- Email: [email protected]
The Pierre Auger Observatory Upgrade Program
Available online at www.sciencedirect.com
Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160 www.elsevier.com/locate/nppp
The Pierre Auger Observatory Upgrade Program C. Di Giulioa , for the Pierre Auger Collaborationb,1 a INFN
Roma Tor Vergata, Via della Ricerca Scientifica 1, 00133 Roma Pierre Auger, Av. San Mart´ın Norte, 304, 5613 Malarg¨ue, Argentina
b Observatorio
Abstract The Pierre Auger Observatory has been taking data since 2004 and with the complete set of detectors since 2008, with an efficiency above 95%, with very low downtime. Our accurancy in the results have had a deep impact on the area of ultra-high energy cosmic rays and new questions and open challenges call for an upgrade of the Observatory. The planned detector upgrade is presented and the expected performance and improved physics sensitivity of the upgraded Auger Observatory are discussed. Keywords: astroparticle physics, cosmic rays, instrumentation: detectors,
1. Introduction In the last ten years the Pierre Auger Observatory [1] has been addressing the most fundamental questions about the nature and origin of the highest-energy cosmic rays. The extremely low flux of the ultra-high energy cosmic rays (UHECRs) requires that the Observatory covers an overall area of 3000 km2 to collect enough statistics. Using simultaneously two complementary techniques to detect the extensive air showers they induce in the atmosphere, the Pierre Auger Observatory has reduced significantly the systematics on the energy scale [2]. The 1600 water Cherenkov surface detectors (SD) are deployed on a triangular grid of 1500 m spacing and measure the lateral distribution of air-shower particles on the ground. The 27 fluorescence telescopes overlook the atmosphere above the SD from four sites located on the periphery of the array. They measure the longitudinal profile of the number of shower particles, through the nitrogen fluorescence light emitted along the shower axis. Email address: [email protected] (C. Di Giulio) 1 Full author list: http://www.auger.org/archive/authors 2015 09.html
http://dx.doi.org/10.1016/j.nuclphysbps.2016.10.022 2405-6014/© 2016 Elsevier B.V. All rights reserved.
In this report, we first present a brief summary of some important Auger results, then we make the case to motivate an upgrade to the Observatory. Following this, the proposed scintillator detector is described, representing the upgrade of the water Cherenkov detector, and the underground muon detector is described as well. Finally, the upgrade performance is also illustrated in detail.
2. Pierre Auger Observatory results Unquestionable evidence for a cosmic-ray flux suppression above 50 EeV is observed after 10 years of the Auger operation as shown in figure 1 [3]. Strong limits have been placed on the photon and neutrino components of the flux indicating that top-down source processes, such as the decay of super-heavy particles, cannot account for a significant part of the observed particle flux [4, 5]. The spectrum and the Xmax data together favor a scenario where the suppression is partly a source effect [6] rather than the GZK process. The limited data on Xmax above 40 EeV is available where the trend towards heavier composition appears [7, 8], and this is due to the low duty cycle of the fluorescence detection
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
154
E [eV] 1018
1019
1020
E 3J(E) eV2 km−2 sr−1 yr−1
[ ] - flux σsys
10
σsys [E]/E = 14 %
38
To reach these aims with high statistics and mass composition discrimination capabilities the Auger Collaboration is proposing an upgrade program with additional detectors and an extension of its operation until 2024. 4. Auger PRIME: the upgrade program
1037
1036 17.5
SD-1500 m vertical SD-750 m vertical SD-1500 m inclined Hybrid
18.0
18.5
19.0
19.5
20.0
20.5
log10(E/eV)
Figure 1: The Auger energy spectra obtained from hybrid events, from SD data obtained with the 1500 m array for vertical and inclined events (SD-1550), and from vertical events with the infilled SD array (SD-750).
technique. The SD has the statistic but it’s not able to discriminate the composition on event-by-event basis. A detailed report on the Auger Collaboration results was also presented at this conference [9] and therefore will not be repeated here.
The main upgrade will be the Surface Scintillator Detector (SSD). It consists in the installation of 4 m2 scintillator detectors on top of the existing water-Cherenkov surface detectors, providing the capability to distinguish between the electromagnetic and muonic components of air showers, which will provide information about the composition of the primary particles. The upgrade also includes the substitution of the current electronics allowing for an improved performance, the completion of the array of the AMIGA muon detectors, and the extension of the duty cycle of the fluorescence telescopes by allowing measurements during nights with more background light. A detailed description of these components of the upgrade program appears in the following subsections. 4.1. The Surface Scintillator Detector
3. Motivations for an upgrade The Auger Collaboration upgrade program called Auger PRIME: (Primary cosmic Ray Identification with Muons and Electrons) has defined the following aims for the upgrade proposal: • to clarify the mass composition and the origin of the flux suppression at the highest energies, i.e. the differentiation between the energy loss effects due to propagation, and the maximum energy of particles injected by astrophysical sources; • to reach a sensitivity to a contribution of protons as small as 10% in the flux suppression region; this will be crucial to predict the flux of secondary gamma rays and neutrinos due to proton interactions for future cosmic-ray, neutrino, and gammaray detectors; • to investigate extensive air showers and hadronic multiparticle production, including the exploration of fundamental particle physics at energies today not accessible with accelerators, studying the constraints on new physics phenomena, such as Lorentz invariance violation or extra dimensions.
Figure 2: 3D view of a Surface Scintillator Detector mounted on a water-Cherenkov detector.
The SSD provides the complementary measurement and is made of a plastic scintillator plane mounted on the existing water-Cherenkov detectors (WCDs) as shown in figure 2. The shower particles will be sampled by the two detectors that respond in different ways to muons and electromagnetic particles as shown for the SSD in figure 3 and for the WCD in figure 4.
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160 4000
Covering the Pierre Auger Observatory area of the surface array, the SSD is designed to be simple, reliable and easily deployed.
155
Real data (scaled) Simulation
3500 3000 Counts
2500
G
2000 1500 1000 500 0
0
100
200
300
400
500
Charge [ADC.bin]
Figure 6: MIP signal from a 2 m2 prototype installed in the Auger array compared with simulations for one minute of data taking. Figure 3: Typical SSD signal response to the electromagnetic and muonic components of the shower.
G
Figure 4: Typical WCD signal response to the electromagnetic and muonic components of the shower.
Less than 5% deviation is expected from uniform detector response over the area of the scintillator. The WCD triggers the SSD providing a clear minimumionizing particle (MIP) signal separated from the background, see figure 6. An engineering array of 10 detectors is scheduled to be installed at the Auger site in 2016. 4.2. The SSD methodology The reconstruction of the muonic shower component is the key element to show the discrimination power for different primary particles. The number of muons Nμ is the least model dependent and most direct composition-sensitive observable that can be obtained by the upgrade of the detector array. The muonic signal in the WCD Sμ,WCD in each detector station can be derived on a station-by-station basis as described in [10], thanks to the signal responses in the SSD (SS S D ) and the WCD (SWCD ) to particles of the electromagnetic and muonic shower components: S μ,WCD = aS WCD + bS S S D ,
Figure 5: SSD schematic view of the two scintillator modules, embedded within an extruded polystyrene foam vessel. The dimensions of the vessel are quoted in the figure.
An SSD unit consists of a box of 3.8 m × 1.3 m, housing two scintillator modules each 1 cm thick and covering an area of 1.9 m2 , see figure 2. In the SSD a double roof, made of corrugated aluminium, is used to reduce the temperature variations. The scintillators are read out by wavelength-shifting fibers guiding the light of the two modules to a photomultiplier tube (PMT) as shown in figure 5.
(1)
where the signals are measured in units of the response to a vertical equivalent muon (VEM) or minimum ionizing particle, respectively. The a and b parameters are derived from detector simulations and have only a very weak dependence on the primary composition and lateral distance from the shower core, as in figure 7 where the bias on the muonic reconstructed signal and the effect on the resolution are shown. Restricted by the limited size of an individual detector station, this method is subject to large fluctuations and is only well suited for deriving mean muon numbers, see figure 8.
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
156
Fitting the lateral distribution function (LDF) is possible to improve the performance of the reconstruction of the muonic signal. Eq. 1 is then applied to the LDF values for the WCD and SSD to calculate the muonic signal at 800 m core distance, Sμ (800). For example at an energy of E=1019.8 eV at θ = 38◦ the reconstruction resolutions of the muonic signal are: σ(S μ (800)) = 22% for protons and 14% for iron. S μ (800) Defining the merit factor as the measure of discrimination power for separating primary i and j using the corresponding observables Si and S j (with the RMS σ) as: S i − S j (2) f MF = σ(S i ) 2 + σ(S j ) 2
Proton Iron
Figure 7: Reconstruction bias (solid symbols) and resolution (open symbols) of the muonic signal contribution for individual detector stations, as a function of distance r from the shower core. The results for proton and iron showers are shown in red and blue, respectively.
using Sμ (800) as a composition estimator, the merit factor for distinguishing between proton and iron primaries is above 1.5 at shower energies larger than 1019.5 eV and small zenith angles.
. 4.3. The water-Cherenkov detector upgrade
0.3
(Nμ, rec- Nμ)/Nμ
p He
0.2
N Fe
0.1 0
Figure 9: First prototype of a SSD scintillator on top of a WCD.
-0.1 -0.2 -0.3 18.8
19
19.2 19.4 19.6 19.8
20
log10(E/eV)
Figure 8: Number of muons Nμ reconstructed (Nμ,rec ) for individual showers using shower universality, compared with the true Nμ as a function of energy for different primary species. Error bars represent the RMS of the distributions.
The WCD is a rotomolded polyethylene tank filled with purified water that produces Cherenkov light when crossed by energetic charged particles associated with cosmic-ray showers. A flexible Tyvek liner inside the tank provides an interface between the water volume and the light sensors that collect Cherenkov light (PMTs) and the light sources that produce calibration pulses (LEDs). Access to the liner, the PMTs and the LEDs is through three hatches located on the top of the tank. An electronics box containing front-end charge amplifiers, shapers, trigger logic, signal buffers, power control, radio transmitter and receivers is located on the top roof of the tank on one of the hatch-covers and is protected by a dome. All the cables connecting the
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
electronics and the light sensors run inside the tank and connect to the electronics via feed-throughs in the hatch covers. Figure 9 shows a picture of an operating WCD. The Surface Detector Electronics (SDE) records the tank signals, makes local triggering decisions, sends timestamps to the central data acquisition system for the global triggers, and stores event data for retrieval when a global trigger condition is satisfied. Because of the small bandwidth (1200 bits/s) available to each tank, the station must operate semi-autonomously, performing calibrations and taking action in response to alarm conditions at the station level. The current SDE was designed 15 years ago using the technology available at that time. Evolution in processors, power consumption of electronics components, and timing systems makes it possible today to design and implement a higher performance electronics system for the Surface Detector array. Furthermore, the proposed electronics provides an interface to allow the scintillator detectors co-located with the surface detector stations to make use of the data processing and communications infrastructure of the stations. The SD stations will be upgraded with new electronics that will process both WCD and SSD signals. Use of the new electronics also aims to increase the data quality (with faster sampling of ADC traces, better timing accuracy, increased dynamic range), to enhance the local trigger and processing capabilities (with a more powerful local station processor and FPGA) and to improve calibration and monitoring capabilities of the surface detector stations. The signals of the WCDs and SSDs will be sampled synchronously at a rate of 120MHz (previously 40MHz) and the GPS timing accuracy will be better than 5ns. The dynamic range of the WCDs will be enhanced by a factor of 32 due to the new electronics and an additional small PMT that will be inserted in one of the filling ports. First prototypes of the new electronics have been built and are being tested. Both the SSDs and the electronics upgrade can be easily deployed, and will have only minimal impact on the continuous data taking of the Observatory. 4.4. The Underground Muon Detector The Underground Muon Detector (UMD) will provide a direct measurement of the muon content of a sub-sample of showers observed by the upgraded Auger surface detector. In the upgrade plan this serves as verification and fine-tuning of the methods used to extract shower muon content using the SSD and WCD stations. The UMD will therefore consist of 61 AMIGA muon detectors deployed on a 750 m grid in the infill area of the Surface Detector, instrumenting a total area of 23.5
157
km2 . The use of the AMIGA muon detectors in this verification role is additional to the rich physics investigations planned for AMIGA in the ankle-region of the energy spectrum[11].
Figure 10: Deployment of a unitary cell AMIGA station with two 10 m2 and two 5 m2 modules at a depth of 2.3 m underground. The big tubes are to provide access to the electronics for development and maintenance purposes, to be replaced by tubes of 31 cm diameter for AMIGA production.
. Each muon detector station will have an area of 30 m2 and will be buried at the side of a surface detector station at a depth of approximately 1.3 m, see figure 10. The distance to the station will be large enough to avoid shadowing from the water tank, guaranteeing uniform shielding, but small enough to represent the same physical point inside the shower front, and allowing shared use of GPS time signals and telecommunications with its associated surface detector. The baseline design for the muon detectors uses the same extruded plastic scintillators already developed and used by the MINOS experiment [12]. They will work as counters (i.e. signals above a tunable threshold are counted) having an appropriate segmentation to prevent pile-up, together with an integrated signal for large muon densities. The light collected in a scintillator will be guided towards a 64 channel Hamamatsu PMT. An AMIGA prototype is displayed in figure 11. It consists of 64× 4 m long scintillator strips. The strips are 4.1 cm wide and 1.0 cm thick and therefore the detector modules have an active area of 10 m2 . The strips have a middle groove which accommodates a wavelength shifter (WLS) fiber of 1.2 mm diameter which is glued into the groove and covered with reflective foil. The scintillator is co-extruded with a TiO2 reflective coating which prevents light from leaving the material.
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
Figure 11: An AMIGA 10 m2 module being manufactured. The black optical connector concentrates the 64 optical fibers coming from the 32 scintillator bars at each side.
. Muon counting and the digitization of the integrated signal is implemented in the AMIGA electronics, including an FPGA with three main functional blocks: counting, data codification, and external communications. The data are sent to the central data acquisition system of the observatory (CDAS) using an independent, commercial radio communications system. 4.5. The fluorescence detector operation upgrade The fluorescence detector provides exceptional information about extensive air showers such as a modelindependent energy reconstruction and direct measurement of the longitudinal development profiles. The main limitation of the FD is its duty cycle, currently below the 15%, see figure 12. Our goal is to increase the exposure for cosmic-ray events above 1019 eV by extending the FD measurement into hours with high night sky background (NSB). The current setup allows this novel operation and we have performed several tests that successfully demonstrate that it is feasible. Safety limits on the long and short term illumination of PMTs by the NSB, and particularly scattered moonlight, define the data taking period of the current FD operation. The duty cycle is therefore limited to about 19%, which is reduced to 15% by bad weather conditions, power cuts and malfunctions. A significant increase of the duty cycle is possible by the extension of the FD operation to times at which a large fraction of the moon in the sky is illuminated. However, during such operations the PMT gain must be reduced by lowering the supplied high voltage (HV) to avoid an excessively
high anode current leading to an irreversible deterioration of the PMT sensitivity. The HV power supplies installed in the FD buildings allow switching between two high voltage levels and the PMTs can be operated at the nominal gain and a lower gain. The first test measurement outside the standard FD data taking period was performed with one FD telescope in the austral autumn of 2015. Telescope 1 at Los Leones was operated during six nights with a highly illuminated moon above the horizon. We have confirmed that the HV change can be done remotely and that the HV stabilizes within a few seconds after its change. The PMT performance can be monitored with the existing calibration setup in exactly the same way as during the standard data taking, so no change in the system is required. By extending the FD operation to higher NSB we can gain up to 40% more events above 1019.5 eV. This will allow us to improve the cross check of results of the upgraded SD array with the FD up to the energy of the suppression in the cosmic-ray flux. 15 14 FD uptime fraction [%]
158
13 12 11 10 9 8
2
4
6
8
10
12
14
16
18
20
22
24
Telescope
Figure 12: Fluorescence telescopes uptime fractions. The colors indicate the 4 sites of the telescopes: blue Los Leones, magenta Los Morados, yellow Loma Amarilla and green Coihueco [13].
5. Auger PRIME expected performance It is complicated to demonstrate the potential of Auger PRIME without knowing what composition to expect in the GZK suppression region [16]. Therefore we show the discrimination power of the additional information obtained with the Auger upgrade in two physics motivated benchmark models. Those are obtained by fitting the Auger flux [14] and composition data [8] for E > 1018.7 eV. In figures 13 and 14 the energy spectra of the two scenarios are compared with that measured by Auger [6].
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
159
To estimate the performance on the mass composition discrimination mock data sets were generated for these two scenarios with a statistics corresponding to 7 years of data taking with Auger PRIME.
[g/cm2]
38
840
-2
-1
-1
E J [eV km sr yr ]
10
3
2
820
10
37
10
36
p, EPOS-LHC Fe, EPOS-LHC
800 780
18
18.5
19
19.5
20
20.5
log (E/eV) 10
Figure 13: Benchmark spectra chosen as representations of a maximum-rigidity scenario [6]. The index of the injection energy spectrum at the sources is about -1. The colors for the different groups are proton-blue, helium-gray, nitrogen-green, and iron-red.
760 740 720 700
Scenario 1
19
19.2
19.4
19.6
Scenario 2
19.8
20
lg(E/eV)
Figure 15: Reconstructed mean depth of shower maximum Xmax with the two scenarios 1 maximum rigidity model and 2 photodisintegration model.
38
10
E J [eV km sr yr ]
3
2
-2
-1
-1
The mean Xmax , RMS (Xmax ), and the relative muon number Rμ are shown in figures 15, 16 and 17 respectively. The RMS contains the intrinsic air-shower fluctuations and the reconstruction resolution. The models predict significantly different extrapolations into the GZK suppression region, while the mean Xmax , RMS (Xmax ) and Rμ are very similar up to 1019.2 eV that is the energy range that is well covered by data of the fluorescence telescopes [7]. The differences between the two scenario are well reproduced with the reconstructed Xmax , RMS (Xmax ) and Rμ and the two scenarios can be distinguished with high significance and statistics.
37
10
36
10
18
18.5
19
19.5
20
20.5
log (E/eV) 10
Figure 14: Benchmark spectra chosen as representations of one photodisintegration scenario [15]. The index of the injection energy spectrum at the sources is about -2. The colors for the different groups are proton-blue, helium-gray, nitrogen-green, and iron-red.
6. Conclusions The Pierre Auger Observatory data until now do not lead to a definitive answer to the origin of the cosmic rays. The Auger upgrade promises high-quality future data, and real scope for new physics uses of existing events. With operation planned from 2018 until 2024, event statistics will more than double compared with the ex-
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
160
RMS(Xmax) [g/cm2]
90
p, EPOS-LHC
80
Fe, EPOS-LHC
70 60 50 40 30 Scenario 1
20 10 18.8
19
Scenario 2
19.2 19.4 19.6 19.8
20
lg(E/eV)
Figure 16: Shower maximum Xmax fluctuation shown in the two scenarios.
isting Auger data set, with the critical added advantage that every event will now have mass information. This will allow us to address some of the most pressing questions in UHECR physics, including that of the origin of the flux suppression, the prospects of light particle astronomy and secondary particle fluxes, and the possibility of new particle physics at extreme energies. Obtaining additional composition-sensitive information will not only help to better reconstruct the properties of the primary particles at the highest energies, but also improve the measurements in the important energy range just above the ankle. Furthermore, measurements with the new detectors will help to reduce systematic uncertainties related to modeling hadronic showers and to limitations of reconstruction algorithms. This improved knowledge of air-shower physics will likely then also allow a re-analysis of existing data for improved energy assignments, for mass composition studies, and for photon and neutrino searches. Finally it should be mentioned that the addition of scintillator detectors across the entire Observatory will also make possible direct comparisons of Auger measurements with those of the surface detectors of the Telescope Array experiment. References
0.25 0.2
0.15
p, EPOS-LHC Fe, EPOS-LHC
0.1 0.05 0 -0.05 -0.1 -0.15
Scenario 1
-0.2 -0.25
Scenario 2
19
19.2
19.4
19.6
19.8
20
lg(E/eV)
Figure 17: Mean number of muons at 38 degree shown in the two scenarios. The number of muons is given relative to that expected for an equal mix of p-He-CNO-Fe as primary particles.
[1] Aab, A., et al. (The Pierre Auger Collaboration) NIM A 798 (2015) 172-213 [2] Verzi, V. et al. (The Pierre Auger Collaboration), Proc of 33rd Int. Cosmic Ray Conf., Rio de Janeiro (2013) [arXiv:1307.5059 [astro-ph.HE]] [3] Vali˜no, I. for the Pierre Auger Collaboration: 2015, PoS(ICRC2015)271 [4] Bleve, C. for the Pierre Auger Collaboration: 2015, PoS(ICRC2015)1103 [5] Aab, A., et al. (The Pierre Auger Collaboration): 2015, PhRvD 91, 092008 [6] Di Matteo, A. for the Pierre Auger Collaboration: 2015, PoS(ICRC2015)249 [7] Aab, A., et al. (The Pierre Auger Collaboration): 2014, PhRvD 90, 122005 [8] Aab, A., et al. (The Pierre Auger Collaboration): 2014, PhRvD 90, 122006 [9] Coutu, S. for the Pierre Auger Collaboration : this volume [10] Letessier-Selvon, A., et al., NIM A 767 (2014) 41-49 [11] Suarez, F. for the Pierre Auger Collaboration, [arXiv:1307.5059 [astro-ph.HE]]. [12] Michael D.G., et al., NIM A 596 (2008) 190-228 [13] Abraham, J., et al. (The Pierre Auger Collaboration) NIM A 620 (2010) 227-251 [14] Aab, A., et al. (The Pierre Auger Collaboration), Proc of 33rd Int. Cosmic Ray Conf., Rio de Janeiro (2013) [arXiv:1307.5059 [astro-ph.HE]] [15] D. Allard, Astropart. Phys. (2012) 39-40 [16] Engel, R. for the Pierre Auger Collaboration: 2015, PoS(ICRC2015)685
Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160 www.elsevier.com/locate/nppp
The Pierre Auger Observatory Upgrade Program C. Di Giulioa , for the Pierre Auger Collaborationb,1 a INFN
Roma Tor Vergata, Via della Ricerca Scientifica 1, 00133 Roma Pierre Auger, Av. San Mart´ın Norte, 304, 5613 Malarg¨ue, Argentina
b Observatorio
Abstract The Pierre Auger Observatory has been taking data since 2004 and with the complete set of detectors since 2008, with an efficiency above 95%, with very low downtime. Our accurancy in the results have had a deep impact on the area of ultra-high energy cosmic rays and new questions and open challenges call for an upgrade of the Observatory. The planned detector upgrade is presented and the expected performance and improved physics sensitivity of the upgraded Auger Observatory are discussed. Keywords: astroparticle physics, cosmic rays, instrumentation: detectors,
1. Introduction In the last ten years the Pierre Auger Observatory [1] has been addressing the most fundamental questions about the nature and origin of the highest-energy cosmic rays. The extremely low flux of the ultra-high energy cosmic rays (UHECRs) requires that the Observatory covers an overall area of 3000 km2 to collect enough statistics. Using simultaneously two complementary techniques to detect the extensive air showers they induce in the atmosphere, the Pierre Auger Observatory has reduced significantly the systematics on the energy scale [2]. The 1600 water Cherenkov surface detectors (SD) are deployed on a triangular grid of 1500 m spacing and measure the lateral distribution of air-shower particles on the ground. The 27 fluorescence telescopes overlook the atmosphere above the SD from four sites located on the periphery of the array. They measure the longitudinal profile of the number of shower particles, through the nitrogen fluorescence light emitted along the shower axis. Email address: [email protected] (C. Di Giulio) 1 Full author list: http://www.auger.org/archive/authors 2015 09.html
http://dx.doi.org/10.1016/j.nuclphysbps.2016.10.022 2405-6014/© 2016 Elsevier B.V. All rights reserved.
In this report, we first present a brief summary of some important Auger results, then we make the case to motivate an upgrade to the Observatory. Following this, the proposed scintillator detector is described, representing the upgrade of the water Cherenkov detector, and the underground muon detector is described as well. Finally, the upgrade performance is also illustrated in detail.
2. Pierre Auger Observatory results Unquestionable evidence for a cosmic-ray flux suppression above 50 EeV is observed after 10 years of the Auger operation as shown in figure 1 [3]. Strong limits have been placed on the photon and neutrino components of the flux indicating that top-down source processes, such as the decay of super-heavy particles, cannot account for a significant part of the observed particle flux [4, 5]. The spectrum and the Xmax data together favor a scenario where the suppression is partly a source effect [6] rather than the GZK process. The limited data on Xmax above 40 EeV is available where the trend towards heavier composition appears [7, 8], and this is due to the low duty cycle of the fluorescence detection
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
154
E [eV] 1018
1019
1020
E 3J(E) eV2 km−2 sr−1 yr−1
[ ] - flux σsys
10
σsys [E]/E = 14 %
38
To reach these aims with high statistics and mass composition discrimination capabilities the Auger Collaboration is proposing an upgrade program with additional detectors and an extension of its operation until 2024. 4. Auger PRIME: the upgrade program
1037
1036 17.5
SD-1500 m vertical SD-750 m vertical SD-1500 m inclined Hybrid
18.0
18.5
19.0
19.5
20.0
20.5
log10(E/eV)
Figure 1: The Auger energy spectra obtained from hybrid events, from SD data obtained with the 1500 m array for vertical and inclined events (SD-1550), and from vertical events with the infilled SD array (SD-750).
technique. The SD has the statistic but it’s not able to discriminate the composition on event-by-event basis. A detailed report on the Auger Collaboration results was also presented at this conference [9] and therefore will not be repeated here.
The main upgrade will be the Surface Scintillator Detector (SSD). It consists in the installation of 4 m2 scintillator detectors on top of the existing water-Cherenkov surface detectors, providing the capability to distinguish between the electromagnetic and muonic components of air showers, which will provide information about the composition of the primary particles. The upgrade also includes the substitution of the current electronics allowing for an improved performance, the completion of the array of the AMIGA muon detectors, and the extension of the duty cycle of the fluorescence telescopes by allowing measurements during nights with more background light. A detailed description of these components of the upgrade program appears in the following subsections. 4.1. The Surface Scintillator Detector
3. Motivations for an upgrade The Auger Collaboration upgrade program called Auger PRIME: (Primary cosmic Ray Identification with Muons and Electrons) has defined the following aims for the upgrade proposal: • to clarify the mass composition and the origin of the flux suppression at the highest energies, i.e. the differentiation between the energy loss effects due to propagation, and the maximum energy of particles injected by astrophysical sources; • to reach a sensitivity to a contribution of protons as small as 10% in the flux suppression region; this will be crucial to predict the flux of secondary gamma rays and neutrinos due to proton interactions for future cosmic-ray, neutrino, and gammaray detectors; • to investigate extensive air showers and hadronic multiparticle production, including the exploration of fundamental particle physics at energies today not accessible with accelerators, studying the constraints on new physics phenomena, such as Lorentz invariance violation or extra dimensions.
Figure 2: 3D view of a Surface Scintillator Detector mounted on a water-Cherenkov detector.
The SSD provides the complementary measurement and is made of a plastic scintillator plane mounted on the existing water-Cherenkov detectors (WCDs) as shown in figure 2. The shower particles will be sampled by the two detectors that respond in different ways to muons and electromagnetic particles as shown for the SSD in figure 3 and for the WCD in figure 4.
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160 4000
Covering the Pierre Auger Observatory area of the surface array, the SSD is designed to be simple, reliable and easily deployed.
155
Real data (scaled) Simulation
3500 3000 Counts
2500
G
2000 1500 1000 500 0
0
100
200
300
400
500
Charge [ADC.bin]
Figure 6: MIP signal from a 2 m2 prototype installed in the Auger array compared with simulations for one minute of data taking. Figure 3: Typical SSD signal response to the electromagnetic and muonic components of the shower.
G
Figure 4: Typical WCD signal response to the electromagnetic and muonic components of the shower.
Less than 5% deviation is expected from uniform detector response over the area of the scintillator. The WCD triggers the SSD providing a clear minimumionizing particle (MIP) signal separated from the background, see figure 6. An engineering array of 10 detectors is scheduled to be installed at the Auger site in 2016. 4.2. The SSD methodology The reconstruction of the muonic shower component is the key element to show the discrimination power for different primary particles. The number of muons Nμ is the least model dependent and most direct composition-sensitive observable that can be obtained by the upgrade of the detector array. The muonic signal in the WCD Sμ,WCD in each detector station can be derived on a station-by-station basis as described in [10], thanks to the signal responses in the SSD (SS S D ) and the WCD (SWCD ) to particles of the electromagnetic and muonic shower components: S μ,WCD = aS WCD + bS S S D ,
Figure 5: SSD schematic view of the two scintillator modules, embedded within an extruded polystyrene foam vessel. The dimensions of the vessel are quoted in the figure.
An SSD unit consists of a box of 3.8 m × 1.3 m, housing two scintillator modules each 1 cm thick and covering an area of 1.9 m2 , see figure 2. In the SSD a double roof, made of corrugated aluminium, is used to reduce the temperature variations. The scintillators are read out by wavelength-shifting fibers guiding the light of the two modules to a photomultiplier tube (PMT) as shown in figure 5.
(1)
where the signals are measured in units of the response to a vertical equivalent muon (VEM) or minimum ionizing particle, respectively. The a and b parameters are derived from detector simulations and have only a very weak dependence on the primary composition and lateral distance from the shower core, as in figure 7 where the bias on the muonic reconstructed signal and the effect on the resolution are shown. Restricted by the limited size of an individual detector station, this method is subject to large fluctuations and is only well suited for deriving mean muon numbers, see figure 8.
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
156
Fitting the lateral distribution function (LDF) is possible to improve the performance of the reconstruction of the muonic signal. Eq. 1 is then applied to the LDF values for the WCD and SSD to calculate the muonic signal at 800 m core distance, Sμ (800). For example at an energy of E=1019.8 eV at θ = 38◦ the reconstruction resolutions of the muonic signal are: σ(S μ (800)) = 22% for protons and 14% for iron. S μ (800) Defining the merit factor as the measure of discrimination power for separating primary i and j using the corresponding observables Si and S j (with the RMS σ) as: S i − S j (2) f MF = σ(S i ) 2 + σ(S j ) 2
Proton Iron
Figure 7: Reconstruction bias (solid symbols) and resolution (open symbols) of the muonic signal contribution for individual detector stations, as a function of distance r from the shower core. The results for proton and iron showers are shown in red and blue, respectively.
using Sμ (800) as a composition estimator, the merit factor for distinguishing between proton and iron primaries is above 1.5 at shower energies larger than 1019.5 eV and small zenith angles.
. 4.3. The water-Cherenkov detector upgrade
0.3
(Nμ, rec- Nμ)/Nμ
p He
0.2
N Fe
0.1 0
Figure 9: First prototype of a SSD scintillator on top of a WCD.
-0.1 -0.2 -0.3 18.8
19
19.2 19.4 19.6 19.8
20
log10(E/eV)
Figure 8: Number of muons Nμ reconstructed (Nμ,rec ) for individual showers using shower universality, compared with the true Nμ as a function of energy for different primary species. Error bars represent the RMS of the distributions.
The WCD is a rotomolded polyethylene tank filled with purified water that produces Cherenkov light when crossed by energetic charged particles associated with cosmic-ray showers. A flexible Tyvek liner inside the tank provides an interface between the water volume and the light sensors that collect Cherenkov light (PMTs) and the light sources that produce calibration pulses (LEDs). Access to the liner, the PMTs and the LEDs is through three hatches located on the top of the tank. An electronics box containing front-end charge amplifiers, shapers, trigger logic, signal buffers, power control, radio transmitter and receivers is located on the top roof of the tank on one of the hatch-covers and is protected by a dome. All the cables connecting the
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
electronics and the light sensors run inside the tank and connect to the electronics via feed-throughs in the hatch covers. Figure 9 shows a picture of an operating WCD. The Surface Detector Electronics (SDE) records the tank signals, makes local triggering decisions, sends timestamps to the central data acquisition system for the global triggers, and stores event data for retrieval when a global trigger condition is satisfied. Because of the small bandwidth (1200 bits/s) available to each tank, the station must operate semi-autonomously, performing calibrations and taking action in response to alarm conditions at the station level. The current SDE was designed 15 years ago using the technology available at that time. Evolution in processors, power consumption of electronics components, and timing systems makes it possible today to design and implement a higher performance electronics system for the Surface Detector array. Furthermore, the proposed electronics provides an interface to allow the scintillator detectors co-located with the surface detector stations to make use of the data processing and communications infrastructure of the stations. The SD stations will be upgraded with new electronics that will process both WCD and SSD signals. Use of the new electronics also aims to increase the data quality (with faster sampling of ADC traces, better timing accuracy, increased dynamic range), to enhance the local trigger and processing capabilities (with a more powerful local station processor and FPGA) and to improve calibration and monitoring capabilities of the surface detector stations. The signals of the WCDs and SSDs will be sampled synchronously at a rate of 120MHz (previously 40MHz) and the GPS timing accuracy will be better than 5ns. The dynamic range of the WCDs will be enhanced by a factor of 32 due to the new electronics and an additional small PMT that will be inserted in one of the filling ports. First prototypes of the new electronics have been built and are being tested. Both the SSDs and the electronics upgrade can be easily deployed, and will have only minimal impact on the continuous data taking of the Observatory. 4.4. The Underground Muon Detector The Underground Muon Detector (UMD) will provide a direct measurement of the muon content of a sub-sample of showers observed by the upgraded Auger surface detector. In the upgrade plan this serves as verification and fine-tuning of the methods used to extract shower muon content using the SSD and WCD stations. The UMD will therefore consist of 61 AMIGA muon detectors deployed on a 750 m grid in the infill area of the Surface Detector, instrumenting a total area of 23.5
157
km2 . The use of the AMIGA muon detectors in this verification role is additional to the rich physics investigations planned for AMIGA in the ankle-region of the energy spectrum[11].
Figure 10: Deployment of a unitary cell AMIGA station with two 10 m2 and two 5 m2 modules at a depth of 2.3 m underground. The big tubes are to provide access to the electronics for development and maintenance purposes, to be replaced by tubes of 31 cm diameter for AMIGA production.
. Each muon detector station will have an area of 30 m2 and will be buried at the side of a surface detector station at a depth of approximately 1.3 m, see figure 10. The distance to the station will be large enough to avoid shadowing from the water tank, guaranteeing uniform shielding, but small enough to represent the same physical point inside the shower front, and allowing shared use of GPS time signals and telecommunications with its associated surface detector. The baseline design for the muon detectors uses the same extruded plastic scintillators already developed and used by the MINOS experiment [12]. They will work as counters (i.e. signals above a tunable threshold are counted) having an appropriate segmentation to prevent pile-up, together with an integrated signal for large muon densities. The light collected in a scintillator will be guided towards a 64 channel Hamamatsu PMT. An AMIGA prototype is displayed in figure 11. It consists of 64× 4 m long scintillator strips. The strips are 4.1 cm wide and 1.0 cm thick and therefore the detector modules have an active area of 10 m2 . The strips have a middle groove which accommodates a wavelength shifter (WLS) fiber of 1.2 mm diameter which is glued into the groove and covered with reflective foil. The scintillator is co-extruded with a TiO2 reflective coating which prevents light from leaving the material.
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
Figure 11: An AMIGA 10 m2 module being manufactured. The black optical connector concentrates the 64 optical fibers coming from the 32 scintillator bars at each side.
. Muon counting and the digitization of the integrated signal is implemented in the AMIGA electronics, including an FPGA with three main functional blocks: counting, data codification, and external communications. The data are sent to the central data acquisition system of the observatory (CDAS) using an independent, commercial radio communications system. 4.5. The fluorescence detector operation upgrade The fluorescence detector provides exceptional information about extensive air showers such as a modelindependent energy reconstruction and direct measurement of the longitudinal development profiles. The main limitation of the FD is its duty cycle, currently below the 15%, see figure 12. Our goal is to increase the exposure for cosmic-ray events above 1019 eV by extending the FD measurement into hours with high night sky background (NSB). The current setup allows this novel operation and we have performed several tests that successfully demonstrate that it is feasible. Safety limits on the long and short term illumination of PMTs by the NSB, and particularly scattered moonlight, define the data taking period of the current FD operation. The duty cycle is therefore limited to about 19%, which is reduced to 15% by bad weather conditions, power cuts and malfunctions. A significant increase of the duty cycle is possible by the extension of the FD operation to times at which a large fraction of the moon in the sky is illuminated. However, during such operations the PMT gain must be reduced by lowering the supplied high voltage (HV) to avoid an excessively
high anode current leading to an irreversible deterioration of the PMT sensitivity. The HV power supplies installed in the FD buildings allow switching between two high voltage levels and the PMTs can be operated at the nominal gain and a lower gain. The first test measurement outside the standard FD data taking period was performed with one FD telescope in the austral autumn of 2015. Telescope 1 at Los Leones was operated during six nights with a highly illuminated moon above the horizon. We have confirmed that the HV change can be done remotely and that the HV stabilizes within a few seconds after its change. The PMT performance can be monitored with the existing calibration setup in exactly the same way as during the standard data taking, so no change in the system is required. By extending the FD operation to higher NSB we can gain up to 40% more events above 1019.5 eV. This will allow us to improve the cross check of results of the upgraded SD array with the FD up to the energy of the suppression in the cosmic-ray flux. 15 14 FD uptime fraction [%]
158
13 12 11 10 9 8
2
4
6
8
10
12
14
16
18
20
22
24
Telescope
Figure 12: Fluorescence telescopes uptime fractions. The colors indicate the 4 sites of the telescopes: blue Los Leones, magenta Los Morados, yellow Loma Amarilla and green Coihueco [13].
5. Auger PRIME expected performance It is complicated to demonstrate the potential of Auger PRIME without knowing what composition to expect in the GZK suppression region [16]. Therefore we show the discrimination power of the additional information obtained with the Auger upgrade in two physics motivated benchmark models. Those are obtained by fitting the Auger flux [14] and composition data [8] for E > 1018.7 eV. In figures 13 and 14 the energy spectra of the two scenarios are compared with that measured by Auger [6].
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
159
To estimate the performance on the mass composition discrimination mock data sets were generated for these two scenarios with a statistics corresponding to 7 years of data taking with Auger PRIME.
38
840
-2
-1
-1
E J [eV km sr yr ]
10
3
2
820
10
37
10
36
p, EPOS-LHC Fe, EPOS-LHC
800 780
18
18.5
19
19.5
20
20.5
log (E/eV) 10
Figure 13: Benchmark spectra chosen as representations of a maximum-rigidity scenario [6]. The index of the injection energy spectrum at the sources is about -1. The colors for the different groups are proton-blue, helium-gray, nitrogen-green, and iron-red.
760 740 720 700
Scenario 1
19
19.2
19.4
19.6
Scenario 2
19.8
20
lg(E/eV)
Figure 15: Reconstructed mean depth of shower maximum Xmax with the two scenarios 1 maximum rigidity model and 2 photodisintegration model.
38
10
E J [eV km sr yr ]
3
2
-2
-1
-1
The mean Xmax , RMS (Xmax ), and the relative muon number Rμ are shown in figures 15, 16 and 17 respectively. The RMS contains the intrinsic air-shower fluctuations and the reconstruction resolution. The models predict significantly different extrapolations into the GZK suppression region, while the mean Xmax , RMS (Xmax ) and Rμ are very similar up to 1019.2 eV that is the energy range that is well covered by data of the fluorescence telescopes [7]. The differences between the two scenario are well reproduced with the reconstructed Xmax , RMS (Xmax ) and Rμ and the two scenarios can be distinguished with high significance and statistics.
37
10
36
10
18
18.5
19
19.5
20
20.5
log (E/eV) 10
Figure 14: Benchmark spectra chosen as representations of one photodisintegration scenario [15]. The index of the injection energy spectrum at the sources is about -2. The colors for the different groups are proton-blue, helium-gray, nitrogen-green, and iron-red.
6. Conclusions The Pierre Auger Observatory data until now do not lead to a definitive answer to the origin of the cosmic rays. The Auger upgrade promises high-quality future data, and real scope for new physics uses of existing events. With operation planned from 2018 until 2024, event statistics will more than double compared with the ex-
C. Di Giulio / Nuclear and Particle Physics Proceedings 279–281 (2016) 153–160
160
RMS(Xmax) [g/cm2]
90
p, EPOS-LHC
80
Fe, EPOS-LHC
70 60 50 40 30 Scenario 1
20 10 18.8
19
Scenario 2
19.2 19.4 19.6 19.8
20
lg(E/eV)
Figure 16: Shower maximum Xmax fluctuation shown in the two scenarios.
isting Auger data set, with the critical added advantage that every event will now have mass information. This will allow us to address some of the most pressing questions in UHECR physics, including that of the origin of the flux suppression, the prospects of light particle astronomy and secondary particle fluxes, and the possibility of new particle physics at extreme energies. Obtaining additional composition-sensitive information will not only help to better reconstruct the properties of the primary particles at the highest energies, but also improve the measurements in the important energy range just above the ankle. Furthermore, measurements with the new detectors will help to reduce systematic uncertainties related to modeling hadronic showers and to limitations of reconstruction algorithms. This improved knowledge of air-shower physics will likely then also allow a re-analysis of existing data for improved energy assignments, for mass composition studies, and for photon and neutrino searches. Finally it should be mentioned that the addition of scintillator detectors across the entire Observatory will also make possible direct comparisons of Auger measurements with those of the surface detectors of the Telescope Array experiment. References
0.25 0.2
0.15
p, EPOS-LHC Fe, EPOS-LHC
0.1 0.05 0 -0.05 -0.1 -0.15
Scenario 1
-0.2 -0.25
Scenario 2
19
19.2
19.4
19.6
19.8
20
lg(E/eV)
Figure 17: Mean number of muons at 38 degree shown in the two scenarios. The number of muons is given relative to that expected for an equal mix of p-He-CNO-Fe as primary particles.
[1] Aab, A., et al. (The Pierre Auger Collaboration) NIM A 798 (2015) 172-213 [2] Verzi, V. et al. (The Pierre Auger Collaboration), Proc of 33rd Int. Cosmic Ray Conf., Rio de Janeiro (2013) [arXiv:1307.5059 [astro-ph.HE]] [3] Vali˜no, I. for the Pierre Auger Collaboration: 2015, PoS(ICRC2015)271 [4] Bleve, C. for the Pierre Auger Collaboration: 2015, PoS(ICRC2015)1103 [5] Aab, A., et al. (The Pierre Auger Collaboration): 2015, PhRvD 91, 092008 [6] Di Matteo, A. for the Pierre Auger Collaboration: 2015, PoS(ICRC2015)249 [7] Aab, A., et al. (The Pierre Auger Collaboration): 2014, PhRvD 90, 122005 [8] Aab, A., et al. (The Pierre Auger Collaboration): 2014, PhRvD 90, 122006 [9] Coutu, S. for the Pierre Auger Collaboration : this volume [10] Letessier-Selvon, A., et al., NIM A 767 (2014) 41-49 [11] Suarez, F. for the Pierre Auger Collaboration, [arXiv:1307.5059 [astro-ph.HE]]. [12] Michael D.G., et al., NIM A 596 (2008) 190-228 [13] Abraham, J., et al. (The Pierre Auger Collaboration) NIM A 620 (2010) 227-251 [14] Aab, A., et al. (The Pierre Auger Collaboration), Proc of 33rd Int. Cosmic Ray Conf., Rio de Janeiro (2013) [arXiv:1307.5059 [astro-ph.HE]] [15] D. Allard, Astropart. Phys. (2012) 39-40 [16] Engel, R. for the Pierre Auger Collaboration: 2015, PoS(ICRC2015)685