- Email: [email protected]
T2K 2KM water Cherenkov detector
Nuclear Physics B (Proc. Suppl.) 159 (2006) 97–100 www.elsevierphysics.com
T2K 2KM water Cherenkov detector K. Okumuraa for T2K 2KM working group a
Research Center for Cosmic Neutrino, Institute for Cosmic Ray Research, Univ. of Tokyo Kashiwa-no-ha 5-1-5, Kashiwa, Chiba 277-8582, JAPAN We propose building a detector at 2km position away from the neutrino production point of the T2K experiment. At this distance, almost the same neutrino flux is measured as that observed at the far detector. Using cancellation mechanism the water Cherenkov detector will enable to measure the νμ flux and νe backgrounds expected at far detector with smaller systematics. In this proceeding the detail studies using Monte Carlo simulation on the performance of the 2KM water Cherenkov detector and νe background measurement will be described.
1. Introduction The Tokai-to-Kamioka (T2K) experiment [1] is the second generation accelerator-based neutrino oscillation experiment next to K2K. The new 40 GeV proton synchrotron at the J-PARC accelerator complex located in Tokai. The SuperKamiokande detector (Super-K) [2] will be used as a far detector target. T2K has a neutrino beam which is on the order of 100 times as powerful as the K2K beam. The peak position of νμ neutrino energy spectrum is about 600 MeV. The major physics of T2K experiment are the precise measurement of Δm2 and θ23 and searching for nonzero θ13 . Δm2 and θ23 will be measured with a few percent accuracy by observing the disappearance of νμ events, and sin2 θ13 is expected to be explored by a factor of 10 from the current limit.
and the size of decay tunnel viewed from near detector is not point-like. Figure 1 shows the near/far neutrino flux ratio at 280 m and 2 km. At 280 m the near/far ratio changes drastically and ranged to ±50 % in the region of the energy peak position. While this effect is small at the 2km position. The 2km is distant enough so that neutrino energy spectra are similar in shape within ±5 % in all energy range.
2. The 2KM detectors We propose building the detector 2km away from the neutrino production point since the observed neutrino energy spectrum at 2km position is almost same as that observed at Super-K detector. In T2K experiment, so-called off-axis technique is used to obtain low-energy narrow band neutrino beam. The neutrino beam will be pointed about 2.5◦ away from Super-K. The energy spectrum observed in the near detector at 280 m position has relatively large difference from that observed at Super-K since it is not enough distant that different fraction of neutrino beam will be observed. The detector has a finite size 0920-5632/$ – see front matter © 2006 Published by Elsevier B.V. doi:10.1016/j.nuclphysbps.2006.08.067
Figure 1. The near/far neutrino flux ratio as a function of energy at 280 m (left) and 2 km position (right)
The 2KM detector complex consist of three sub-system, Liquid Argon (LAr) detector [3], water Cherenkov detector and muon range detector (MRD). Figure 2 shows the geometry of the 2KM water Cherenkov detector. The shape is cylindri-
98
K. Okumura / Nuclear Physics B (Proc. Suppl.) 159 (2006) 97–100
cal and aligned horizontally. Its size is 13.8 m long and 9.3 m in diameter, and the fiducial volume is 100 tons. The detector size is determined by two reasons. One is that most of the produced muons in the fiducial volume should be contain inside detector volume in order to reconstruct the precise momentum. Another is that the detector should not be so large that there is more than one neutrino interaction per spill on average.
neutrinos and it becomes intrinsic background of electron appearance. The basic philosophy of the 2KM water Cherenkov detector is to use the same detection method, same target, and same analysis method as in Super-K so as to cancel out as many systematic error between the two detectors as possible. Additionally, the water Cherenkov detector is relatively low cost and well known technology. With 2km detector measurement, the systematic errors due to flux are expected to be canceled between near and far detector, though systematic error related to detector is not correlated. Using this philosophy the 2KM water Cherenkov detector aims at measuring beam νμ flux and νe backgrounds before oscillation with an ultimate accuracy. 3. Monte Carlo simulation study
Figure 2. The cross-sectional view of the 2KM water Cherenkov detector.
Also, the LAr fine grain detector and the MRD will be located in the front of and behind the water Cherenkov detector, respectively. The LAr detector has excellent performance on track separation and low-energy particle detection. It will provide the independent measurement on neutrino interaction, such as non-QE/QE ratio, NC π0 . MRD will be placed downstream of water Cherenkov detector and detects the range of high energy muons over about 1 GeV which exit from the water Cherenkov detector. MRD is very important and essential to 2km detector complex. The high energy neutrinos is the source of neutral current backgrounds for νe appearance. Most of them are produced by kaon decay so it will be also important to constrain kaon produced neutrino flux, because kaon decay produce electron
We performed the simulation study on the 2km detector to check the performance. The simulator is newly developed with GEANT4 frame work and has a flexibility to change detector size and layout including the size of the PMTs. In order to tune the detector response, we simulate cosmic ray muons with the configuration of 1kton water Cherenkov detector at K2K and compared with real data. The event reconstruction code is used as that used at Super-K which are slightly modified and tuned to fit 2km detector condition. We have done quantitative studies to determine the optimum configuration for matching relevant Super-K resolutions and efficiencies. We have studied single-ring muon selection efficiency, ringcounting and particle identification (PID) with 8inch and 20-inch PMT configurations. The single-ring event selection efficiencies for T2K νμ interactions are compared for reconstructing single-ring muons. We can achieve an event selection efficiency same as SuperKamiokande’s to within 1% for 8-inch PMTs. The difference is 5% if 20-inch PMTs are used. Figure 3 shows the PID likelihood distributions of muons produced in the fiducial volume of the Cherenkov detector. Clearly, the tail of the distribution has a more sharply cut off tail, and fewer misidentified events than in the 8-inch case. Also, the single/multi-ring separation in ring-counting
K. Okumura / Nuclear Physics B (Proc. Suppl.) 159 (2006) 97–100
is better for the detector with 8-inch PMTs configuration. Therefore, we conclude that a configuration with 8-inch PMTs is superior to one with 20-inch PMTs.
10
3
10
2
μ-like
e-like
99
tions which fake a single-ring electron, νe interactions from the intrinsic νe flux component of the T2K beam, misidentification of CC-νμ interactions. If there is no observed appearance signal, or the signal is quite small, the error or sensitivity will be dominated by how well we can determine the background to the search. Figure 4 shows this effect as a function of exposure for Super-K with several errors on the total background normalization assumed.
10
1
-10
-8
-6
-4
-2
0
2
4
6
8
10
Figure 3. Likelihood parameter of particle identification estimated by the 2KM Cherenkov detector. The PID distributions obtained with the 8inch PMTs configuration (red) and 20-inch PMTs (black). The events for which the likelihood parameter value is less than zero are misidentified as electrons.
4. Physics with 2KM water Cherenkov detector This section describes the role of the 2KM detectors in characterizing the flux and interactions of neutrinos at the Super-K far detector. The primary concerns are prediction of the background for the νe appearance search and prediction of the νμ flux for the disappearance measurement. The goal of the 2KM detector is to control systematic uncertainties and provide a direct comparison of the unoscillated event rate, independent of Monte Carlo prediction. In this proceedings the νe background measurement is described. 4.1. Measurement of the background for νe appearance When searching for νe appearance in Super-K there will be both an irreducible intrinsic νe background and a background due to event misidentification. They come from NC single-π 0 interac-
Figure 4. Sensitivity to θ13 as a function of exposure for three uncertainties in the background prediction. The first arrow is the exposure for a five year T2K run, the second for five years of a upgraded J-PARC beam with Hyper-Kamiokande as the target.
As can be seen, if the total background uncertainty is allowed to approach 20% the θ13 sensitivity flattens out after 5 years of T2K running as the result becomes systematics-limited. Therefore our goal is to control the total uncertainty to 10% and individual uncertainties to 5%. Using the water Cherenkov detector we will measure the total background for the νe appearance search. We apply the same set of cuts to 2km data as will be applied to the SK far detector data. The background for νe analysis at SuperKamiokande, may be extrapolated from the mea-
100
K. Okumura / Nuclear Physics B (Proc. Suppl.) 159 (2006) 97–100
SK 8.4 14.1 0.4
2KM extrapolated 9.6 ± 0.4 ± 0.8 13.2 ± 0.5 ± 1.0 0.6 ± 0.04 ± 0.03
NC beam νe CC-νμ Table 1 The number of background events for the νe appearance search at Super-K, based on 5 years of running at 1021 pot/yr. The first column (SK) was obtained with the backgrounds at Super-K far detector estimated by the full simulation. The second column (2KM extrapolated) was deduced by the geometrical scaling method. The uncertainties listed first are statistical, derived from the event rate at 2KM. The second uncertainties are systematic. The systematic uncertainty of the fiducial volume are not included in these numbers.
evts/22.5kt/5yrs/100MeV
Predicted BG for νe app. at SK 9 8 from SK MC+reconstruction
7
analysis. In the following study, the ratio of effiSK ciencies 2km in equation 1 is assumed to be equal to 1. We then take into account systematics on the fiducial volume and the energy scale, which we conservatively set to 4% and 3% respectively (based on K2K 1kton experience [4]). Applying this method, we obtain a predicted background of 23.4 events with a statistical uncertainty of 0.6 events and a systematic uncertainty of 1.6 events. The total error is approximately 7.5%, within our 10% goal. The extrapolated background is plotted in Figure 5 where it is compared with the fully simulated and reconstructed background at Super-K. There is good agreement between the two shapes inside the signal window, demonstrating that this simple scaling method is well-suited to predict the background at SK. The above analysis was performed assuming 5-year exposure of the T2K beam with 1021 protons-ontarget per year. After one year of exposure the total error using the same technique would be approximately 8.7%, still within our 10% goal.
6
5. Summary
5 extrapolated from 2KM measurement
4 3 2 1 0 0
1000
2000
3000
4000 5000 reconstructed Eeν (MeV)
Figure 5. Extrapolated background for νe appearance at SK, using the scaling method. The reconstructed energy window is indicated by the dashed lines. Total background for νe appearance at SK. The red and blue points correspond to the results of the full T2K simulation at Super-K and from the 2KM extrapolation, respectively.
surement at 2km using a simple scaling method: 2 MSK SK LSK NSK = N2km × × × , (1) L2km M2km 2km where L is the distance from the detector the neutrino source and M the fiducial mass used in the
The 2KM water Cherenkov detector will be able to measure the νμ flux and νe backgrounds observed at Super-K precisely using cancellation mechanism, and has a potential to improve the sensitivity of oscillation parameter measurement at T2K experiment. REFERENCES 1. Y. Itow et al., The JHF-Kamioka neutrino project, (2001), hep-ex/0106019 2. Y. Fukuda et al., Nucl. Instrum. Meth. A501, 418 (2003) 3. A. Meregaglia, these proceedings 4. Ahn, M. H. et al., Phys. Rev. Lett. 93 051801 (2004)
T2K 2KM water Cherenkov detector K. Okumuraa for T2K 2KM working group a
Research Center for Cosmic Neutrino, Institute for Cosmic Ray Research, Univ. of Tokyo Kashiwa-no-ha 5-1-5, Kashiwa, Chiba 277-8582, JAPAN We propose building a detector at 2km position away from the neutrino production point of the T2K experiment. At this distance, almost the same neutrino flux is measured as that observed at the far detector. Using cancellation mechanism the water Cherenkov detector will enable to measure the νμ flux and νe backgrounds expected at far detector with smaller systematics. In this proceeding the detail studies using Monte Carlo simulation on the performance of the 2KM water Cherenkov detector and νe background measurement will be described.
1. Introduction The Tokai-to-Kamioka (T2K) experiment [1] is the second generation accelerator-based neutrino oscillation experiment next to K2K. The new 40 GeV proton synchrotron at the J-PARC accelerator complex located in Tokai. The SuperKamiokande detector (Super-K) [2] will be used as a far detector target. T2K has a neutrino beam which is on the order of 100 times as powerful as the K2K beam. The peak position of νμ neutrino energy spectrum is about 600 MeV. The major physics of T2K experiment are the precise measurement of Δm2 and θ23 and searching for nonzero θ13 . Δm2 and θ23 will be measured with a few percent accuracy by observing the disappearance of νμ events, and sin2 θ13 is expected to be explored by a factor of 10 from the current limit.
and the size of decay tunnel viewed from near detector is not point-like. Figure 1 shows the near/far neutrino flux ratio at 280 m and 2 km. At 280 m the near/far ratio changes drastically and ranged to ±50 % in the region of the energy peak position. While this effect is small at the 2km position. The 2km is distant enough so that neutrino energy spectra are similar in shape within ±5 % in all energy range.
2. The 2KM detectors We propose building the detector 2km away from the neutrino production point since the observed neutrino energy spectrum at 2km position is almost same as that observed at Super-K detector. In T2K experiment, so-called off-axis technique is used to obtain low-energy narrow band neutrino beam. The neutrino beam will be pointed about 2.5◦ away from Super-K. The energy spectrum observed in the near detector at 280 m position has relatively large difference from that observed at Super-K since it is not enough distant that different fraction of neutrino beam will be observed. The detector has a finite size 0920-5632/$ – see front matter © 2006 Published by Elsevier B.V. doi:10.1016/j.nuclphysbps.2006.08.067
Figure 1. The near/far neutrino flux ratio as a function of energy at 280 m (left) and 2 km position (right)
The 2KM detector complex consist of three sub-system, Liquid Argon (LAr) detector [3], water Cherenkov detector and muon range detector (MRD). Figure 2 shows the geometry of the 2KM water Cherenkov detector. The shape is cylindri-
98
K. Okumura / Nuclear Physics B (Proc. Suppl.) 159 (2006) 97–100
cal and aligned horizontally. Its size is 13.8 m long and 9.3 m in diameter, and the fiducial volume is 100 tons. The detector size is determined by two reasons. One is that most of the produced muons in the fiducial volume should be contain inside detector volume in order to reconstruct the precise momentum. Another is that the detector should not be so large that there is more than one neutrino interaction per spill on average.
neutrinos and it becomes intrinsic background of electron appearance. The basic philosophy of the 2KM water Cherenkov detector is to use the same detection method, same target, and same analysis method as in Super-K so as to cancel out as many systematic error between the two detectors as possible. Additionally, the water Cherenkov detector is relatively low cost and well known technology. With 2km detector measurement, the systematic errors due to flux are expected to be canceled between near and far detector, though systematic error related to detector is not correlated. Using this philosophy the 2KM water Cherenkov detector aims at measuring beam νμ flux and νe backgrounds before oscillation with an ultimate accuracy. 3. Monte Carlo simulation study
Figure 2. The cross-sectional view of the 2KM water Cherenkov detector.
Also, the LAr fine grain detector and the MRD will be located in the front of and behind the water Cherenkov detector, respectively. The LAr detector has excellent performance on track separation and low-energy particle detection. It will provide the independent measurement on neutrino interaction, such as non-QE/QE ratio, NC π0 . MRD will be placed downstream of water Cherenkov detector and detects the range of high energy muons over about 1 GeV which exit from the water Cherenkov detector. MRD is very important and essential to 2km detector complex. The high energy neutrinos is the source of neutral current backgrounds for νe appearance. Most of them are produced by kaon decay so it will be also important to constrain kaon produced neutrino flux, because kaon decay produce electron
We performed the simulation study on the 2km detector to check the performance. The simulator is newly developed with GEANT4 frame work and has a flexibility to change detector size and layout including the size of the PMTs. In order to tune the detector response, we simulate cosmic ray muons with the configuration of 1kton water Cherenkov detector at K2K and compared with real data. The event reconstruction code is used as that used at Super-K which are slightly modified and tuned to fit 2km detector condition. We have done quantitative studies to determine the optimum configuration for matching relevant Super-K resolutions and efficiencies. We have studied single-ring muon selection efficiency, ringcounting and particle identification (PID) with 8inch and 20-inch PMT configurations. The single-ring event selection efficiencies for T2K νμ interactions are compared for reconstructing single-ring muons. We can achieve an event selection efficiency same as SuperKamiokande’s to within 1% for 8-inch PMTs. The difference is 5% if 20-inch PMTs are used. Figure 3 shows the PID likelihood distributions of muons produced in the fiducial volume of the Cherenkov detector. Clearly, the tail of the distribution has a more sharply cut off tail, and fewer misidentified events than in the 8-inch case. Also, the single/multi-ring separation in ring-counting
K. Okumura / Nuclear Physics B (Proc. Suppl.) 159 (2006) 97–100
is better for the detector with 8-inch PMTs configuration. Therefore, we conclude that a configuration with 8-inch PMTs is superior to one with 20-inch PMTs.
10
3
10
2
μ-like
e-like
99
tions which fake a single-ring electron, νe interactions from the intrinsic νe flux component of the T2K beam, misidentification of CC-νμ interactions. If there is no observed appearance signal, or the signal is quite small, the error or sensitivity will be dominated by how well we can determine the background to the search. Figure 4 shows this effect as a function of exposure for Super-K with several errors on the total background normalization assumed.
10
1
-10
-8
-6
-4
-2
0
2
4
6
8
10
Figure 3. Likelihood parameter of particle identification estimated by the 2KM Cherenkov detector. The PID distributions obtained with the 8inch PMTs configuration (red) and 20-inch PMTs (black). The events for which the likelihood parameter value is less than zero are misidentified as electrons.
4. Physics with 2KM water Cherenkov detector This section describes the role of the 2KM detectors in characterizing the flux and interactions of neutrinos at the Super-K far detector. The primary concerns are prediction of the background for the νe appearance search and prediction of the νμ flux for the disappearance measurement. The goal of the 2KM detector is to control systematic uncertainties and provide a direct comparison of the unoscillated event rate, independent of Monte Carlo prediction. In this proceedings the νe background measurement is described. 4.1. Measurement of the background for νe appearance When searching for νe appearance in Super-K there will be both an irreducible intrinsic νe background and a background due to event misidentification. They come from NC single-π 0 interac-
Figure 4. Sensitivity to θ13 as a function of exposure for three uncertainties in the background prediction. The first arrow is the exposure for a five year T2K run, the second for five years of a upgraded J-PARC beam with Hyper-Kamiokande as the target.
As can be seen, if the total background uncertainty is allowed to approach 20% the θ13 sensitivity flattens out after 5 years of T2K running as the result becomes systematics-limited. Therefore our goal is to control the total uncertainty to 10% and individual uncertainties to 5%. Using the water Cherenkov detector we will measure the total background for the νe appearance search. We apply the same set of cuts to 2km data as will be applied to the SK far detector data. The background for νe analysis at SuperKamiokande, may be extrapolated from the mea-
100
K. Okumura / Nuclear Physics B (Proc. Suppl.) 159 (2006) 97–100
SK 8.4 14.1 0.4
2KM extrapolated 9.6 ± 0.4 ± 0.8 13.2 ± 0.5 ± 1.0 0.6 ± 0.04 ± 0.03
NC beam νe CC-νμ Table 1 The number of background events for the νe appearance search at Super-K, based on 5 years of running at 1021 pot/yr. The first column (SK) was obtained with the backgrounds at Super-K far detector estimated by the full simulation. The second column (2KM extrapolated) was deduced by the geometrical scaling method. The uncertainties listed first are statistical, derived from the event rate at 2KM. The second uncertainties are systematic. The systematic uncertainty of the fiducial volume are not included in these numbers.
evts/22.5kt/5yrs/100MeV
Predicted BG for νe app. at SK 9 8 from SK MC+reconstruction
7
analysis. In the following study, the ratio of effiSK ciencies 2km in equation 1 is assumed to be equal to 1. We then take into account systematics on the fiducial volume and the energy scale, which we conservatively set to 4% and 3% respectively (based on K2K 1kton experience [4]). Applying this method, we obtain a predicted background of 23.4 events with a statistical uncertainty of 0.6 events and a systematic uncertainty of 1.6 events. The total error is approximately 7.5%, within our 10% goal. The extrapolated background is plotted in Figure 5 where it is compared with the fully simulated and reconstructed background at Super-K. There is good agreement between the two shapes inside the signal window, demonstrating that this simple scaling method is well-suited to predict the background at SK. The above analysis was performed assuming 5-year exposure of the T2K beam with 1021 protons-ontarget per year. After one year of exposure the total error using the same technique would be approximately 8.7%, still within our 10% goal.
6
5. Summary
5 extrapolated from 2KM measurement
4 3 2 1 0 0
1000
2000
3000
4000 5000 reconstructed Eeν (MeV)
Figure 5. Extrapolated background for νe appearance at SK, using the scaling method. The reconstructed energy window is indicated by the dashed lines. Total background for νe appearance at SK. The red and blue points correspond to the results of the full T2K simulation at Super-K and from the 2KM extrapolation, respectively.
surement at 2km using a simple scaling method: 2 MSK SK LSK NSK = N2km × × × , (1) L2km M2km 2km where L is the distance from the detector the neutrino source and M the fiducial mass used in the
The 2KM water Cherenkov detector will be able to measure the νμ flux and νe backgrounds observed at Super-K precisely using cancellation mechanism, and has a potential to improve the sensitivity of oscillation parameter measurement at T2K experiment. REFERENCES 1. Y. Itow et al., The JHF-Kamioka neutrino project, (2001), hep-ex/0106019 2. Y. Fukuda et al., Nucl. Instrum. Meth. A501, 418 (2003) 3. A. Meregaglia, these proceedings 4. Ahn, M. H. et al., Phys. Rev. Lett. 93 051801 (2004)