Studying the nuclear mass composition of Ultra-High 2 Energy Cosmic Rays with the Pierre Auger 1 0 Observatory 2 n a L. Cazon1 for The Pierre Auger Collaboration2,3 J 0 1 LIP,Av Elias Garcia 14-1,1000-149 Lisboa, Portugal 3 2 Av. San Mart´ın Norte304 (5613) Malargu¨e, Argentina 3 A full author list and affiliations can befound at ] http://www.auger.org/archive/authors 2011 09.html E H E-mail: [email protected] . h Abstract. p TheFluorescenceDetectorofthePierreAugerObservatorymeasurestheatmosphericdepth, - o Xmax, wherethelongitudinal profileof thehigh energy air showers reaches itsmaximum. This r is sensitive to the nuclear mass composition of the cosmic rays. Due to its hybrid design, the t s PierreAugerObservatoryalsoprovidesindependentexperimentalobservablesobtainedfromthe a SurfaceDetectorforthestudyofthenuclearmasscomposition. WepresentXmax-distributions [ and an update of the average and RMS values in different energy bins and compare them to 1 the predictions for different nuclear masses of the primary particles and hadronic interaction v models. We also present the results of the composition-sensitive parameters derived from the 5 ground level component. 6 2 6 . 1 0 1. Introduction 2 ThedeterminationofthenuclearmasscompositionofUHECR(UltraHighEnergyCosmicRays) 1 : is fundamental to unveil the origin of the most energetic particles known in nature. UHECR v are detected by means of the extensive air showers created in the Earth’s atmosphere, which are i X composed by a cascade of hadrons and electromagnetic (EM) particles. The depth at which the r EM shower reaches its maximum, Xmax, strongly correlates with the depth where the primary a firstly interacted. The Xmax-distribution carries information aboutthe primary particle and the physical processes in the cascade. The hadronic cascade is composed mostly by pions, of which the charged pions might decay in flight into a muon and a neutrino. Due to their long lifetime and low cross section, muons can leave the core of the hadronic cascade traveling kilometers away and be detected. Electrons spread out in time and space because of Coulomb scattering, whereas muons practically travel following straight lines. The time structure of the shower disc encodes the history of the shower, allowing us to recover information about the longitudinal evolution by analyzing the time distribution of the particles arriving at ground level. The Pierre Auger Observatory, located on the high plateau of the Pampa Amarilla, is the largest cosmic ray observatory ever built. Its hybrid design allows to collect the shower particles by a surface detector (SD) and to observe the longitudinal development of the EM profile by collecting the UV light with a fluorescence detector (FD). The SD spans 1600 Cherenkov 2 detectorsina1.5kmtriangulargridover3000km ,whereastheFDiscomposedby24telescopes distributed over 4 sites overlooking the array. The baseline design is being complemented with enhancements both on the SD and FD, and new detection techniques [1]. Inthispaperwepresentthelatest hXmaxiandRMS-Xmax resultsas afunctionof energy, and theXmax-distributionsforthehighestandlowestenergybins. Wealsopresenttheevolutionwith energy of SD observables extracted from the time distributions of the signal which are sensitive to the longitudinal evolution of the shower. We compare our results with the predictions of hadronic interaction models [2]. 2. Electromagnetic Shower Maximum We have used data taken between December 2004 and September 2010, following the analysis reported in [3] and updated in [4]. We have considered those showers reconstructed by the FD that have at least a signal in one of the SD stations. The longitudinal profile was fitted with a Gaisser-Hillas function, and a series of cuts depending on the atmospheric optical conditions and on the quality of the reconstructed profile were applied [3, 4]. In order to avoid that the data sample is biased regarding to the nuclear mass, a number of cuts in the geometry were imposed. −2 The resolution in Xmax is around 20 gcm . Uncertainties in the atmospheric conditions, −2 calibration, event selection andreconstruction resultinasystematic uncertainty of ≤ 13gcm . The average values as a function of the energy are displayed in the left panel of Fig. 1 (third plot from top), alongside with the predictions for different hadronic interaction models. The elongation rate D10 = dhXmaxi/dlog10E is better described using two slopes, with 2 χ /ndf=7.4/9, compared to 54/11 if we used a single power law. The change of the elongation rate could be interpreted as a change on the nuclear mass composition, provided that there is no change on fundamental properties of the hadronic physics. The comparison of hXmaxi with the different predictions of the hadronic interaction models could be interpreted as a change of composition towards heavier elements at the highest energies. The left panel of Fig. 1 also displaystheRMS-Xmax,whichhasbeencorrectedbythedetectorresolution. Itrapidlydecreases with energy around the same energy of the change on the elongation rate. The distributions of Xmax for all energy bins have been published in [4]. Fig. 2 plots the (Xmax −hXmaxi)-distributions for the highest and lowest energy bins. The predictions of the different hadronic models show a nearly universal shape. At low energies, the shape of the data distribution is compatible with a very light or mixed composition, whereas at high energies a heavier composition is favored. The high Xmax tail, characteristic of very light components, is suppressed with respect to the lowest energies. 3. Asymmetry of the signal risetime In each SD event, the Cherenkov detectors record the signal as a function of time. The risetime, t1/2, defined as the time elapsed between the 10% and 50% of the integrated signal, depends on the distance to the shower maximum [5], the zenith angle θ and the distance to the core r. In inclined showers, the average risetime depends on an angle defined on the perpendicular plane ζ as ht1/2/ri = a + bcosζ, due to the different effective distances of the tanks to the shower maximum. The evolution of b/a with zenith angle reaches a maximum at Θmax which is different for different primary masses [6]. In the right central panel of Fig. 1 an example of b/a as a function of ln(secθ) is shown for the energy bin log (E/eV) = 18.85−19.00. 10 18 ◦ Events with energy above 3.16×10 eV and θ ≤ 60 were selected [8]. Results of Θmax as a function of energy are displayed in the left panel of Fig. 1. The systematic uncertainty amounts to <∼ 10% of the proton-iron separation predicted by models. The numberof muons predicted by the hadronic interaction models differs from data [9]. A preliminary study including this effect indicates a possible change of about ≤ 5% of the proton-iron difference. Figure 1. Left Panel: Results on shower evolution sensitive observables compared with models predictions. The error bars correspond to the statistical uncertainty, whereas the systematic uncertainty is represented by the shaded bands. Right Panel, top: Typical longitudinal development of the muon production. Middle: Average asymmetry in the risetime. Bottom: Typical longitudinal development on the energy deposit. 4. The muon production depth An approximated relation between the arrival time delay of a muon with respect to the shower front plane and the distance of the muon production point to ground can be used to transform the arrival time distributions of muons into production depth distributions [10]. In [11], this ◦ ◦ technique was applied to Auger data in an angular window between 55 and 65 for stations with r ≥ 1800 m, as a result of a trade off between EM contamination rejection and the intrinsic resolutionofthemethod. InFig.1,righttoppanel,anexampleforarealeventatE = 94±3EeV is shown. Similarly to the EMcomponent, the depth at which themuon production rate reaches a maximum, Xµ is a good indicator of the primary mass composition. After a series of cuts max [11], the measured values of hXµ i are presented in the upper panel of Fig. 1. The systematic max uncertainty duetoreconstruction bias,coreposition, rejection oftheEMcomponentandquality Figure 2. Centered distributions Xmax −hXmaxi, for the lower and highest energy bins. The mixed composition corresponds to 50% p and Fe. −2 cuts amounts to 11 gcm , corresponding to 14% of the proton-iron separation. Note that the discrepancy on the total number of muons observed between models and data does not affect this result, but only the normalization of the muon production depth distribution. 5. Conclusions Provided that the present models give a fair description of the physical processes and their systematics, the average primary composition of UHECR could be inferred from the left panel of Fig. 1, showing a change towards heavier mass composition at high energies. The evolution of hXmaxi, Θmax , and hXmµaxi is similar in the overlapping energy regions, despite the fact of having completely independent systematics. RMS-Xmax evolution can be accommodated with a variety of compositions since it is influenced by the shower-to-shower fluctuations and the relative differences on hXmaxi. The analysis of the shape of the Xmax-distributions also points to the direction of heavier composition as we increase the energy. On the other hand, it is not guaranteed that current hadronic interaction models describe well particle physics at these energies [9], and a different interpretation would therefore apply. References [1] H. Klages for thePierre AugerCollaboration, these proceedings [2] N.N. Kalmykov,S.S.Ostapchenko,andA.I.Pavlov,Nucl.Phys.Proc.Suppl.52B(1997) 17-28; E.J.Ahnet al., Phys. Rev.D80 (2009) 094003; T. Pierog and K.Werner, Phys.Rev. Lett.101 (2008) 171101 [3] The Pierre AugerCollaboration, Phys.Rev.Lett. 104 (2010) 091101. [4] P. Facal for the Pierre Auger Collaboration, Proceedings of the 32nd ICRC at Beijing, China (2011). arXiv:1107.4804 [5] H. Wahlberg, for the Pierre Auger Collaboration, (2009) Proc. 31st ICRC, Lodz, Poland (2009). arXiv:0906.2319v1 [6] M.T.Dova,M.E.Mancenido, A.G.Mariazzi, H.Wahlberg,F.Arqueros,D.Garcia-Pinto, Astropart.Phys. 31 (2009) 312-319. [7] The Pierre AugerCollaboration, Nucl.Instrum. Meth. A613 (2010) 29-39. [8] D.Garcia-PintoforthePierreAugerCollaboration,Proc.32ndICRC,Beijing,China(2011).arXiv:1107.4804 [9] J. Allen for the Pierre Auger Collaboration, Proc. 32nd ICRC, Beijing, China (2011). arXiv:1107.4804 ; R. Ulrich for thePierre AugerCollaboration, theseproceedings [10] L.Cazon, R. A.Vazquez, E. Zas, Astropart. Phys. 23 (2005) 393-409. [11] D. Garcia-Gamez for the Pierre Auger Collaboration, Proc. 32nd ICRC, Beijing, China (2011). arXiv:1107.4804