9 0 Search for photons at the Pierre Auger Observatory 0 2 M. Rissea∗ for the Pierre Auger Collaboration n a aBergische Universit¨at Wuppertal, 42119 Wuppertal, Germany J 6 1 ThePierreAugerObservatoryhasauniquepotentialtosearchforultra-highenergyphotons(above∼1EeV). First experimentallimits on photonswere obtained duringconstruction of the southern part of theObservatory. ] Remarkably, already these limits have proven useful to falsify proposals about the origin of cosmic rays, and to E perform fundamental physicsbyconstraining Lorentzviolation. A finaldiscovery ofphotonsat theupperendof H theelectromagnetic spectrum is likely to impact various branchesof physicsand astronomy. . h p - o 1. Introduction 2. Search with hybrid data r t s Therearegoodreasonstosearchforultra-high ShowersinitiatedbyUHEphotonsdevelopdif- a energy(UHE,above∼1018 eV=1EeV)photons ferently from showers induced by nuclear pri- [ at the Auger Observatory [1]. An UHE photon maries. Particularly, observables related to the 1 producesa“normal”airshower,easilydetectable development stage or “age” of a shower (such as v 5 with a giant shower observatory. Still, such pho- the depth of shower maximum Xmax) and to the 2 ton showers have characteristics that make them content of shower muons provide good sensitiv- 5 welldistinguishablefromshowersinitiatedbypri- ity to identify primary photons. Photon showers 2 mary hadrons. The calculation of these photon areexpectedtodevelopdeeperintheatmosphere 1. showercharacteristicscanbedonewithhighcon- (larger Xmax), see Fig. 1. This is connected to 0 fidence because QED effects dominate, such that the smaller multiplicity in electromagnetic inter- 9 solid conclusions can be drawn from the data. actions compared to hadronic ones, such that a 0 A substantial flux of UHE photons is predicted larger number of interactions is required to de- : v in top-down models [2]. Even in conventional grade the energy to the critical energy where the i cosmic-ray scenarios, UHE photons may be pro- cascading process stops. Additionally, the LPM X duced at detectable level [3]; as their expected effect [7] results in a suppression of the pair pro- r a flux at Earth depends on uncertain parameters duction and bremsstrahlung cross-sections. Pho- such as source density, injection spectrum or ex- ton showers also contain fewer secondary muons, tragalactic radio background, findings on UHE since photoproduction and direct muon pair pro- photons will provide valuable astrophysics infor- ductionareexpectedtoplayonlyasub-dominant mation. role. Inthefollowing,thecurrentstatusofthesearch Using high-quality hybrid events (air showers forUHEphotonsatthePierreAugerObservatory registeredbyboththefluorescencetelescopesand is briefly summarized. Detailed descriptions of the ground array which fullfil certain strict re- the data analyses can be found in Refs. [4,5]. A quirements to be accepted for the analysis), it is general review on the search for UHE photons possible to compare directly the observed Xmax using air showers is given in Ref. [6]. values of UHE events with expectations for pho- ton primaries of same energy and arrival direc- tion. An example of a measured air shower and the comparison of its Xmax value to results from photon simulations is shown in Fig. 2. One can ∗CurrentlyatUniversityofSiegen,Germany 1 2 -2m) 1200 Fly´s Eye 2)) 35 Event 1687849 <>X (g cmax11901000000 HHYYCHSDPaaAIiiERRCAkkSGeeuuESAssRttE- ss-2MABkk-0V -L022IAUA400AI00LNR15CCOAABNICC photon pwhitoht opnreshower 15-0eV/(g cm 223050 EX m~a x1~6 7E8e0V g cm-2 800 TUNKA proton X (1 15 700 d 10 E/ 600 iron d 5 QGSJET 01 500 QGSJET II 0500 600 700 800 900 1000 1100 1200 400 SIBYLL 2.1 atmospheric depth X (g cm-2) 14 15 16 17 18 19 20 21 10 10 10 10 10 10 1E0lab (eV10) (cid:255) dXmax00..000078 Event 1687849: N/ d0.006 N 1/0.005 data photon simulation Figure 1. Average depth of shower maximum 0.004 <Xmax> versus energy simulated for primary 0.003 photons, protons and iron nuclei. Depending on 0.002 the specific particle trajectory through the geo- 0.001 magneticfield,photonsaboveafewtimes1019eV 0 700 800 900 1000 1100 1200 1300 can also create a preshower (see e.g. Ref. [8] and X (g cm-2) max references therein): as indicated by the splitting of the photon line, the averageXmax values then do not only depend on primary energy but also Figure 2. (Top) Example of a reconstructedlon- arrival direction. For nuclear primaries, calcula- gitudinal energy deposit profile (points) and the tionsfordifferenthadronicinteractionmodelsare fit by a Gaisser-Hillas function (line). (Bottom) displayed. Oneseesthatat10EeV,photonspen- etrate deeper into the atmosphere than hadron Xmax measured in the shower shown in the top primaries by (on average)∼200 g cm−2 or more. plot(pointwitherrorbar)comparedtotheXmγax distribution expected for photon showers (solid This is to be comparedwiththe resolutionofflu- orescencetelescopesoftypically∼20−30gcm−2. line). The observed Xmax is well below the val- ues assuming primary photons. (Figures taken (Figure taken from Ref. [4].) from Ref. [4].) see that the observedXmax is well below the val- ues assuming primary photons. It is not likely ing Auger ground array data, see below). More- thatthisairshowerwasinitiatedbyanUHEpho- over,theAugerhybriddatasetwasthefirst(and ton. so far only) one where the fluorescence technique Inspecting 29 such high-quality air showers with its direct observation of Xmax was applied with energies above 10 EeV, it turned out that to search for UHE photons. in all cases the photon predictions exceed by far the observed Xmax values. From this, an up- 3. Search with ground array data per limit of 16% (95% c.l.) on the cosmic-ray photon fraction could be obtained as published To exploit the larger number of UHE events in 2007 [4]. Despite the very small number of recorded by the ground array only (i.e. without events, this limit was the best one at that time requiring additional fluorescence telescope data), above 10 EeV (to be superseded by the limit us- two observables of the array detectors were cho- 3 sen in an analysis published in 2008 [5] which were found that exceed the Auger photon limits have significantly different behavior for nuclear by about a factor 10. Corresponding constraints primaries when compared to photons: the rise- onLorentzviolationparametersimproveprevious time of the recorded shower signal and the ra- constraintsbyseveralordersofmagnitudedueto dius of curvature of the showerfront(see Fig.3). the extreme energy in case of UHE photons. As an energy estimator, the total signal deposit S(1000) derived for a ground detector at 1000 m 5. Outlook core distance is used by comparing the measured value to photon shower simulations. Data taken during the first phase of construc- In Fig. 4, for each measured showera quantity tion of the southern Auger Observatory allowed isplottedasafunctionofenergyoftheeventthat the derivation of UHE photons limits that ap- combines the discriminating observables of rise- proach(intermsoffraction)the10−2 levelaten- time andradius of curvature. Also shownare ex- ergies above 10 EeV. Both detector components pectationsfromphotonshowersimulations. Ana (telescopes and array) offer excellent discrimina- priori cuthadbeendefinedsuchthatdatapoints tionpowerbetweenphotonandhadronprimaries lying above the mean of the photon distribution suchthatthesensitivitylevelwillcontinuetoim- are considered as photon candidates. Account- prove over the next years while the data accu- ingforinefficiencies,upper limitsonthepresence mulate. The upper range of predictions for GZK of photons in the primary cosmic-ray beam were photons(seeFig.5)aswellastheyetunexplored placed; in particular a limit of 2% (at 95% c.l.) energy range below 10 EeV can be tested soon. above10EeVwasderived(seeFig.5,lowerplot). With a main goal of full sky coverage, the Also,inthisanalysisthefirstdirectlimittothe AugerObservatoryistobecompletedbyanorth- fluxofUHEphotons(insteadofthefraction)was ernsite. Currentplansaimatasignificantly(fac- obtained (Fig. 5, upper plot). In case of ground tor ∼7) larger array to proceed with cosmic-ray arrays,theflux(oralimitontheflux)isthemore astronomy. With such an enlarged northern site, robust experimental quantity (see Refs. [5,6]). photonfractionsatorbelow the10−3 levelarein reach within few years of operation. Asforanyexploratorysearch,thetimescalefor 4. Implications the final discovery of UHE photons is uncertain. As can be seen from Fig. 5, particularly from However,eventheveryfirstresultsfromthepho- the comparison of experimental UHE photon ton searchatthe AugerObservatoryprovedvery flux limits and model predictions, contempo- helpful for cosmic-ray physics (e.g. discrimina- rary top-down models are severely constrained tion between different cosmic-ray source scenar- now: thenon-observationofUHEphotonsbythe ios) and for fundamental physics (test of Lorentz AugerObservatoryisdifficulttoexplaininexotic invariance). It seems reasonable to believe that physics scenarios (see e.g. also Refs. [9,10]). this ability of the UHE photon search to provide The photon limits also reduce systematic un- substantial physics results will persist. The un- certainties in the derivation of the total cosmic- precedented and presently unique sensitivity of ray flux spectrum [11], since the energies of pho- theAugerObservatorytoUHEphotons,particu- ton showers could be misreconstructed. As can larlywhencompletedbythenorthernsite,allows also be seen in Fig. 4, there are no indications of for the first time a realistic search for UHE pho- an increasing photon component towardshighest tons alsofrom conventionalcosmic-rayscenarios. energies at the present sensitivity level. A final discovery of UHE photons would open The photonbounds havealsoprovenuseful for a new − and the most extreme − “window” of fundamental physics. In Ref. [12], the fraction of photon astronomy. Experience shows that this is GZK photons expected at Earth was computed usually accompanied with radically new, and of- assuming that Lorentz invariance violation sup- tenunexpected,insightsaboutthebizarreinhab- presses pair production by UHE photons. Values itantsoftheuniverseandtheir(inter-)actions[13] 4 (forasurelyincompletelistofpossibleimpactsof a discovery of UHE photons on various research fields, see Ref. [6]). Acknowledgments. It is a pleasure to thank m] 24 Photon the CRIS organizers for an excellent and in- e [k 22 Data spiring conference. (The sometimes quite “Ger- atur 20 v man” weather made me feel at home even ur 18 C more.) Partial support from the German Min- of 16 s istry for Education and Research BMBF (Ver- diu 14 a bundforschungAstroteilchenphysik)andfromthe R 12 DFG are kindly acknowledged. 10 QGSJet 8 Sibyll REFERENCES 6 4 1. J. Abraham et al. (Pierre Auger Collabora- 2 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 tion),Nucl.Instrum.Meth.A523,50(2004). sec(q ) 2. P. Bhattacharjee, G. Sigl, Phys. Rep. 327, 109 (2000). m600 Photon k 3. [Gar.XGive:lamstirnoi-,phO/.07K06a.l2a1s8h1e]v;, D.VG. .SemiSkiogzl,, s] at 1 500 Data n Phys. Rev. D 75, 103001 (2007) [t1/2400 [arXiv:astro-ph/0703403]. 4. J. Abraham et al. (Pierre Auger Collabora- 300 Sibyll tion), Astropart. Phys. 27 (2007) 155 5. J. Abraham et al. (Pierre Auger Collabora- QGSJet 200 tion), Astropart. Phys. 29 (2008) 243 6. M. Risse,P.Homola,Mod.Phys.Lett. A 22, 100 749 (2007). 7. L.D. Landau, I.Ya. Pomeranchuk, Dokl. 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 sec(q ) Akad. Nauk SSSR 92, 535 & 735 (1953); A.B. Migdal, Phys. Rev. 103, 1811 (1956). 8. P. Homola et al., Comp. Phys. Comm. 173, 71 (2005) Figure 3. Parameterization of the mean behav- 9. R. Aloisio, F. Tortorici, Astropart. Phys. 29, ior of the radius of curvature R (upper plot) 307 (2008). and shower risetime at 1000 m core distance t1/2 10. M. Kachelrieß,arXiv:0810.3017v2 (lower plot) for 20 EeV primary photons as a 11. Pierre Auger Collaboration,Phys. Rev. Lett. function of the zenith angle using two different 101, 061101 (2008) hadronic interaction models. An increase (a de- 12. M. Galaverni, G. Sigl, Phys. Rev. Lett. 100, crease) of R (of t1/2) with zenith angle is ex- 021102 (2008). pectedduetothegenerallylongerpathlengthsto 13. A. Lawrence, arXiv:0704.0809 ground in case of larger inclination. Real events of 19–21 EeV (photon energy scale) are added. The significant deviation of the observed values from those expected for primary photons is visi- ble. (Figures taken from Ref. [5].) 5 -1]yr limits at 95% CL SHDM -1r SHDM’ -2sm10-1 TZD Burst k GZK Photons [E0 A Limit (E>E0) > E n10 or eviatio 8 MDaCt aPhotons n Flux f10-2 ent D 6 Photo on 4 p m 10-3 o 2 C al p 0 nci 1019 1020 Pri-2 E0 [eV] -4 -6 [%]E0100 HPlimits at 95% CL A A2 -8 E> HP r AY -101 1.2 1.4 1.6 1.8 2 2.2 on fo A Y Log(Energy/EeV) Fracti 10 FD Y SSTHHDDDMM’ n Z Burst Photo LGiZmKi tP h(oEt>oEns0) Figure4. Plottedisaquantitythatcombinesthe measurements of radius of curvature and shower risetime (cf. Fig. 3) for data (black crosses) and 1 photon simulations (open red circles) as a func- tionoftheprimaryenergy(photonenergyscale). Data lying above the dashed line, which indi- 1019 1020 E [eV] 0 cates the mean of the distribution for photons, are taken as photon candidates. No event meets this requirement. Moreover,notrend isvisible of higher-energyevents becoming more photon-like. Figure 5. The upper limits onthe integralflux of (Figure taken from Ref. [5].) photons (top) and on the fraction of photons in theintegralcosmic-rayflux(bottom)asafunction ofthethresholdenergyasmeasuredbytheAuger arraydetector(blackarrows)alongwithprevious experimental limits and model predictions. (Fig- ures taken from Ref. [5].)