- Email: [email protected]
SNO & Solar Neutrino Results
Nuclear Physics B (Proc. Suppl.) 168 (2007) 84–89 www.elsevierphysics.com
SNO & Solar Neutrino Results J. Maneiraa
∗
a
Laborat´orio de Intrumenta¸c˜ao e F´ısica Experimental de Part´ıculas (LIP-Lisboa), Av. Elias Garcia, 14, 1, 1000-149 Lisboa, Portugal
In this talk, the latest results from the presently active solar neutrino experiments, SNO and Super-Kamiokande were reviewed, ranging from 8 B neutrino flux and spectrum measurements and their impact on neutrino oscillation paramenters, to measurements of time variations and searches for hep and relic neutrinos. Prospects for the upcoming results from new analyses or data from new data taking phases were also discussed.
1. Introduction Currently active solar neutrino experiments have measured the flux, spectrum and rate variations in time of the high energy 8 B neutrinos using the water Cherenkov technique. SuperKamiokande (SK)[1] is a 50 kton (22.5 kton fiducial size) water detector located in the Kamioka laboratory in Japan, under an overburden of 2700 mwe. SK detects solar neutrinos via elastic scattering on electrons. From 1996 to 2001, SK took data with 11146 20 inch PMTs and a photo coverage of 40 %, a period referred to as ”SK-I”. After the 2001 incident, the detector was refurbished with 5200 PMTs and took data with a coverage of 19 % from2002 to 2005, the period referred to as ”SK-II”. In July 2006, the detector was rebuilt to its original coverage and the ”SK-III” period started. Here we will report on preliminary analyses of the SK-II data. For a detailed discussion on prospects for SK-III, see [2]. The Sudbury Neutrino Observatory (SNO) [3] is a 1 kton heavy water (D2 O) detector located at a depth of 2 km (6000 mwe) in the INCO Creighton mine near Sudbury, Canada. The heavy water is contained in a 12 m diameter acrylic vessel (AV) that sepearates it from about 5 kton of normal water that serves as a gamma and neutron shield. The AV and heavy water volume are surrounded by an array of 9456 PMTs ∗ On behalf of the SNO collaboration. Thanks to the Super-K collaboration for assistance in the update of their latest results.
0920-5632/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.nuclphysbps.2007.02.061
mounted on a stainless steel support structure. SNO detects solar neutrinos mainly through reactions on the deuteron, but also through elastic scattering with electrons, as in SK: νe + d → p + p + e− νx + d → p + n + νx νx + e− → νx + e−
(CC), (NC), (ES).
While the charged current (CC) reaction is sensitive only to νe ’s, the neutral current reaction (NC) is equally sensitive to all active neutrino flavors (x = e, μ, τ ). The elastic scattering (ES) reaction is sensitive to all electron flavors as well, but with reduced sensitivity to νμ and ντ . The observation of neutrons from the NC reaction provides a model-independent measurement of the total flux of active 8 B solar neutrinos. The neutron counting is carried out in different ways in each of the three phases of SNO: detection of a 6.25 MeV γ ray following capture on a deuteron in the first phase; detection of a 8.6 MeV γ ray cascade following capture on 35 Cl in the second phase, where 2 tonnes of salt were added to the D2 O. In the third phase, the neutrons are detected by an array of 40 3 He counters (Neutral Current Detectors, or NCDs), independently from Cherenkov events. Here we will report on the current and forthcoming analyses of data from phases I and II of SNO, and discuss prospects for phase III.
J. Maneira / Nuclear Physics B (Proc. Suppl.) 168 (2007) 84–89
2.
8
85
B Solar Neutrino Measurements
2.1. Flux measurements The results for the 8 B flux, from the full SNO salt phase dataset[4] are, in units of 106 cm−2 s−1 : ΦCC
+0.08 = 1.68 ± 0.06(stat.)−0.09 (syst.)
ΦN C
+0.38 = 4.94 ± 0.21(stat.)−0.34 (syst.)
ΦSN OES
= 2.35 ± 0.22(stat.) ± 0.15(syst.)
The difference bertween ΦN C and ΦCC shows at a 7σ level that there is a non-νe component in the solar neutrino flux, demonstrating flavor change. These results are in agreement with the previous results from phase I[5], and the NC flux measurement is in good agreement with Solar Standard Model (SSM) predictions[6]. The preliminary result for the 8 B flux from the full SK-II period[7,8], measured with the ES reaction, is, in units of 106 cm−2 s−1 : ΦSKES
=
2.38 ± 0.05(stat.)+0.16 −0.15 (syst.)
These results are in agreement with the more precise results from the SK-I period [9] and with the SNO measurements of the different flavor components of the 8 B flux. 2.2. Spectrum measurements Figure 1 shows the measured spectrum of the ES reaction in SUper-Kamiokande, relative to the predicted spectrum, using the flux and shape from the SSM[6]. The preliminary results for SKII are compatible with measurements from SK-I, even if at a higher threshold and with less precision, and with an undistorted 8 B shape. The pattern of hit PMTs for the gamma cascade from neutron capture in 35 Cl is more smeared than for the single electron CC events, and this allows the signal extraction in the SNO salt phase to use an isotropy parameter (β14 ), in addition to position of the event and direction with respect to the Sun. So the energy spectrum does not need to be used and can be measured for the CC and ES reactions. Figure 2 shows the extracted spectrum of the CC reaction, compared to the SSM and LMA-oscillation predicted shapes. Within the systematic error band (shown in green), the results are compatible both with the standard and oscillated 8 B shape.
Figure 1. Super-Kamiokande elastic scattering spectrum, normalized with respect to the SSM 8 B prediction.
2.3. Global analysis of neutrino oscillations The total flux of 8 B neutrinos from the SNO NC measurement, the spectrum measurements of the CC and ES reactions in SNO and SK-I (day and night spectra considered separately) were included in a global two-flavor neutrino oscillation analysis, together with the radiochemical solar neutrino measurements and rate and shape predictions from the Solar Standard Model. The data from the KamLAND reactor experiment[11] were also included. Figure 3 shows the allowed regions from this oscillation analysis. The best-fit point is: Δm2 θ
−5 = 8.0+0.6 eV 2 −0.4 × 10
= 33.9+2.4 −2.2 deg
Only the so-called LMA-I region is allowed by the current data, and maximal mixing is excluded at more than 4 σ. 2.4. Searches for time variation of the event rate Both Super-Kamiokande and SNO have performed analyses of the time variations of their signal rates, with sensitivity to periodicities of very different magnitudes. A difference between the average day and night rates would be evidence for
J. Maneira / Nuclear Physics B (Proc. Suppl.) 168 (2007) 84–89
300 Data
6 m2 (10 eV 2)
Systematic uncertainties
250
SSM 8B model shape LMA 8B model shape
200
20 68% CL
-5
Events/(0.5 MeV)
86
150
100
95% CL
15
99.73% CL
10
50
0
6
7
8
9
10
11
12
13
T eff (MeV)
5 0
Figure 2. Extracted CC energy spectrum compared to that predicted with the best-fit LMA parameters. Tef f is the effective electron kinetic energy.
an oscillation-in-matter induced regeneration of electron neutrinos when crossing the Earth. This effect is predicted at the few % level for a part of the allowed LMA oscillation parameter region. Variations with longer time scales, in particular those correlated with the solar cycles, could be due to neutrino spin-flip conversion in the magnetic fields of the Sun. The search for day/night rate variations is parametrized with the day/night assymetry: ADN
=
2(Day − N ight) (Day + N ight)
The results for phases I and II of SK (preliminary for SK-II, systematics still under study) and combined phase I and II of SNO are ADN (SKI) = −0.021 ± 0.020(stat.)+0.013 −0.012 (syst.) ADN (SKII) = −0.064 ± 0.043(stat.) ADN (SN O)
= 0.037 ± 0.040(stat. + syst.),
so there is no evidence of a day-night asymmetry in the present data from both experiments. Figure 4 shows the variation of the SK event rate in funcion of time, including also the preliminary analysis of the more recent SK-II dataset.
0.2
0.4
0.6
0.8
1 tan2e
Figure 3. Allowed region of neutrino oscillation parameters from a global analysis of all solar neutrino data, combined with the results of the KamLAND reactor neutrino experiment.
Figure 5 shows the variation of the signal rate in SNO (relative to the average rate), but in function of the time since the perihelion, so all data in folded in the same one-year range. In both cases, the expected rate variation due to the eccentricity of the Earth orbit is observed (and SNO fits the eccentricity to be 0.0143 ± 0.0086, compared to 0.0167). More sophisticated periodicity analyses were performed on the datasets of both experiments by searching for peaks in a frequency spectrum with the Lomb-Scargle periodogram technique [12,10]. A large number of Monte Carlo simulated datasets were generated assuming no variation beyond eccentricity, and for SK, 19% of those (35% for SNO) gave a peak at least as large as the largest peak observed in the data. Independent analyses of the binned SK and SNO data were also performed (see, for instance[13], also finding no variations beyond what is expected from statistical fluctuations, so there is no evidence for time modulation in the SK and SNO data.
87
Relative rate
J. Maneira / Nuclear Physics B (Proc. Suppl.) 168 (2007) 84–89
1.1
1.05 1
0.95 0.9 0
50
100
150
200
250
300
350
Days since perihelion
Figure 5. Relative SNO event rate as a function of days since perihelion, normalized to the mean rate. The solid line represents the expected variation due to the eccentricity of the Earth orbit. Figure 4. Super-Kamiokande event rate in function of time. The solid line represents the expected variation due to the eccentricity of the Earth orbit.
3. Searches for hep neutrinos In addition to 8 B neutrinos, the SSM predicts a high energy component of the solar neutrino spectrum that can, in principle, be detected by water Cherenkov detectors. These are the hep netrinos that have a predicted flux about 600 times smaller than 8 B, but a higher endpoint, at 18.77 MeV. In the energy region between the endpoint of the 8 B neutrinos and the rise of the atmospheric neutrinos, the background is small enough to allow a search for the hep neutrinos. Both SK[9] and SNO[14] have performed that analysis. SK has chosen the energy interval 18-21 MeV that maximizes the hep/8 B ratio based on Monte Carlo simulations. The observation of 4.9 events, when 1.06 are expected from the hep signal and 1.72 from 8 B, places the upper limit on the hep flux at 7.3×104cm−2 s−1 (90% C.L.), assuming all hep neutrinos are electron-type. Applying a correction for the effect of oscillations, this is equivalent to a limit on the total flux of hep neutrinos of 1.5 × 105 cm−2 s−1 . The most sensitive energy interval was chosen for the SNO analysis by evaluating the expected signal and background levels, before examining the data, and the limits were set using a modi-
fied Felman-Cousins technique. The systematics of the energy response, important for the prediction of the 8 B background, was evaluated from the analysis of three different calibration sources: 16 N (6.13 MeV gamma), pT (3 H(p, γ)4 He, 19.8 MeV gamma source), and Michel electrons (from cosmic muon decay). The background contribution from atmospheric neutrinos, and the effect of neutrino oscillations were considered. In the selected energy interval of 14.3-20 MeV, 0.99 signal and 3.1 background events were expected. Two events were observed, placing a limit on the total flux of hep neutrinos at 2.3 × 104 cm−2 s−1 , that is, 2.9× the SSM prediction and an improvement of ×6.5 over the previous limit. 4. Outlook 4.1. Low energy threshold analysis Since the analysis published in [4], several improvements to the Monte Carlo and analysis were developed, mainly to the energy reconstruction, with the aim of improving the uncertainties on energy scale and resolution, reduce the low energy background tails and reduce the energy threshold for signal extraction. The analysis of the optical calibration data allowed the determination of the relative efficiencies of a large fraction ( 70%) of the PMTs and the use of these parameters both in the Monte Carlo and energy estimators allowed a better modelling of the energy resolution. A new energy estimator using both prompt and late
Events / 1.425 MeV
88
J. Maneira / Nuclear Physics B (Proc. Suppl.) 168 (2007) 84–89 20 18 16
Data Estimated Background
14
Predicted hep signal
12 10
hep signal box
8 6 4 2 0
13
14
15
16
17
18
19
20
21
Teff (MeV)
Figure 6. Distribution of events in the region of the 8 B endpoint. Also shown are the estimated number of background events, including systematic uncertainties, and the standard model prediction for the hep signal.
light was developed and the previous estimator was also improved. Both show a spread of about 0.5% on the difference between data and MC over the radial position of the 16 N calibration source, compared to a spread of 0.7% in previous analyses. The joint analysis of the phase I and II data is still underway but an energy threshold of 4.0 MeV (in Tef f ) appears possible. 4.2. Prospects for Phase III In Phase III of SNO, that started in the Summer of 2004 and ended November 2006, an array of 36 proportional counters filled with 3 He (plus 4 counters filled with 4 He for background measurement) was installed in the D2 Ovolume. These counters, called Neutral Current Detectors (NCDs) detect neutrons through the process 3 He(n, p)3 H, releasing a proton and a tritium nucleus with Q-value of 764 keV. The counters are about 10 m long and have a diameter of 5 cm. They are attached to the bottom of the SNO
acrylic vessel on a grid of 1 × 1 m. The neutrons will preferentially capture on the 3 He, so that the detection of Neutral Current events is done in the NCDs on an event-by-event basis, independently from CC and ES Cherenkov signals seen by the PMT array. An extensive program of in-situ and ex-situ calibrations was carried out during phase III of the experiment, to measure the response of the array to both the neutron signal and the backgrounds, mainly from α particles. Figure 7 shows a fit to SNO phase III data with a characteristic neutron spectrum obtained from calibration. The NCD electronics allows the digitization of the signal pulse and the shape of that pulse is used to distinguish neutrons from background, and reconstructing position of the event. The Monte Carlo and analysis of the PMT array data was significantly updated for phase III, to take into the account the optical effects of shadows and reflections of the Cherenkov on the proportional counters. The method for the optical calibration of SNO was based on normalizing the PMT occupancy of an off-center laser run by the occupancy of the same PMT in a central run (in order to cancel the PMT efficiency from the occupancy model prediction). Now, in the third phase, in order to maintain the simplicity of the calibration model and reject partially shadowed PMTs, this method resulted in a significant statistics loss. So a new method was developed, using estimates of the relative PMT efficiencies, and correcting for the effect of reflections. In all other analyses partially shadowed PMTs are not rejected, so the loss of light due to the NCDs is not large, and is mostly compensated by upgrades to the electronics that were made between phases II and III, that allow the PMT thresholds to be set lower. The NC measurement in the salt phase of SNO is expected to have smaller and significantly different systematic errors than the NC measurement in the previous phases. The CC measurement is also expected to improve since the large correlation with NC in the signal extraction in previous phases is now almost eliminated.
J. Maneira / Nuclear Physics B (Proc. Suppl.) 168 (2007) 84–89
89
D68 (2003) 092002. 13. G. Ranucci, Phys. Rev. D73 (2006) 103003. See hep-ph/0605212 for an analysis of the SNO data. 14. SNO Collaboration, Astrophys.J., 653 (2006) 1545.
Figure 7. Fit to SNO phase III data with a characteristic neutron spectrum obtained from calibration.
REFERENCES 1. Super-Kamiokande Collaboration, Nucl. Instr. and Meth. A501 (2003) 418-462 . 2. M. Smy, these proceedings. 3. SNO Collaboration, Nucl. Instr. and Meth. A449 (2000) 172. 4. SNO Collaboration, Phys. Rev. C72 (2005) 055502. 5. SNO Collaboration, Phys. Rev. Lett. 89 (2002) 011301. 6. J.N. Bahcall and M.H. Pinsonneault, Phys. Rev. Lett. 92 (2004) 121301. 7. A. Hime, ”Active Solar Neutrino Experiments”, XXII International Conference on Neutrino Physics and Astrophysics, Santa Fe, June 13-19, 2006. 8. Y. Takeuchi, ”Updated solar neutrino results from Super-Kamiokande”, XXXIII International Conference on High Energy Physics, Moscow, July 26-August 2, 2006. 9. Super-Kamiokande Collaboration, Phys. Rev. D73 (2006) 112001. 10. SNO Collaboration, Phys. Rev. D72 (2005) 052010. 11. KamLAND Collaboration, Phys. Rev. Lett. 94 (2005) 081801. 12. Super-Kamiokande Collaboration, Phys. Rev.
SNO & Solar Neutrino Results J. Maneiraa
∗
a
Laborat´orio de Intrumenta¸c˜ao e F´ısica Experimental de Part´ıculas (LIP-Lisboa), Av. Elias Garcia, 14, 1, 1000-149 Lisboa, Portugal
In this talk, the latest results from the presently active solar neutrino experiments, SNO and Super-Kamiokande were reviewed, ranging from 8 B neutrino flux and spectrum measurements and their impact on neutrino oscillation paramenters, to measurements of time variations and searches for hep and relic neutrinos. Prospects for the upcoming results from new analyses or data from new data taking phases were also discussed.
1. Introduction Currently active solar neutrino experiments have measured the flux, spectrum and rate variations in time of the high energy 8 B neutrinos using the water Cherenkov technique. SuperKamiokande (SK)[1] is a 50 kton (22.5 kton fiducial size) water detector located in the Kamioka laboratory in Japan, under an overburden of 2700 mwe. SK detects solar neutrinos via elastic scattering on electrons. From 1996 to 2001, SK took data with 11146 20 inch PMTs and a photo coverage of 40 %, a period referred to as ”SK-I”. After the 2001 incident, the detector was refurbished with 5200 PMTs and took data with a coverage of 19 % from2002 to 2005, the period referred to as ”SK-II”. In July 2006, the detector was rebuilt to its original coverage and the ”SK-III” period started. Here we will report on preliminary analyses of the SK-II data. For a detailed discussion on prospects for SK-III, see [2]. The Sudbury Neutrino Observatory (SNO) [3] is a 1 kton heavy water (D2 O) detector located at a depth of 2 km (6000 mwe) in the INCO Creighton mine near Sudbury, Canada. The heavy water is contained in a 12 m diameter acrylic vessel (AV) that sepearates it from about 5 kton of normal water that serves as a gamma and neutron shield. The AV and heavy water volume are surrounded by an array of 9456 PMTs ∗ On behalf of the SNO collaboration. Thanks to the Super-K collaboration for assistance in the update of their latest results.
0920-5632/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.nuclphysbps.2007.02.061
mounted on a stainless steel support structure. SNO detects solar neutrinos mainly through reactions on the deuteron, but also through elastic scattering with electrons, as in SK: νe + d → p + p + e− νx + d → p + n + νx νx + e− → νx + e−
(CC), (NC), (ES).
While the charged current (CC) reaction is sensitive only to νe ’s, the neutral current reaction (NC) is equally sensitive to all active neutrino flavors (x = e, μ, τ ). The elastic scattering (ES) reaction is sensitive to all electron flavors as well, but with reduced sensitivity to νμ and ντ . The observation of neutrons from the NC reaction provides a model-independent measurement of the total flux of active 8 B solar neutrinos. The neutron counting is carried out in different ways in each of the three phases of SNO: detection of a 6.25 MeV γ ray following capture on a deuteron in the first phase; detection of a 8.6 MeV γ ray cascade following capture on 35 Cl in the second phase, where 2 tonnes of salt were added to the D2 O. In the third phase, the neutrons are detected by an array of 40 3 He counters (Neutral Current Detectors, or NCDs), independently from Cherenkov events. Here we will report on the current and forthcoming analyses of data from phases I and II of SNO, and discuss prospects for phase III.
J. Maneira / Nuclear Physics B (Proc. Suppl.) 168 (2007) 84–89
2.
8
85
B Solar Neutrino Measurements
2.1. Flux measurements The results for the 8 B flux, from the full SNO salt phase dataset[4] are, in units of 106 cm−2 s−1 : ΦCC
+0.08 = 1.68 ± 0.06(stat.)−0.09 (syst.)
ΦN C
+0.38 = 4.94 ± 0.21(stat.)−0.34 (syst.)
ΦSN OES
= 2.35 ± 0.22(stat.) ± 0.15(syst.)
The difference bertween ΦN C and ΦCC shows at a 7σ level that there is a non-νe component in the solar neutrino flux, demonstrating flavor change. These results are in agreement with the previous results from phase I[5], and the NC flux measurement is in good agreement with Solar Standard Model (SSM) predictions[6]. The preliminary result for the 8 B flux from the full SK-II period[7,8], measured with the ES reaction, is, in units of 106 cm−2 s−1 : ΦSKES
=
2.38 ± 0.05(stat.)+0.16 −0.15 (syst.)
These results are in agreement with the more precise results from the SK-I period [9] and with the SNO measurements of the different flavor components of the 8 B flux. 2.2. Spectrum measurements Figure 1 shows the measured spectrum of the ES reaction in SUper-Kamiokande, relative to the predicted spectrum, using the flux and shape from the SSM[6]. The preliminary results for SKII are compatible with measurements from SK-I, even if at a higher threshold and with less precision, and with an undistorted 8 B shape. The pattern of hit PMTs for the gamma cascade from neutron capture in 35 Cl is more smeared than for the single electron CC events, and this allows the signal extraction in the SNO salt phase to use an isotropy parameter (β14 ), in addition to position of the event and direction with respect to the Sun. So the energy spectrum does not need to be used and can be measured for the CC and ES reactions. Figure 2 shows the extracted spectrum of the CC reaction, compared to the SSM and LMA-oscillation predicted shapes. Within the systematic error band (shown in green), the results are compatible both with the standard and oscillated 8 B shape.
Figure 1. Super-Kamiokande elastic scattering spectrum, normalized with respect to the SSM 8 B prediction.
2.3. Global analysis of neutrino oscillations The total flux of 8 B neutrinos from the SNO NC measurement, the spectrum measurements of the CC and ES reactions in SNO and SK-I (day and night spectra considered separately) were included in a global two-flavor neutrino oscillation analysis, together with the radiochemical solar neutrino measurements and rate and shape predictions from the Solar Standard Model. The data from the KamLAND reactor experiment[11] were also included. Figure 3 shows the allowed regions from this oscillation analysis. The best-fit point is: Δm2 θ
−5 = 8.0+0.6 eV 2 −0.4 × 10
= 33.9+2.4 −2.2 deg
Only the so-called LMA-I region is allowed by the current data, and maximal mixing is excluded at more than 4 σ. 2.4. Searches for time variation of the event rate Both Super-Kamiokande and SNO have performed analyses of the time variations of their signal rates, with sensitivity to periodicities of very different magnitudes. A difference between the average day and night rates would be evidence for
J. Maneira / Nuclear Physics B (Proc. Suppl.) 168 (2007) 84–89
300 Data
6 m2 (10 eV 2)
Systematic uncertainties
250
SSM 8B model shape LMA 8B model shape
200
20 68% CL
-5
Events/(0.5 MeV)
86
150
100
95% CL
15
99.73% CL
10
50
0
6
7
8
9
10
11
12
13
T eff (MeV)
5 0
Figure 2. Extracted CC energy spectrum compared to that predicted with the best-fit LMA parameters. Tef f is the effective electron kinetic energy.
an oscillation-in-matter induced regeneration of electron neutrinos when crossing the Earth. This effect is predicted at the few % level for a part of the allowed LMA oscillation parameter region. Variations with longer time scales, in particular those correlated with the solar cycles, could be due to neutrino spin-flip conversion in the magnetic fields of the Sun. The search for day/night rate variations is parametrized with the day/night assymetry: ADN
=
2(Day − N ight) (Day + N ight)
The results for phases I and II of SK (preliminary for SK-II, systematics still under study) and combined phase I and II of SNO are ADN (SKI) = −0.021 ± 0.020(stat.)+0.013 −0.012 (syst.) ADN (SKII) = −0.064 ± 0.043(stat.) ADN (SN O)
= 0.037 ± 0.040(stat. + syst.),
so there is no evidence of a day-night asymmetry in the present data from both experiments. Figure 4 shows the variation of the SK event rate in funcion of time, including also the preliminary analysis of the more recent SK-II dataset.
0.2
0.4
0.6
0.8
1 tan2e
Figure 3. Allowed region of neutrino oscillation parameters from a global analysis of all solar neutrino data, combined with the results of the KamLAND reactor neutrino experiment.
Figure 5 shows the variation of the signal rate in SNO (relative to the average rate), but in function of the time since the perihelion, so all data in folded in the same one-year range. In both cases, the expected rate variation due to the eccentricity of the Earth orbit is observed (and SNO fits the eccentricity to be 0.0143 ± 0.0086, compared to 0.0167). More sophisticated periodicity analyses were performed on the datasets of both experiments by searching for peaks in a frequency spectrum with the Lomb-Scargle periodogram technique [12,10]. A large number of Monte Carlo simulated datasets were generated assuming no variation beyond eccentricity, and for SK, 19% of those (35% for SNO) gave a peak at least as large as the largest peak observed in the data. Independent analyses of the binned SK and SNO data were also performed (see, for instance[13], also finding no variations beyond what is expected from statistical fluctuations, so there is no evidence for time modulation in the SK and SNO data.
87
Relative rate
J. Maneira / Nuclear Physics B (Proc. Suppl.) 168 (2007) 84–89
1.1
1.05 1
0.95 0.9 0
50
100
150
200
250
300
350
Days since perihelion
Figure 5. Relative SNO event rate as a function of days since perihelion, normalized to the mean rate. The solid line represents the expected variation due to the eccentricity of the Earth orbit. Figure 4. Super-Kamiokande event rate in function of time. The solid line represents the expected variation due to the eccentricity of the Earth orbit.
3. Searches for hep neutrinos In addition to 8 B neutrinos, the SSM predicts a high energy component of the solar neutrino spectrum that can, in principle, be detected by water Cherenkov detectors. These are the hep netrinos that have a predicted flux about 600 times smaller than 8 B, but a higher endpoint, at 18.77 MeV. In the energy region between the endpoint of the 8 B neutrinos and the rise of the atmospheric neutrinos, the background is small enough to allow a search for the hep neutrinos. Both SK[9] and SNO[14] have performed that analysis. SK has chosen the energy interval 18-21 MeV that maximizes the hep/8 B ratio based on Monte Carlo simulations. The observation of 4.9 events, when 1.06 are expected from the hep signal and 1.72 from 8 B, places the upper limit on the hep flux at 7.3×104cm−2 s−1 (90% C.L.), assuming all hep neutrinos are electron-type. Applying a correction for the effect of oscillations, this is equivalent to a limit on the total flux of hep neutrinos of 1.5 × 105 cm−2 s−1 . The most sensitive energy interval was chosen for the SNO analysis by evaluating the expected signal and background levels, before examining the data, and the limits were set using a modi-
fied Felman-Cousins technique. The systematics of the energy response, important for the prediction of the 8 B background, was evaluated from the analysis of three different calibration sources: 16 N (6.13 MeV gamma), pT (3 H(p, γ)4 He, 19.8 MeV gamma source), and Michel electrons (from cosmic muon decay). The background contribution from atmospheric neutrinos, and the effect of neutrino oscillations were considered. In the selected energy interval of 14.3-20 MeV, 0.99 signal and 3.1 background events were expected. Two events were observed, placing a limit on the total flux of hep neutrinos at 2.3 × 104 cm−2 s−1 , that is, 2.9× the SSM prediction and an improvement of ×6.5 over the previous limit. 4. Outlook 4.1. Low energy threshold analysis Since the analysis published in [4], several improvements to the Monte Carlo and analysis were developed, mainly to the energy reconstruction, with the aim of improving the uncertainties on energy scale and resolution, reduce the low energy background tails and reduce the energy threshold for signal extraction. The analysis of the optical calibration data allowed the determination of the relative efficiencies of a large fraction ( 70%) of the PMTs and the use of these parameters both in the Monte Carlo and energy estimators allowed a better modelling of the energy resolution. A new energy estimator using both prompt and late
Events / 1.425 MeV
88
J. Maneira / Nuclear Physics B (Proc. Suppl.) 168 (2007) 84–89 20 18 16
Data Estimated Background
14
Predicted hep signal
12 10
hep signal box
8 6 4 2 0
13
14
15
16
17
18
19
20
21
Teff (MeV)
Figure 6. Distribution of events in the region of the 8 B endpoint. Also shown are the estimated number of background events, including systematic uncertainties, and the standard model prediction for the hep signal.
light was developed and the previous estimator was also improved. Both show a spread of about 0.5% on the difference between data and MC over the radial position of the 16 N calibration source, compared to a spread of 0.7% in previous analyses. The joint analysis of the phase I and II data is still underway but an energy threshold of 4.0 MeV (in Tef f ) appears possible. 4.2. Prospects for Phase III In Phase III of SNO, that started in the Summer of 2004 and ended November 2006, an array of 36 proportional counters filled with 3 He (plus 4 counters filled with 4 He for background measurement) was installed in the D2 Ovolume. These counters, called Neutral Current Detectors (NCDs) detect neutrons through the process 3 He(n, p)3 H, releasing a proton and a tritium nucleus with Q-value of 764 keV. The counters are about 10 m long and have a diameter of 5 cm. They are attached to the bottom of the SNO
acrylic vessel on a grid of 1 × 1 m. The neutrons will preferentially capture on the 3 He, so that the detection of Neutral Current events is done in the NCDs on an event-by-event basis, independently from CC and ES Cherenkov signals seen by the PMT array. An extensive program of in-situ and ex-situ calibrations was carried out during phase III of the experiment, to measure the response of the array to both the neutron signal and the backgrounds, mainly from α particles. Figure 7 shows a fit to SNO phase III data with a characteristic neutron spectrum obtained from calibration. The NCD electronics allows the digitization of the signal pulse and the shape of that pulse is used to distinguish neutrons from background, and reconstructing position of the event. The Monte Carlo and analysis of the PMT array data was significantly updated for phase III, to take into the account the optical effects of shadows and reflections of the Cherenkov on the proportional counters. The method for the optical calibration of SNO was based on normalizing the PMT occupancy of an off-center laser run by the occupancy of the same PMT in a central run (in order to cancel the PMT efficiency from the occupancy model prediction). Now, in the third phase, in order to maintain the simplicity of the calibration model and reject partially shadowed PMTs, this method resulted in a significant statistics loss. So a new method was developed, using estimates of the relative PMT efficiencies, and correcting for the effect of reflections. In all other analyses partially shadowed PMTs are not rejected, so the loss of light due to the NCDs is not large, and is mostly compensated by upgrades to the electronics that were made between phases II and III, that allow the PMT thresholds to be set lower. The NC measurement in the salt phase of SNO is expected to have smaller and significantly different systematic errors than the NC measurement in the previous phases. The CC measurement is also expected to improve since the large correlation with NC in the signal extraction in previous phases is now almost eliminated.
J. Maneira / Nuclear Physics B (Proc. Suppl.) 168 (2007) 84–89
89
D68 (2003) 092002. 13. G. Ranucci, Phys. Rev. D73 (2006) 103003. See hep-ph/0605212 for an analysis of the SNO data. 14. SNO Collaboration, Astrophys.J., 653 (2006) 1545.
Figure 7. Fit to SNO phase III data with a characteristic neutron spectrum obtained from calibration.
REFERENCES 1. Super-Kamiokande Collaboration, Nucl. Instr. and Meth. A501 (2003) 418-462 . 2. M. Smy, these proceedings. 3. SNO Collaboration, Nucl. Instr. and Meth. A449 (2000) 172. 4. SNO Collaboration, Phys. Rev. C72 (2005) 055502. 5. SNO Collaboration, Phys. Rev. Lett. 89 (2002) 011301. 6. J.N. Bahcall and M.H. Pinsonneault, Phys. Rev. Lett. 92 (2004) 121301. 7. A. Hime, ”Active Solar Neutrino Experiments”, XXII International Conference on Neutrino Physics and Astrophysics, Santa Fe, June 13-19, 2006. 8. Y. Takeuchi, ”Updated solar neutrino results from Super-Kamiokande”, XXXIII International Conference on High Energy Physics, Moscow, July 26-August 2, 2006. 9. Super-Kamiokande Collaboration, Phys. Rev. D73 (2006) 112001. 10. SNO Collaboration, Phys. Rev. D72 (2005) 052010. 11. KamLAND Collaboration, Phys. Rev. Lett. 94 (2005) 081801. 12. Super-Kamiokande Collaboration, Phys. Rev.