuclear interaction cross sections. Located in the Booster Neutrino Beam (BNB) at Fermilab at a baseline of 470 m, MicroBooNE is the first experiment in the U.S. utilizing a large liquid-argon time-projection chamber (LArTPC) [1].

The primary physics goal of MicroBooNE is to address the excess of data events with an electromagnetic object in the regime of neutrino energy of 200 – 500 MeV reported by the MiniBooNE experiment [2]. As a Cherenkov detector filled with mineral oil, MiniBooNE was not able to distinguish electrons from photons. If the excess comes from events with an electron, it may imply existence of a sterile neutrino from the interpretation of the oscillation. On the other hand, if the electromagnetic object in those events is a photon, it may indicate an unknown background component. A LArTPC is able to distinguish electrons and photons by looking for the topology at the start of an electromagnetic shower, and can thereby be exploited to investigate the MiniBooNE anomaly.

Measuring neutrino-argon interaction cross sections is another goal of the MicroBooNE experiment. Neutrino-nuclear interactions are currently not well understood, and have significant impacts on the precision of neutrino oscillation measurements. In particular, one of the most important neutrino experiments in the next generation, Deep Underground Neutrino Experiment (DUNE) [3], aims to address the CP-invariance violation in the lepton sector and the neutrino mass ordering by measuring rates and energy spectra of neutrino oscillation with the LArTPC technology. Therefore, precise measurements of neutrino-argon cross sections at MicroBooNE will be relevant.

In addition, the MicroBooNE detector is utilized to explore astroparticle and exotic physics, such as the detection of neutrinos from core-collapse supernova explosions, and searches for nucleon decays, nucleon oscillation, as well as dark matter candidates. The size of the MicroBooNE detector is not sensitive to most of the searches. However, we can demonstrate the technique, study backgrounds, and probe the thresholds of large LArTPCs. Moreover, LArTPC R&D and detector physics can be performed with MicroBooNE. For example, we demonstrate the purification of liquid argon, the design of the high voltage system, the cold electronics, readout and data acquisition systems. The effects of detector noise, electron recombination and attenuation can also be characterized at MicroBooNE. All the results will provide the DUNE experiment with valuable information.

2 Detector and its Performance

MicroBooNE consists of one time-projection chamber (TPC), with 89 tons of liquid argon in its active volume. As shown in Fig. 1, the TPC has dimensions of 10.4 m in the BNB direction, 2.3 m in vertical, and 2.5 m between the cathode and anode, along which a high voltage electric field of 273 V/cm is applied. There are three wire planes at the anode, each oriented by a degree of and with respect to the vertical, reading out the deposited charges at 2 MHz. A light collection system consisting of 32 8-inch photomultiplier tubes (PMTs) is mounted behind the anode wire planes. More details about the detector can be found in [4].

Charged particles produced from neutrino-argon interactions ionize argon atoms and create scintillation light. The scintillation light, produced in a time scale of 6 nanoseconds, is collected by the PMTs, determining the event timing, while the ionization electrons slowly drift towards the anode. Those electrons pass through the first two wire planes, leaving induced current, and then are collected in the third wire planes. Assuming a constant drift velocity ( mm/s), the electron drift time is proportional to the drift distance, and therefore each wire plane provides a two-dimensional image with the granularity of 3 mm (wire pitch) times 0.6 mm (the sampling rate of digitization under the current high voltage configuration). Fig. 2 illustrates the high spatial resolution of MicroBooNE LArTPC and the capability of characterizing a complicated event topology. A three-dimensional event can be reconstructed from the three two-dimensional images from the three wire planes.

Figure 1: (a) Schematic drawing of the MicroBooNE cryostat which hosts the TPC. (b) The expected components and energy spectra of the BNB flux at MicroBooNE in the neutrino mode.
Figure 2: An annotated display of an event from MicroBooNE data, taken from the third wire plane, where the wires are vertical. The color scale indicates the amount of deposited charges. The horizontal direction of the display represents the wire numbers, while the vertical direction shows the drift time, or the drift distance. The display characterizes a neutrino-argon interaction with a few tracks and two electromagnetic showers in the final state, overlaid with multiple tracks from cosmic rays.

MicroBooNE started taking BNB neutrino data on October 15th, 2015. The composition of the BNB flux, dominated by , can be found in Fig. 1. The detector and the data acquisition have performed stably [5], and protons on target (POT) have been recorded as of the date of the conference. Fig. 3 shows the cumulative efficiency of the data acquisition at MicroBooNE.

Figure 3: The recorded protons on target (POT) per week (histograms) and the cumulatively delivered and recorded POT (curves).

3 Reconstruction of Physics Objects

Fully automated reconstruction algorithms are required to tackle the great amount of charge deposition from particles produced in neutrino-argon interactions, and from particles induced by cosmic rays during the long readout window (4.8 milliseconds) in an event. They are also needed to reduce the bias introduced by a visual scan. To process data collected at the TPC, we start with filtering the noise from the detector electronics [6], and then extract hits from the digital waveforms. Multiple clustering algorithms are applied, associating the hits originating from the same charged particles [7].

We use a three-dimensional track fitter to reconstruct tracks and remove hits associated to through-going tracks, which likely represent cosmic rays. Fig. 4 illustrates reconstructed tracks in an event taken outside the BNB operation window. Subsequently, we reconstruct the remaining hits and obtain tracks, electromagnetic showers, and neutrino-argon interaction vertices. An event containing a neutrino-argon interaction with reconstructed tracks and showers in the final state can be found in Fig. 4.

Figure 4: Three-dimensional reconstructed events from data collected at MicroBooNE: (a) An event containing cosmic rays collected outside the BNB window. The three boxes show the full readout length per event, corresponding to 4.8 milliseconds. The red highlighted box outlines the 1.6 milliseconds after the trigger time. The colored lines represent reconstructed tracks from cosmic rays. (b) An event containing a -argon charged-current interaction with a production, selected by a visual scan. The white points are the reconstructed locations of deposited charges in the three-dimensional volume in the TPC. The colored cones represent the geometry of the reconstructed electromagnetic showers, possibly originating from the photons from the decay of the .

4 First Analyses

Utilizing the reconstructed physics objects, we develop different selection criteria for various analyses. In this section, the first analyses from MicroBooNE will be discussed.

4.1 Michel Electrons

To further understand the detector response in the tens of MeV energy range and the muon identification, we study the energy spectrum of Michel electrons, electrons in the decay products of stopping muons [8].

In this analysis, we use muons from cosmic rays. The data sample contains 280,751 events collected outside of the BNB operation windows, corresponding to 1,347 seconds. A set of clustering algorithms are developed, profiling the deposited charges based on the highly resoluted topological and calorimetric information provided by the third wire plane, which collects ionization electrons. We identify the tracks as stopping muons by looking for an increase in the charge deposition per unit length towards the end of the track, and the electron candidate is recognized as the track coming after the identified muon stopping point at an angle with respect to the muon track. The reconstruction and selection algorithms are fully automated.

The energy of Michel electrons is calculated from the reconstructed charges at the third (collection) wire plane with appropriate electronic calibration factors and correction factors. The correction factors account for two effects,

  • recombination of argon ions and ionization electrons,

  • attenuation of the ionization electrons during the drift path owing to the electronegative contamination in the liquid argon.

In this analysis, constant correction factors are applied; further development and studies are underway to better model the two effects and the corresponding correction factors.

Figure 5: The reconstructed energy spectrum of Michel electrons from data and Monte Carlo simulation. The uncertainty in both data and Monte Carlo simulation accounts for the statistic uncertainty.

The distribution of the reconstructed energy from Michel electrons is shown in Fig. 5. The energy distribution of Michel electrons typically has a sharp edge at 52.2 MeV, half the mass of muons. The distortion of the spectrum is owing to the fact that the radiated photons, which can start the pair production in tens of centimeters, are poorly included in the energy reconstruction. Nonetheless, the reasonable agreement between the spectra from the Monte Carlo (MC) simulation and the data demonstrates our understanding of Michel electrons in LArTPCs. The remaining difference in the two spectra may come from variation of calibration factors in different TPC wire channels. Further studies are ongoing to improve the analysis.

4.2 -argon Charged-current Interactions

The measurement of the -argon charged-current interaction cross section provides us with a foundation for comparisons to theoretical calculations and other experimental results. In addition, it serves as a common starting point for further measurements of exclusive interaction channels, such as the charged-current interaction with the production of a . Owing to the surface location, the recorded events at MicroBooNE are dominated by cosmic rays, and identifying and removing those background events are challenging. In this analysis, we present multiple kinematic distributions of muons produced from the -argon charged-current interaction. The analysis outlines required tools for data quality and detector stability checks, for physics object reconstructions and event selections. It also guides us towards strategies and required improvements for all MicroBooNE analyses [9].

A data sample of 546,910 events, corresponding to protons on target, is analyzed. The MC event generator GENIE is used to simulate the neutrino-argon interactions, while the particles induced by cosmic rays in these events are modeled by the CORSIKA simulation program. The passages of particles through the detector are simulated by GEANT4. Further, we exploit data collected outside the BNB operation windows for an estimation of the background events containing no neutrino-argon interaction.

Utilizing the reconstruction algorithms described in Sec. 3, we develop fully automated selection schemes. As a -argon charged-current interaction produces a muon, leaving a long track in the detector, we select such events coincident with the BNB beam timing. One of the schemes and the consequent kinematic distributions are presented in these proceedings. We require

  1. a light signal above 50 photoelectrons (P.E.) within the BNB operation window (1.6 s, much shorter than the TPC readout window, 4.8 ms), indicating activities coincident with the beam timing,

  2. at least a track longer than 70 cm, identifying as the muon candidate,

  3. a light signal above 50 P.E. in agreement with the position of the candidate muon track in the beam direction,

  4. the reconstructed interaction vertex within the fiducial volume (20 cm from the border in vertical, and 10 cm from the border in the other dimensions), removing events induced by cosmic rays or other background interactions,

  5. at least a track starting within 3 cm from the interaction vertex, ensuring the production of charged particles near the vertex,

  6. multiple sets of selection criteria on track kinematics for different charged particle multiplicities.

More details can be found in [9].

In the selected sample, we obtain an efficiency convoluted with acceptance of 30%, while obtaining the purity of 65%. The dominant background events originate from cosmic rays. Fig. 6 shows kinematic distributions of the candidate muon tracks, including the length, the angle with respect to the neutrino beam direction (), and the azimuthal angle around the beam direction (). The pure cosmic ray background events, determined using the data collected outside of the BNB operation window, have been subtracted from those distributions.

Figure 6: Kinematic distributions of the candidate muon tracks in the selected events: (a) track length, (b) , where denotes the angle of the track with respect to the neutrino beam direction, and (c) , the azimuthal angle around the beam direction. The number of events in the simulation is normalized to that in the data. Events from pure cosmic rays are subtracted. The uncertainty in the Monte Carlo simulation and the data accounts for the statistical uncertainty only. The pions produced from neutral-current interactions misidentified as muon tracks are typically shorter, and contribute to the first two bins in the track length distribution. The efficiency at , corresponding to the vertical direction, is lower because the candidate muon tracks in vertical are more likely to be identified as cosmic rays and are removed.

The distributions from the MC simulation agree reasonably well with those from the data, indicating our capability of modeling the signal, background events, as well as the detector response. Studies of systematic uncertainties in MicroBooNE are currently underway; nonetheless, we expect the major contributions of the systematic uncertainties would originate from the modeling of the BNB flux, the detector effects (e.g. the detector noise, non-uniformity of the electric field), and the simulation of neutrino-nucleus interactions.

4.3 -argon Charged-current Interactions with Production

Reconstruction algorithms for electromagnetic showers are the key step towards the oscillation analysis. The reconstructed invariant mass is important to demonstrate the performance of our electromagnetic shower reconstruction algorithms, as it requires both the direction and the energy of the reconstructed electromagnetic showers originating from the photons from the decays of s. Out of the candidate events of -argon charged-current interactions, we visually select a few events potentially containing a production. As illustrated in Fig. 7, the decays into two photons, each travels a few to a few tens of centimeters and then starts developing an electromagnetic shower. We thereby identify events containing two detached electromagnetic showers pointing back to the interaction vertex, which can be anchored by the beginning of the candidate muon track [10].

Figure 7: An event containing a -argon charged-current interaction with a production candidate, selected by a visual scan: (a) display with raw digital waveforms collected at the third wire plane, (b) display with reconstructed hits (black points), tracks (red lines), and electromagnetic showers (colored triangles) projected into the third wire plane.

Fig. 7 shows the reconstructed shower cones and the hits used to form the cones. Developments on calorimetry and studies on systematic uncertainties are currently underway. Further, we have made progress towards fully automated selection exclusively on this interaction channel with a fair efficiency, and plan to obtain results in the near future.

5 Outlook and Summary

MicroBooNE is unique in its physics goals of addressing the anomaly reported by the MiniBooNE experiment, of measuring neutrino-argon cross sections, and of conducting R&D for both astroparticle and exotic physics searches and LArTPC performances. It has been fully operational and stably taking neutrino data for 10 months, recording POT on tape. In these proceedings, we present the first fully automated reconstruction and event selection algorithms for LArTPCs and the first results, the energy spectrum of Michel electrons, the kinematic distributions of -argon charged-current interactions, and the demonstration of the shower reconstruction algorithms with the events containing a production. The energy spectrum of Michel electrons is a standard tool used for energy calibrations, while the distributions and the demonstration of the -argon charged-current interactions guide us towards the developments and studies for the final cross section measurements and the neutrino oscillation analyses.

It is important to precisely measure the neutrino-nucleus cross sections as their uncertainties are the major components of the systematic uncertainties of the neutrino oscillation measurements. As of today, only few neutrino-argon cross sections have been reported [11], and therefore the cross sections measured by MicroBooNE would significantly improve our knowledge. In particular, the energy regime of neutrinos produced by BNB, from 200 MeV to 2 GeV, corresponds to the second oscillation maximum of the DUNE experiment. We plan to deliver an inclusive neutrino-argon cross section measurement in 2017, and will obtain several measurements of exclusive and differential cross sections.

In 2018/2019, the Short Baseline Neutrino (SBN) Program will be operational, aiming to answer the question of existence of sterile neutrinos, which could potentially explain the MiniBooNE anomaly in the low energy regime and the earlier anomaly reported by the Liquid Scintillator Neutrino Detector (LSND) [12]. The SBN program will utilize the BNB neutrino beam as the neutrino source, and will contain a near and a far detectors to MicroBooNE. The Short Baseline Near Detector, SBND, will characterize the neutrino beam flux, and have large statistics for neutrino-argon cross section measurements, while the far detector, ICARUS, will be sensitive to the relevant parameter space. MicroBooNE will continue operating as part of the SBN program, and continue to deliver valuable information for the design of future LArTPC experiments, including both the detectors in the SBN program and the detectors in the DUNE experiment.


  1. The MicroBooNE Collaboration, The MicroBooNE Technical Design Report (2012), http://www-microboone.fnal.gov/publications.
  2. The MiniBooNE Collaboration, Unexplained Excess of Electronlike Events from a 1-GeV Neutrino Beam, PRL 102, 101802 (2009).
  3. R. Acciari et al., Long-Baseline Neutrino Facility (LBNF) and Deep Underground Neutrino Experiment (DUNE), arXiv:1512.06148 (2015).
  4. R. Acciari et al., Design and construction of the MicroBooNE detector, JINST 12 (2017).
  5. The MicroBooNE collaboration, MicroBooNE Detector Stability, MICROBOONE-NOTE-1013-PUB (2016).
  6. The MicroBooNE collaboration, Noise Characterization and Filtering in the MicroBooNE TPC, MICROBOONE-NOTE-1016-PUB (2016).
  7. The MicroBooNE collaboration, The Pandora Multi-algorithm Approach to Automated Pattern Recognition in LArTPC Detectors, MICROBOONE-NOTE-1015-PUB (2016).
  8. R. Acciari et al., Michel Electron Reconstruction Using Cosmic-Ray Data from the MicroBooNE LArTPC, arXiv:1704.02927 (2017).
  9. The MicroBooNE collaboration, Selection and kinematic properties of charged-current inclusive events in 5E19 POT of MicroBooNE data, MICROBOONE-NOTE-1010-PUB (2016).
  10. The MicroBooNE collaboration, Demonstration of 3D Shower Reconstruction on MicroBooNE Data, MICROBOONE-NOTE-1012-PUB (2016).
  11. G.P. Zeller, Neutrino Cross Section Measurements, Chin. Phys. C 40, 100001 (2016).
  12. R. Acciari et al., A Proposal for a Three Detector Short-Baseline Neutrino Oscillation Program in the Fermilab Booster Neutrino Beam, arXiv:1503.01520 (2015).
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
Add comment
Loading ...
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description