aa r X i v : . [ a s t r o - ph ] O c t Ultra High Energy Cosmic Rays from earth-basedobservatories
Sergio PetreraINFN and Physics Department, L’Aquila UniversityI-67010 L’Aquila, ITALY
The origin of the highest energy cosmic rays is one of the most exciting questionsof astroparticle physics. Even though a general concept linking magnetic field andsize of possible sources (the so called “Hillas plot” [1]) is the basis of our currentunderstanding, up to now there are no generally accepted source candidates knownto be able to produce particles of such extreme energies.At these energies cosmic rays are expected to exhibit a suppression in the en-ergy spectrum because of their interaction with the microwave background radiation(CMB). This feature, known as the Greisen-Zatsepin-Kuz’min (GZK) effect [2], is atabout ∼ · eV for protons. It limits the horizon from which these particles canbe observed to a distance below about 100 Mpc (depending on the primary mass).The non-observation of the GZK effect in the data of the AGASA experiment [3]has motivated several theoretical and phenomenological models trying to explain theabsence of the GZK effect. More recently both HiRes [4] and Auger [5] have shownevidence of a suppression such as expected from the GZK effect with high statisticalsignificance. The recent observation of directional correlations of the most energeticAuger events with the positions of nearby Active Galactic Nuclei [6] complements theobservation of the GZK effect very nicely.Mass composition is another important key to discriminate among different modelsabout the origin of high-energy cosmic rays. Such measurements are difficult due totheir strong dependence on hadronic interaction models. Only primary photons canbe discriminated safely from protons and nuclei and recent upper limits to their fluxlargely constrain existing top-down models.In this paper, prepared for the Physics in Collisions 2008 Conference, each of thesetopics are exposed and reviewed. In the UHE region two detection methods are effective for extensive air showers (EAS):arrays of surface detectors and air fluorescence detectors. A comprehensive review of1hese experimental methods can be found in [7]. In this Section recent experimentsdedicated to the detection of cosmic rays are briefly described.
AGASA . The Akeno Giant Air Shower Array, located in Japan at the latitude ofabout 35 ◦ N and altitude of 900 m above sea level was in operation from 1990 until2004. It was a large surface array [8], designed to measure the front of the cosmic rayshowers as they reach ground. The array consisted of 111 plastic scintillators withsize of 2.2 m deployed with separation of 1 km and covering an area of 100 km . Thearray was complemented by 27 muon detectors consisting of proportional countersplaced below absorbers. HiRes . It is the new and sophisticated version of the pioneering Fly’s Eye instrumentof the Utah group based on the detection of the fluorescence light from the nitrogenmolecules excited by the charged particles of the cosmic ray showers. It was inoperation from 1997 until 2006. The HiRes instrument [9] consists of two sites 12.6km apart located at Dugway in Utah (USA) hosting 22 telescopes at HiRes I and42 at HiRes II. The telescopes cover the full 360 ◦ azimuth and in elevation from 3 ◦ up to 17 ◦ (HiRes I) and from 3 ◦ up to 31 ◦ (HiRes II). The main components of eachtelescope are a spherical mirror of about 4 m size and an array of 256 photomultipliersas sensitive element. UV filters to cut light outside the 300-400 nm interval of thenitrogen fluorescence were also used. Auger . Two observatories, one in the Northern and one in the Southern hemisphereare foreseen for the Pierre Auger Observatory project, to achieve a full explorationof the sky. The Southern Auger Observatory [10] is located near the small town ofMalarg¨ue in the province of Mendoza (Argentina) at the latitude of about 35 ◦ S andaltitude of 1400 above see level. The Observatory is a hybrid system, a combination ofa large surface array and a fluorescence detector. The surface detector (SD) is a largearray of 1600 water Cherenkov counters spaced at a distance of 1.5 km and covering atotal area of 3000 km . Each counter is a plastic tank of cylindrical shape with size 10m × . ◦ . The fluorescence detector (FD) consists of24 telescopes located in four stations which are built on the top of small elevations onthe perimeter of the site. The telescopes measure the shower development in the airby observing the fluorescence light. Each telescope has a 12 m spherical mirror withcurvature radius of 3.4 m and a camera with 440 photomultipliers. The field of viewof each telescope is 30 ◦ × ◦ . UV filters were used as in HiRes. The Southern AugerObservatory started to collect data in 2004. The Observatory and has been completedin Summer 2008. The Northern Auger Observatory which is now being designed willbe located in Colorado (USA). We note that the present Auger Observatory is theonly detector exploring the Southern hemisphere. Telescope Array (TA) . It is being built by a US - Japan - Korea collaboration2n Millard County, Utah, USA. Like the Auger Observatory, the TA is a hybriddetector [11]. It covers an area of 860 km and comprises 576 scintillator stationsand three FD sites on a triangle with about 35 km separation each equipped with 12fluorescence telescopes.Fig. 1 shows a comparison of the exposures accumulated by various experimentsat the end of 2007. More details about the exposure calculations can be found in [12,13]. It can be seen that the largest exposure has been achieved with the AugerObservatory and it will continue to deliver more than about 7000 km sr for eachyear of operation. TA is not shown in the figure since it started full operation sinceMarch 2008. It is important to notice the different behaviour with energy of theFigure 1: Accumulated exposures of various experiments at the end of 2007 (see ref. [12,13]). apertures for arrays of surface detectors and fluorescence detectors. In case of arraysof detectors with a regular pattern, the aperture can be calculated in a straightforward and model independent way, once the energy threshold for CR detection andreconstruction is exceeded. The situation is different for fluorescence telescopes. Here,the maximum distance out to which showers can be observed increases with increasingfluorescence light and thereby increasing energy. This condition makes the aperturecalculation dependent on Monte Carlo simulation and then on primary mass andon the hadronic interaction models employed. This dependence can be considerablyreduced by applying quality cuts to geometry parameters (e.g. the distance of theshower), but this is possible only if geometry is well determined as in the cases of3ybrid or stereo detection.
Most of the energy spectrum data available today at UHE are provided by AGASA,HiRes and Auger (see Fig. 2). The two last experiments recently published spectrumanalyses [4, 5] showing evidence of a flux suppression as expected by the GZK effectwith significances of about 5 and 6 σ respectively at slightly different energies (5.6and 4 × eV). Shifting the energy scale by about +15% for Auger and about -25%for AGASA with respect to HiRes the three spectra agree rather well up to about 5 × eV. At higher energies the AGASA data do not exhibit any flux suppressionand thus are inconsistent with the other data. Energy [eV/particle] ] e V - s r - s - J ( E ) [ m S c a l ed f l u x E AGASAHiRes IHiRes IIAuger PRL 2008
Figure 2:
Cosmic ray flux measurements (multiplied by E ) from AGASA [3], HiRes [4]and Auger [5]. Typical uncertainties of the energy scale are on the order of 20 ÷ Spectrum data from instruments with exposure less than 1000 km sr yr have not been consid-ered in this review. ◦ (S ) is usedas reference and the zenith angle dependence of the energy estimator is determinedassuming that the arrival directions are isotropically distributed. This procedure istraditionally called Constant Intensity Cut . The absolute calibration of S is derived lg(E FD /eV)18.5 19 19.5 / V E M ) l g ( S FD )/E FD (E - E -0.8 -0.4 -0 0.4 0.8 N u m b er o f E v e n t s Mean 0.01 ± ± lg(E/eV) - - . J / A x E -1-0.500.51 E (eV) · · · · Auger Hires I
Figure 3:
Left : Auger calibration of SD data: correlation between surface detector signaland FD energy. The fractional differences between the two energy estimators are inset.
Right : Fractional difference between Auger and HiRes I data relaltive to a spectrum withindex of 2.69. from the hybrid events using the calorimetric energy measured by the FD whichis then corrected for the missing energy using the mean value between proton andiron (uncertainty about 4% at 10 eV). This absolute calibration, which defines theenergy scale, is at present affected by a systematic error of about 20%, mainly due touncertainties on the fluorescence yield and on the calibration of the FD telescopes.The energy calibration, obtained from the subset of hybrid events (see Fig. 3) is thenused for the full set of events with higher statistics measured by the SD.5he flux suppression in Auger and HiRes as well as the possible difference in theirenergy scales is evident when plotting the fractional difference with respect to a powerlaw spectrum. Fig. 3, right panel, shows this fact for a spectral index of 2.69 whichis the one fitted by Auger below 4 × eV. lg(E/eV) ] < X m ax > [ g / c m Auger ICRC07HiRes ICRC07HiRes/MIAFlys Eye
QGSJETII−03QGSJET01SIBYLL2.1EPOS1.6 p r o t o n i r o n Figure 4: h X max i as a function of energy compared with proton and iron predictions usingdifferent hadronic interaction models. Measuring the composition of cosmic rays is crucial to obtain a full understandingof their acceleration processes, propagation and relation with galactic particles. Theatmospheric depth X max denotes the longitudinal position of the shower maximum,which is directly accessible with the FD. It grows logarithmically with the energyof the primary particle. The behaviour of X max for different primary particles likephotons, protons and heavier nuclei can be conceptually understood in the frameworkof the Heitler and superposition models [15], which provides good agreement withdetailed Monte Carlo simulations. New results based on HiRes-Stereo and Augerhybrid data at the ICRC [16, 17] are reported in Fig. 4. Both data sets agreevery well up to ∼ · eV but differ slightly at higher energies. The differencesbetween the two experiments are within the differences observed between p- and Fe-predictions for different hadronic interaction models. With these caveat kept in mind,6oth experiments observe an increasingly lighter composition towards the ankle. Athigher energies, the HiRes measurement yields a lighter composition than Auger.Another important issue concerning the primary composition is the search forphotons and neutrinos in EAS. The Auger Observatory has set new photon limits withboth the hybrid and SD detection methods [18, 19]. The new limits are compared toprevious results and to theoretical predictions in Fig. 5 for the photon fraction. Interms of the photon fraction, the current bound at 10 EeV approaches the percentlevel while previous bounds were at the 10 percent level. A discovery of a substantialphoton flux could have been interpreted as a signature of top-down (TD) models.In turn, the experimental limits now put strong constraints on these models. Forinstance, certain SHDM (Super Heavy Dark Matter) or TD models discussed in theliterature [20] predict fluxes that exceed the limits by a factor 10. [eV] E [ % ] P h o t o n F r a c t i o n f o r E > E SHDMSHDM’TDGZK Photons) Limit (E>E
A A A2HP HP AYY YFD limits at 95% CL
Neutrino Energy [eV] ] - s r - s - f( E ) [ G e V c m E -8 -7 -6 -5 -4 -3
10 AMANDA HiResHiResBaikal RICE’06GLUE’04 ANITA-liteFORTE’04GZK neutrinos
Auger integrated Auger differential
Figure 5:
Left : The upper limits on the fraction of photons in the integral cosmic rayflux derived from Auger SD (black arrows) along with previous experimental limits(HP: Haverah Park; A1, A2: AGASA; AY: AGASA-Yakutsk; Y: Yakutsk; FD: Augerhybrid limit). Also shown are predictions from top-down models (SHDM, SHDM’,TD) and predictions of the GZK photon fraction. For references see [19].
Right :Limits at the 90% C.L. for a diffuse flux of ν τ assuming a 1:1:1 ratio of the 3 neutrinoflavors and the expected flux of GZK neutrinos. For references see [21]Neutrino induced showers can be also identified if they occur deep in the atmo-sphere under large zenith angles, or by their special topology in the case of Earth-skimming tau neutrinos. Identification criteria have been developed to find EAS thatare generated by tau neutrinos emerging from the Earth. Auger has searched for tauneutrinos in the data collected up to August 2007. No candidates have been foundand an upper limit on the diffuse tau neutrino flux has been set. In Fig. 5 thisresult [21] is shown. 7 Arrival Directions
Most of the recent results are from the Auger Collaboration who have started adetailed investigation of the angular directions of the cosmic rays. While no excesshas been found from the Galactic Centre in the EeV energy range, evidence foranisotropy has been found in the extreme energy region.Observation of an excess from the region of the Galactic Centre in the EeV energyregion were reported by AGASA [22] and SUGAR [23]. The Auger Observatory issuitable for this study because the Galactic Centre (constellation of Sagittarius) lieswell in the field of view of the experiment. The angular resolution of the SD of Augerdepends on the number of tanks activated by the shower and it is better than onedegree at high-energy. However, with statistics much greater than previous data, theAuger search [24] does not show abnormally over-dense regions around the GC.A big step towards the discovery of the UHECR sources has been recently madeby the Pierre Auger Collaboration [6, 25]. The highest energy events recorded so farwere scanned for correlations with relatively nearby AGNs ( z ≤ .
024 correspondingto D ≤
100 Mpc) listed in the V´eron-Cetty/V´eron catalogue [27]. AGNs where usedonly up to a maximal redshift z max , which was a free parameter in the correlationscan. Two other free parameters were the minimal energy of the cosmic ray eventsE thr and the maximum separation between reconstructed cosmic ray direction andthe AGN position ψ . The scan was performed over data taken during the first twoyears of stable operation (01/2004 - 05/2006) and a significant minimum of the chanceprobability calculated assuming isotropic arrival directions was observed. After theparameters of this explorative scan ( z max = 0 . thr = 56 EeV , ψ = 3 . ◦ ) werefixed, the consecutive data set (06/2006-08/2007) was used to verify the correlationsignal and the hypothesis of an isotropic source distribution could be rejected at morethen 99% confidence level. A sky map of the 27 events above the energy thresholdof E thr = 56 EeV together with the selected AGN is shown in Fig. 6. Also shownare the events selected during a follow-up analysis of stereo data from the HiResCollaboration [26], which do not show a significant correlation.The interpretation of the observed anisotropy is ongoing and a much larger eventstatistics will be needed to investigate, for example, whether the AGNs act only astracers for the underlying true sources and whether the angular separation betweenAGN and UHECR can be related to magnetic deflections.
In recent years UHECRs have shown a variety of exciting features: the flux suppres-sion at energies as the one expected for the GZK cutoff and possible correlations withsources are the most attactive. The two phenomena are strictly related one to each8 entaurus A Virgo A Fornax A +30-30+60-600+60+120 -60 -120
Figure 6:
The sky seen with UHECRs with energy above 56 EeV detected with the sur-face array of Auger (red circles, [6, 25]) and with the HiRes detector in stereo mode (bluesquares, [26]) in galactic coordinates. Filled markers denote cosmic rays within 3.1 ◦ fromAGNs with redishift z < .
018 (black stars, [27]). The relative exposures of the two exper-iments are not shown for simplicity. Very roughly Auger (HiRes) is blind to a part of theleft (right) side of this plot and then their exposures are rather complementary. Detailedexposures can be found in the original papers. other. In particular the correlation scenario is compatible with suitable spectrumshapes and mass compositions in the GZK region. This because cosmic ray propaga-tion through galactic fields and their interactions with the photon background affectnot only directions, but also the energy and type of particles observed on Earth.Coming years are expected to be fruitful. New data will come from the NorthernHemisphere: Telescope Array, now, and Auger North, in a few years, will join thisfascinating exploration.
References [1] A. M. Hillas, Annu. Rev. Astron. Astrophys., , 425 (1984).[2] K. Greisen, Phys. Rev. Lett. , 748 (1966) and G.T. Zatsepin and V.A. Kuz’min,Sov. Phys. JETP Lett. (Engl. Transl.), , 78 (1966).[3] M. Takeda et al. [AGASA Collaboration], Astropart. Phys. , 447 (2003).[4] R.U. Abbasi et al. [HiRes Collaboration], Phys. Rev. Lett. , 101101 (2008).95] J. Abraham et al. [Pierre Auger Collaboration], Phys. Rev. Lett. , 061101(2008).[6] J. Abraham et al. [Pierre Auger Collaboration], Science , 938 (2007).[7] M. Nagano and A. A. Watson, Rev. Mod. Phys. , 689 (2000).[8] N. Chiba et al., Nucl. Instrum. Meth., A311 , 338 (1992).[9] T. Abu-Zayyad et al., Nucl. Instrum. Meth.
A450 , 253 (2000).[10] J. Abraham et al., Nucl. Instrum. Meth.
A523 , 50 (2004).[11] M. Fukushima et al. [Telescope Array Collaboration], Proc. 30th ICRC, M´erida,p. 955 (2007).[12] K.-H. Kampert, J. Phys. Conf. Ser., , 062002 (2008).[13] D. G. Bergman and J. W. Belz, J. Phys. G, R359 (2007).[14] D. Newton et al., Astropart. Phys. , 414 (2007).[15] J. Matthews, Astropart. Phys. , 387 (2005)[16] Y. Fedorova et al. [HiRes Collaboration], Proc. 30th ICRC, M´erida, p.1236(2007).[17] M. Unger et al. [Pierre Auger Collaboration], Proc. 30th ICRC, M´erida, p.594(2007).[18] J. Abraham et al. [Pierre Auger Collaboration], Astropart. Phys. , 155 (2007)[19] J. Abraham et al. [Pierre Auger Collaboration], Astropart. Phys. , 243 (2008).[20] References to theoretical models can be found e.g. in [19].[21] J. Abraham et al. [Pierre Auger Collaboration], Phys. Rev. Lett. , 211101(2008).[22] M. Teshima et al., Proc. 27th ICRC, Hamburg, vol. 1, p. 337 (2001).[23] J.A. Bellido et al., Astropart. Phys. , 167 (2001).[24] J. Abraham et al. [Pierre Auger Collaboration], Astropart. Phys. , 244 (2007).[25] J. Abraham et al. [Pierre Auger Collaboration], Astropart. Phys. , 288 (2008).[26] R.U. Abbasi et al. [HiRes Collaboration], arXiv eprints: 0804.0382v1 (2008).[27] M.-P. V´eron-Cetty, P. V´eron, Astron. Astrophys.455