Proc. 2016 Int. Conf. Ultra-High Energy Cosmic Rays (UHECR2016) https://doi.org/10.7566/jpscp.19.011033 Auger at the Telescope Array: Recent Progress Toward a Direct Cross-Calibration of Surface-Detector Stations Sean Quinn 1,, C. Covault 1, J. Johnsen 2, P. Lebrun 3, P. Mantsch 3, J.A.J. Matthews 4, P. Mazur 3, F. Sarazin 2 and R. Sato 5 for the Pierre Auger Collaboration a ; R. Cady 6, T. Fujii 7, J.N. Matthews 6, T. Nonaka 7, S. Ogio 8, H. Sagawa 7 and R. Takeishi 7 for the Telescope Array Collaboration b ; S. Colognes 9, B. Courty 9, B. Genolini 10, L. Guglielmi 9, R. Halliday 1, R. Lorek 1, M. Marton 11, E. Rauly 10, T. Trung 10, O. Wolf 2 1 Dept. of Physics, Case Western Reserve Univ., 2076 Adelbert Road, Cleveland, Ohio 44106, USA 2 Physics Department, Colorado School of Mines, Golden CO, USA 3 Fermilab, Batavia IL, USA 4 Department of Physics and Astronomy, University of New Mexico, Albuquerque, NM, USA 5 Pierre Auger Observatory, Malargüe, Argentina 6 Department of Physics and Astronomy, University of Utah, Salt Lake City, UT, USA 7 Institute for Cosmic Ray Research, University of Tokyo, Kashiwa, Chiba, Japan 8 Graduate School of Science, Osaka City University, Osaka, Osaka, Japan 9 Laboratoire Astroparticules et Cosmologie (APC), Université Paris 7, CNRS-IN2P3, Paris, France 10 Institut de Physique Nucléaire d Orsay (IPNO), Université Paris 11, CNRS-IN2P3, Orsay, France 11 Laboratoire de Physique Subatomique et de Cosmologie (LPSC), Université Joseph Fourier, INPG, CNRS-IN2P3, Grenoble, France a Auger author list: auger.org/archive/authors_2016_09.html b TA author list: telescopearray.org/images/papers/201703-authorlist.pdf E-mail: spq@cwru.edu (Received Mar 29, 2017) Since 2007 the Telescope Array Project (TA) and the Pierre Auger Observatory (Auger) have collected extensive data sets spanning several orders of magnitude in energy of the cosmic-ray spectrum. In both experiments the bulk of data is generated from the surface-detector (SD) array, which is energetically calibrated with fluorescence detectors using a hybrid approach. However, each experiment has implemented a different SD station design, resulting in different sensitivities of extensive airshower channels. Understanding these differences and any potential unforeseen systematic errors is essential for future joint analyses. In this paper we present an update on the progress of an in-situ cross-calibration program. We focus on recent hardware installations which enable the read out of co-located Auger and TA SD stations at the TA central laser facility (CLF). We also present a preliminary analysis of event signals observed at the CLF. KEYWORDS: (cross) calibration, data acquisition, instrumentation 1. Introduction 1.1 Participants The Pierre Auger Observatory is a hybrid cosmic-ray (CR) observatory located in Mendoza province, Argentina, which detects extensive air showers using four fluorescence-telescope detectors (FD) and 1660 surface-detector (SD) stations. The Observatory began formal data collection in 2004. Fluorescence observations provide high-quality data of shower energy as well as depth of maximum shower development, but have a limited duty cycle. A subset of hybrid events passing stringent quality cuts is used to calibrate SD only events which represent most of the experiment s data. The 011033-1 2018 The Author(s) This article is available under the terms of the Creative Commons Attribution 4.0 License. Any further distribution of this work must maintain attribution to the author(s) and the title of the article, journal citation, and DOI.
011033-2 Telescope Array Project is also a hybrid CR observatory located in Utah, USA. In addition to its three FD, TA operates 507 SD stations, with an additional 400 detectors being added over the next few years to quadruple the detection area [1]. Both experiments implement the hybrid approach and utilize SD stations with self-contained electronics, communications, and solar power systems. The detection medium in Auger is used to count mainly air-shower muons, while TA detectors will count any ionizing particles. The Auger SD station is a water-cherenkov detector (WCD). Relativistic leptons and high-energy photons generate signals via Cherenkov radiation and pair production, respectively, which are collected by photomultiplier tubes (PMTs). A detailed technical description can be found in [2]. One advantage of the WCD design is the ability to distinguish between muonic and electromagnetic shower components. The Telescope Array SD station uses plastic scintillator panels. Fluorescent scintillation light is collected by wavelength-shifting fibers and guided to PMTs in a dual layer setup. A complete description of the experiment is in [3]. The nature of this medium makes it difficult to distinguish particle types. 1.2 Motivation To better understand the energy spectrum and origin of ultrahigh-energy cosmic rays (UHECRs) the TA and Auger Collaborations have performed analyses of a joint data set [4, 5]. These studies benefit from increased statistics and full sky coverage. A recent analysis [5] has concluded that UHECR composition results of both experiments are in agreement. The energy spectra also agree within systematic errors up to the ankle, but diverge toward the highest energies near the Greisen, Zatsepin, Kuz min (GZK) cutoff [6, 7]. The nature of this discrepancy is unknown. Speculative explanations might include: the result of energy-dependent systematic errors, or the manifestation of different physics scenarios in the northern and southern skies. In terms of arrival-direction clustering, TA observes a hot spot [8] while Auger reports no significant deviation from isotropy [9] at the time of writing. We intend to investigate the possibility of energy-dependent systematics using a direct comparison of surface-detection methods through a two-phase joint cross-calibration program. We are nearing completion of phase one where data for station-level responses to the same air shower is compared. In this paper we review results reported from earlier work [10], provide an update of newly deployed hardware, present station trace data and first data of shower-triggered station-tostation comparisons. 2. Detectors The components required to complete phase one have been installed at the TA CLF site. This site was chosen since it has easy access to roads and the TA wireless local-area network (WLAN). The CLF itself is near the center of the SD grid within the Long Ridge subarray surrounded by four adjacent stations, being closer to the western neighbors. We are currently collecting data from: one Auger south (AS) WCD, one prototype Auger north (AN) and two TA stations, see Figure 1. By forming a doublet between AN and AS we are able to build an empirical mapping function to translate signals between these detectors. Thus, when a microarray of AN stations is deployed in the future, we will be able to map these results to an equivalent micro-array of AS detectors. Both AS and AN detectors consist of a 3.6 m diameter, 1.2 m high, reflectively lined (Tyvek) tank filled with 12,000 L of pure water. The AN station uses only one downward facing 9 PMT, compared to AS which uses three symmetrically distributed PMTs 1.2 m from the central axis [2]. The AS PMTs run at positive anode high voltage (HV) which is sourced from the base electronics, set by a slow control process in the AS electronics. The electronics receive low-gain (AC coupled anode) and high-gain (8th dynode) channels for each PMT. Digitization is performed by 10 bit 40
Proceedings of 2016 International Conference on Ultra-High Energy Cosmic Rays (UHECR2016) 011033-3 MHz semi-flash ADCs. These traces are then analyzed and time-stamped (GPS reference clock) by a programmable logic device running triggering firmware. A central data-acquisition (DAQ) system running on a single board computer (SBC), more details given in section 3.1, evaluates the timestamps of potentially interesting station-level events. Coincident time-stamps in a certain geometry generate a physics trigger, which instructs the stations to transfer ADC traces for the event. Online calibration data from the previous minute are also included to express the ADC trace as a muonequivalent signal. The AN electronics design is similar to AS, but uses fewer components since only a single PMT signal is digitized. The key differences are: a new 10 bit 100 MHz flash ADC (FADC) which processes four channels for increased dynamic range. The electronics box resides inside the tank, so the HV module is integrated into the electronics motherboard instead of PMT base. The anode signal is split into 0.1, 1, and 30 channels, and instead of the 8th dynode, the 5th stage is used. The TA station uses two layers of scintillators 3 m2 in area and 1.2 cm thick. Light is guided through 104 wavelength shifting fibers to a 30 mm PMT. Like the AS design, the PMT base provides HV from a reference signal sent by the electronics. The anode is digitized by 12 bit 50 MHz FADCs which are processed by triggers implemented on field-programmable gate arrays. Like Auger, a hierarchical triggering system is implemented, based on minimum-ionizing particles (MIP). Potentially interesting station-level events are communicated using a custom wireless protocol. Physics triggers are similar to those in Auger, requiring a coincidence of station-level events in a certain geometry. The four detectors are co-located at the CLF site within a few dozen meters. The Auger doublet is located in the northeast corner, one TA station is in the northwest, and the other in the southwest corner. The maximum separation is between the Auger doublet and the southwest TA station, which is roughly 44 m. In the next section we describe how we trigger and read out the detectors. Fig. 1. Wide angle photo of hardware deployed at the TA CLF site. Custom cabling was setup along guy wires from stations to the CLF for data acquisition. Due north is into the page. 3. 3.1 DAQ system High-level description In order to retrieve data from the Auger stations, which are designed to operate wirelessly, modifications were made to communicate over a physical wire which is connected to a SBC housed inside the CLF. An external trigger is required to read out the Auger doublet for low-energy events: this is provided by a primitive threshold-comparator circuit attached to a bare local TA SD station (i.e., without TA electronics). For a direct comparison between Auger and TA waveforms, a second fully functional TA station is installed in the southwest corner and is able to send data wirelessly using the standard TA wireless communication protocols. When this station is triggered, the relevant time
011033-4 stamp is also communicated to the Auger computer. 3.2 TA global (physics) trigger During March 2016 a TA SD electronics box was brought online at the CLF site. It uses a parabolic radio antenna to listen for physics triggers sent to the TA SD array. The firmware was modified to relay time-stamp information for shower candidates observed by the TA DAQ computer. This data is forwarded over a RS-232 serial connection to the SBC. As previously mentioned, the southwest global TA station uses standard TA electronics and operates in normal TA DAQ mode. Unlike active TA stations, station-level trigger time stamps are ignored hence this station does not participate with adjacent stations to form a physics trigger. However, when an event is observed in the array, this station transmits data in a normal fashion. The relevant traces for this project are only collected when a core lands in a constrained area of neighboring stations. Nevertheless, since Auger traces are retrieved for the global trigger timestamp, a direct comparison of station waveforms is possible. Moreover it will be possible to use Auger data simultaneously with TA data for a combined dual-detector extensive air-shower reconstruction. 3.3 CLF vicinity local trigger The northwest TA station, installed August 2016, will be used to investigate showers of lowerand intermediate-energy with core positions close to the CLF. Currently it only operates as an external trigger for the Auger doublet until a way to collect the TA waveforms over a hardwired connection is devised. The northwest TA is roughly 32 m from the doublet. The PMT is operated nominally at 1.2 kv, and when the anode output of both PMTs simultaneously crosses a threshold of < -92 mv, the circuit transmits a logic pulse over a RG58 cable to a development FPGA board (MicroZed) running time-tagging firmware referenced to GPS time. The signal is time-stamped and sent over a serial connection to the Auger SBC (Raspberry Pi 2 Model B). This station is operated at a base rate of 3 15 Hz, consisting of two-fold coincidences between the layers, and is powered by the TA solar panel and battery. 3.4 Trigger decision and software Auger stations are designed to transmit data and receive commands over radio. Custom interconnects and cabling bring AN CANbus and AS RS-232 serial data into the CLF in place of radios. A radio-protocol emulation program on the SBC is used to decode AS data and send control or read out commands to the AS station, while similar software handles AN. These packages provide real-time station-level event lists (Auger T2 triggers). These lists, combined with another program which parses the CLF local and global trigger, are analyzed for coincident time stamps. A read out request (Auger T3 trigger) is sent if the time difference between the Auger doublet event and global event is < 100 µs, or if the local trigger and Auger doublet difference is < 20 µs. These values are chosen to account for any systematic timing offsets in electronics, observed to be about 10 µs. A high-level diagram of the setup is displayed in Figure 2. Events are archived by day (UTC) and uploaded to Case Western Reserve University and TA servers where data decompression and analysis are performed using a micro Auger central data acquisition code written for this project. Monitoring and detector-performance information is updated daily. 4. Analysis & Results 4.1 Auger and TA trace comparison To date we have collected 20 global coincidences between the AS and neighboring TA stations. The CLF global station (DET2421) triggered for 15 of these events. In Figure 3 we show FADC
011033-5 Fig. 2. Left: block diagram of cabling and hardware deployed at the CLF site (see text for details). Right: trigger-condition flow chart showing the do while loop to compare example time stamps. Items in parentheses refer to the global trigger. waveforms for a single event. For AS, the anode channel (low gain) was used. To translate from hardware units to vertical-equivalent muons (VEM) the AS signal is integrated and rescaled using the muon peak of the charge histogram, generated over a 60 s interval, and a nominal gain of 32. For the TA signal, the trace is integrated and scaled using the peak in the pedestal and charge histograms, generated from data over the previous 10 min., with a gain factor of 1.5. The baseline is estimated using the mean of quiescent, low-dispersion trace windows, and interpolated through signal regions. We find particle densities of 98.5 VEM and 81.9 MIP for this event, using the arithmetic mean of PMTs for both detectors. 20 PMT 1 PMT 2 PMT 3 8 PMT 1 PMT 2 15 6 Signal [VEM] 10 Signal [MIP] 4 5 2 0 220 230 240 250 260 270 280 290 300 310 Time [25 ns] 0 0 20 40 60 80 100 Time [20 ns] Fig. 3. Global trigger event: Auger south (left) and TA (right) ADC traces. 4.2 MIP vs. VEM calibration curve Using 15 coincident global events we can calculate the AS and TA signals, see Figure 4. With more statistics, this curve will form the basis to compare detector responses to the same shower and it should be possible to investigate how any potential difference in detector sensitivity depends on air-shower parameters. In future studies we intend to compare these empirical results to simulations using the Auger Offline framework including the WCD and a version of scintillator detector.
011033-6 200 DET2421 Signal [MIP] 150 100 50 PRELIMINARY 0 0 50 100 150 200 AS Signal [VEM] Fig. 4. Data points show the processed signals for the same air shower. N.B. Aside from baseline removal and integration, no additional post processing or quality cuts have been made on this data. 5. Summary and future outlook In this work we have highlighted progress made since the publication of [10]. Key improvements include installation of an SBC which allows remote control and read out of Auger stations that are coincident with a local CLF trigger and a TA array-wide global trigger. We have presented preliminary traces for sample events along with the VEM vs MIP correlations between them. We intend to continue collecting data using this setup, while preparing to implement phase two of the project, which involves deploying a micro array of Auger stations in the TA grid. This array will independently observe and reconstruct showers detected by both collaborations, allowing us to cross-validate each experiment s complete reconstruction pipeline. Acknowledgements This project was supported in part by JSPS Grant-in-Aid for Scientific Research (A) 25247035. S. Quinn s attendance was supported in part by a travel grant from the American Physical Society. References [1] H. Sagawa, Proc. of the 34th Int. Cosmic Ray Conf., The Hague, The Netherlands, paper 657 (2015) [2] The Pierre Auger Collaboration, Nucl. Instr. Meth. Phys. Res. A 798, 172 (2015) [3] T. Abu Zayyad et al., Nucl. Instr. Meth. Phys. Res. A 689, 87 (2012) [4] O. Deligny, Proc. of the 34th Int. Cosmic Ray Conf., The Hague, The Netherlands, paper 395 (2015) [5] M. Unger, Proc. of the 34th Int. Cosmic Ray Conf., The Hague, The Netherlands, paper 307 (2015) [6] K. Greisen, Phys. Rev. Lett. 16, 748 (1966) [7] G. T. Zatsepin, V. A. Kuz min, JETPL 4, 78Z (1966) [8] P. Tinyakov, Proc. of the 34th Int. Cosmic Ray Conf., The Hague, The Netherlands, paper 326 (2015) [9] The Pierre Auger Collaboration, ApJL 762, L13 (2013) [10] R. Takeishi, Proc. of the 34th Int. Cosmic Ray Conf., The Hague, The Netherlands, paper 393 (2015)