Diffuse Galactic gamma-ray emission with H.E.S.S
H.E.S.S. Collaboration, A. Abramowski, F. Aharonian, F. Ait Benkhali, A.G. Akhperjanian, E.O. Angüner, M. Backes, S. Balenderan, A. Balzer, A. Barnacka, Y. Becherini, J. Becker Tjus, D. Berge, S. Bernhard, K. Bernlöhr, E. Birsin, J. Biteau, M. Böttcher, C. Boisson, J. Bolmont, P. Bordas, J. Bregeon, F. Brun, P. Brun, M. Bryan, T. Bulik, S. Carrigan, S. Casanova, P.M. Chadwick, N. Chakraborty, R. Chalme-Calvet, R.C.G. Chaves, M. Chrétien, S. Colafrancesco, G. Cologna, J. Conrad, C. Couturier, Y. Cui, I.D. Davids, B. Degrange, C. Deil, P. deWilt, A. Djannati-Ataï, W. Domainko, A. Donath, L.O'C. Drury, G. Dubus, K. Dutson, J. Dyks, M. Dyrda, T. Edwards, K. Egberts, P. Eger, P. Espigat, C. Farnier, S. Fegan, F. Feinstein, M.V. Fernandes, D. Fernandez, A. Fiasson, G. Fontaine, A. Förster, M. Füßling, S. Gabici, M. Gajdus, Y.A. Gallant, T. Garrigoux, G. Giavitto, B. Giebels, J.F. Glicenstein, D. Gottschall, M.-H. Grondin, M. Grudzińska, D. Hadasch, S. Häffner, J. Hahn, J. Harris, G. Heinzelmann, G. Henri, G. Hermann, O. Hervet, A. Hillert, J.A. Hinton, W. Hofmann, P. Hofverberg, M. Holler, D. Horns, A. Ivascenko, A. Jacholkowska, C. Jahn, M. Jamrozy, M. Janiak, F. Jankowsky, I. Jung-Richardt, M.A. Kastendieck, K. Katarzyński, U. Katz, S. Kaufmann, B. Khélifi, M. Kieffer, et al. (127 additional authors not shown)
aa r X i v : . [ a s t r o - ph . H E ] N ov Diffuse Galactic gamma-ray emission with H.E.S.S.
H.E.S.S. Collaboration, A. Abramowski , F. Aharonian , , , F. Ait Benkhali , A.G. Akhperjanian , , E.O. Ang¨uner , M. Backes ,S. Balenderan , A. Balzer , A. Barnacka , , Y. Becherini , J. Becker Tjus , D. Berge , S. Bernhard , K. Bernl¨ohr , ,E. Birsin , J. Biteau , , M. B¨ottcher , C. Boisson , J. Bolmont , P. Bordas , J. Bregeon , F. Brun , P. Brun , M. Bryan ,T. Bulik , S. Carrigan , S. Casanova , , P.M. Chadwick , N. Chakraborty , R. Chalme-Calvet , R.C.G. Chaves , M. Chr´etien ,S. Colafrancesco , G. Cologna , J. Conrad , , C. Couturier , Y. Cui , I.D. Davids , , B. Degrange , C. Deil , P. deWilt ,A. Djannati-Ata¨ı , W. Domainko , A. Donath , L.O’C. Drury , G. Dubus , K. Dutson , J. Dyks , M. Dyrda , T. Edwards ,K. Egberts , ∗ P. Eger , P. Espigat , C. Farnier , S. Fegan , F. Feinstein , M.V. Fernandes , D. Fernandez , A. Fiasson ,G. Fontaine , A. F¨orster , M. F¨ußling , S. Gabici , M. Gajdus , Y.A. Gallant , T. Garrigoux , G. Giavitto , B. Giebels ,J.F. Glicenstein , D. Gottschall , M.-H. Grondin , M. Grudzi´nska , D. Hadasch , S. H¨affner , J. Hahn , J. Harris ,G. Heinzelmann , G. Henri , G. Hermann , O. Hervet , A. Hillert , J.A. Hinton , W. Hofmann , P. Hofverberg , M. Holler ,D. Horns , A. Ivascenko , A. Jacholkowska , C. Jahn , M. Jamrozy , M. Janiak , F. Jankowsky , I. Jung-Richardt ,M.A. Kastendieck , K. Katarzy´nski , U. Katz , S. Kaufmann , B. Kh´elifi , M. Kieffer , S. Klepser , D. Klochkov ,W. Klu´zniak , D. Kolitzus , Nu. Komin , K. Kosack , S. Krakau , F. Krayzel , P.P. Kr¨uger , H. Laffon , G. Lamanna ,J. Lefaucheur , V. Lefranc , A. Lemi`ere , M. Lemoine-Goumard , J.-P. Lenain , T. Lohse , A. Lopatin , C.-C. Lu ,V. Marandon , A. Marcowith , R. Marx , G. Maurin , N. Maxted , M. Mayer , T.J.L. McComb , J. M´ehault , , P.J. Meintjes ,U. Menzler , M. Meyer , A.M.W. Mitchell , R. Moderski , M. Mohamed , K. Mor˚a , E. Moulin , T. Murach , M. de Naurois ,J. Niemiec , S.J. Nolan , L. Oakes , H. Odaka , S. Ohm , B. Opitz , M. Ostrowski , I. Oya , M. Panter , R.D. Parsons ,M. Paz Arribas , N.W. Pekeur , G. Pelletier , P.-O. Petrucci , B. Peyaud , S. Pita , H. Poon , G. P¨uhlhofer , M. Punch ,A. Quirrenbach , S. Raab , I. Reichardt , A. Reimer , O. Reimer , † M. Renaud , R. de los Reyes , F. Rieger , C. Romoli ,S. Rosier-Lees , G. Rowell , B. Rudak , C.B. Rulten , V. Sahakian , , D. Salek , D.A. Sanchez , A. Santangelo ,R. Schlickeiser , F. Sch¨ussler , A. Schulz , U. Schwanke , S. Schwarzburg , S. Schwemmer , H. Sol , F. Spanier , G. Spengler ,F. Spies , L. Stawarz , R. Steenkamp , C. Stegmann , , F. Stinzing , K. Stycz , I. Sushch , , J.-P. Tavernet , T. Tavernier ,A.M. Taylor , R. Terrier , M. Tluczykont , C. Trichard , K. Valerius , C. van Eldik , B. van Soelen , G. Vasileiadis , J. Veh ,C. Venter , A. Viana , P. Vincent , J. Vink , H.J. V¨olk , F. Volpe , M. Vorster , T. Vuillaume , S.J. Wagner , P. Wagner ,R.M. Wagner , M. Ward , M. Weidinger , Q. Weitzel , R. White , A. Wierzcholska , P. Willmann , A. W¨ornlein ,D. Wouters , R. Yang , V. Zabalza , , D. Zaborov , M. Zacharias , A.A. Zdziarski , A. Zech , H.-S. Zechlin , and Y. Fukui Universit¨at Hamburg, Institut f¨ur Experimentalphysik,Luruper Chaussee 149, D 22761 Hamburg, Germany Max-Planck-Institut f¨ur Kernphysik, P.O. Box 103980, D 69029 Heidelberg, Germany Dublin Institute for Advanced Studies, 31 Fitzwilliam Place, Dublin 2, Ireland National Academy of Sciences of the Republic of Armenia,Marshall Baghramian Avenue, 24, 0019 Yerevan, Republic of Armenia Yerevan Physics Institute, 2 Alikhanian Brothers St., 375036 Yerevan, Armenia Institut f¨ur Physik, Humboldt-Universit¨at zu Berlin, Newtonstr. 15, D 12489 Berlin, Germany University of Namibia, Department of Physics, Private Bag 13301, Windhoek, Namibia University of Durham, Department of Physics, South Road, Durham DH1 3LE, U.K. GRAPPA, Anton Pannekoek Institute for Astronomy, University of Amsterdam,Science Park 904, 1098 XH Amsterdam, The Netherlands Obserwatorium Astronomiczne, Uniwersytet Jagiello´nski, ul. Orla 171, 30-244 Krak´ow, Poland now at Harvard-Smithsonian Center for Astrophysics,60 Garden St, MS-20, Cambridge, MA 02138, USA Department of Physics and Electrical Engineering, Linnaeus University, 351 95 V¨axj¨o, Sweden Institut f¨ur Theoretische Physik, Lehrstuhl IV: Weltraum und Astrophysik,Ruhr-Universit¨at Bochum, D 44780 Bochum, Germany GRAPPA, Anton Pannekoek Institute for Astronomy and Institute of High-Energy Physics,University of Amsterdam, Science Park 904, 1098 XH Amsterdam, The Netherlands Institut f¨ur Astro- und Teilchenphysik, Leopold-Franzens-Universit¨at Innsbruck, A-6020 Innsbruck, Austria Laboratoire Leprince-Ringuet, Ecole Polytechnique, CNRS/IN2P3, F-91128 Palaiseau, France now at Santa Cruz Institute for Particle Physics, Department of Physics,University of California at Santa Cruz, Santa Cruz, CA 95064, USA Centre for Space Research, North-West University, Potchefstroom 2520, South Africa LUTH, Observatoire de Paris, CNRS, Universit´e Paris Diderot, 5 Place Jules Janssen, 92190 Meudon, France LPNHE, Universit´e Pierre et Marie Curie Paris 6,Universit´e Denis Diderot Paris 7, CNRS/IN2P3,4 Place Jussieu, F-75252, Paris Cedex 5, France Institut f¨ur Astronomie und Astrophysik, Universit¨at T¨ubingen, Sand 1, D 72076 T¨ubingen, Germany Laboratoire Univers et Particules de Montpellier,Universit´e Montpellier 2, CNRS/IN2P3, CC 72,Place Eug`ene Bataillon, F-34095 Montpellier Cedex 5, France DSM/Irfu, CEA Saclay, F-91191 Gif-Sur-Yvette Cedex, France Astronomical Observatory, The University of Warsaw, Al. Ujazdowskie 4, 00-478 Warsaw, Poland Instytut Fizyki J¸adrowej PAN, ul. Radzikowskiego 152, 31-342 Krak´ow, Poland School of Physics, University of the Witwatersrand,1 Jan Smuts Avenue, Braamfontein, Johannesburg, 2050 South Africa Landessternwarte, Universit¨at Heidelberg, K¨onigstuhl, D 69117 Heidelberg, Germany Oskar Klein Centre, Department of Physics, Stockholm University,Albanova University Center, SE-10691 Stockholm, Sweden Wallenberg Academy Fellow, School of Chemistry & Physics, University of Adelaide, Adelaide 5005, Australia APC, AstroParticule et Cosmologie, Universit´e Paris Diderot,CNRS/IN2P3, CEA/Irfu, Observatoire de Paris, Sorbonne Paris Cit´e,10, rue Alice Domon et L´eonie Duquet, 75205 Paris Cedex 13, France Univ. Grenoble Alpes, IPAG, F-38000 Grenoble, FranceCNRS, IPAG, F-38000 Grenoble, France Department of Physics and Astronomy, The University of Leicester,University Road, Leicester, LE1 7RH, United Kingdom Nicolaus Copernicus Astronomical Center, ul. Bartycka 18, 00-716 Warsaw, Poland Institut f¨ur Physik und Astronomie, Universit¨at Potsdam,Karl-Liebknecht-Strasse 24/25, D 14476 Potsdam, Germany Laboratoire d’Annecy-le-Vieux de Physique des Particules,Universit´e de Savoie, CNRS/IN2P3, F-74941 Annecy-le-Vieux, France DESY, D-15738 Zeuthen, Germany Universit´e Bordeaux 1, CNRS/IN2P3, Centre d’ ´Etudes Nucl´eaires de Bordeaux Gradignan, 33175 Gradignan, France Universit¨at Erlangen-N¨urnberg, Physikalisches Institut,Erwin-Rommel-Str. 1, D 91058 Erlangen, Germany Centre for Astronomy, Faculty of Physics, Astronomy and Informatics,Nicolaus Copernicus University, Grudziadzka 5, 87-100 Torun, Poland Funded by contract ERC-StG-259391 from the European Community, Department of Physics, University of the Free State, PO Box 339, Bloemfontein 9300, South Africa GRAPPA, Institute of High-Energy Physics, University of Amsterdam,Science Park 904, 1098 XH Amsterdam, The Netherlands and Department of Astrophysics, Nagoya University, Chikusa-ku, Nagoya 464-8602, Japan
Diffuse γ -ray emission is the most prominent observable signature of celestial cosmic-ray interactions at high energies. While already being investigated at GeV energies overseveral decades, assessments of diffuse γ -ray emission at TeV energies remain sparse. Aftercompletion of the systematic survey of the inner Galaxy, the H.E.S.S. experiment is in aprime position to observe large-scale diffuse emission at TeV energies. Data of the H.E.S.S.Galactic Plane Survey are investigated in regions off known γ -ray sources. Corresponding γ -ray flux measurements were made over an extensive grid of celestial locations. Longitudinaland latitudinal profiles of the observed γ -ray fluxes show characteristic excess emission notattributable to known γ -ray sources. For the first time large-scale γ -ray emission along theGalactic Plane using imaging atmospheric Cherenkov telescopes has been observed. Whilethe background subtraction technique limits the ability to recover modest variation on thescale of the H.E.S.S. field of view or larger, which is characteristic of the inverse Comptonscatter-induced Galactic diffuse emission, contributions of neutral pion decay as well asemission from unresolved γ -ray sources can be recovered in the observed signal to a largefraction. Calculations show that the minimum γ -ray emission from π -decay represents asignificant contribution to the total signal. This detection is interpreted as a mix of diffuseGalactic γ -ray emission and unresolved sources. PACS numbers: 95.85.Pw, 98.70.Sa, 98.38.Cp
I. INTRODUCTION
Cosmic rays permeate our Galaxy and thereby undergointeractions, producing amongst other particles diffuse γ -rays at high energies. Interactions capable of producing ∗ [email protected] † [email protected] γ -rays are the production and subsequent decay of neu-tral pions in the interstellar medium, inverse Comptonscattering on radiation fields and bremsstrahlung. Eachof these processes contributes differently depending onenergy and line-of-sight integrated densities of matter orradiation fields. Diffuse γ -ray emission was observed firstby SAS-2 [10] and further investigated by COS-B [19] andEGRET [14]. The most recent and detailed survey of the γ -ray sky employed the Fermi-LAT instrument studyingthe energy range between 200 MeV and 100 GeV [3]. Atthese energies, the diffuse emission constitutes the princi-pal component of the γ -ray sky, and represents emissionoriginating from cosmic-ray interactions, not dominatedby unresolved sources [23].Towards higher energies, the γ -ray flux from resolvedsources increasingly dominates the total observed celes-tial γ -ray emission. Accordingly, the respective signalat these energies contains a potentially large fraction ofunresolved γ -ray sources. At energies close to ∼ E >
100 GeV) γ -rays are inverse Compton scatteringand neutral pion decay [7]. In this energy regime, obser-vations of diffuse γ -ray emission have been reported bythe Milagro experiment [1] at a median energy of 15 TeV,and also by ARGO-YBJ [18]. These experiments oper-ate under favourable duty cycles and observe large fieldsof the sky, yet they are limited in γ -hadron separationquality as well as angular resolution, which further com-plicates conclusive discrimination between γ -ray sourcesand diffuse emission signatures. Profiting from a substan-tially lower energy threshold and arc-minute scale angu-lar resolution, imaging atmospheric Cherenkov telescopeshave the potential to improve substantially on these mea-surements. Particularly the High Energy StereoscopicSystem (H.E.S.S.) is privileged due to the comparativelylarge field of view (5 ◦ in diameter) and its location inNamibia, which allows for an excellent view on the cen-tral part of the Galactic Plane.Presented here is a first study of the diffuse γ -rayemission utilizing the imaging atmospheric Cherenkovtechnique. Problems arising in this measurement andmethodological limitations imposed by the technique arediscussed. The resulting signal is interpreted with re-spect to cosmic-ray interactions via neutral pion decay,inverse Compton scattering, and contributions of unre-solved sources. II. DATA AND ANALYSIS METHODOLOGYA. The H.E.S.S. Galactic Plane Survey
H.E.S.S. is a system of imaging atmospheric Cherenkovtelescopes in the Khomas highland of Namibia [5]. A sys-tem of four telescopes has been taking data since 2003.Since 2012 H.E.S.S. advanced to its second phase featur-ing a central telescope with a sixfold mirror area com-pared to the original 12 m diameter telescopes.A substantial part of the H.E.S.S. I data set is theH.E.S.S. Galactic Plane Survey (HGPS), which was ac-cumulated over the past 10 years with the four tele-scope system. Although having revealed a wealth ofnew sources [12], and already reported extended emis-sion from the Galactic Center ridge [6], a measurementof the large-scale diffuse γ -ray emission remains challeng-ing regarding sensitivity and analysis methodology. B. Analysis of the Galactic Plane Survey
Our investigation of diffuse emission relates to thoseregions where no γ -ray sources are detected.As such an analysis aims at discovery of a very faintsignal, the application of a sensitive analysis methodis required. The results presented here are obtainedusing semi-analytical modeling of the air shower pro-duced in γ -ray interactions in the atmosphere, result-ing in an improved sensitivity compared to conventionalanalysis methods [13]. Good-quality data with point-ings in the region between − . ◦ and 62 . ◦ in Galac-tic longitude l and between − . ◦ and 4 . ◦ in Galacticlatitude b are used for a measurement in the range of − ◦ < l < ◦ and − ◦ < b < ◦ , where the expo-sure of the HGPS is largest. The data amount to a totalof 2484 . C. Background Subtraction
Hadrons and electrons can produce air showers thatlook like γ -ray showers and it is therefore necessary tosubtract remaining background events that survive eventselection cuts. The standard procedure to avoid system-atic effects caused by changing atmospheric or instru-ment conditions, is to determine the background levelfrom data, from the same field of view, in regions of noknown γ -ray sources [8]. However, this method placesconstraints on the size of the emission region that can beprobed, which has to be smaller than the field-of-view toallow for background estimation regions. Special atten-tion needs to be paid to the selection of celestial regionsapplicable to a background measurement. In order to de-fine the regions excluded from background subtraction,an iterative procedure is adopted. At each step, a sig-nificance map of the Galactic Plane region is computedusing the ring background technique [12] with an over-sampling radius of 0 . ◦ (suitable for slightly extendedsources). The following exclusion conditions apply: Eachpixel with a significance s above 4 σ with at least oneneighboring pixel with s > . σ is excluded and viceversa. In order to include also tails in the PSF used todescribe the γ -ray sources, the obtained exclusion regionsare extended by 0 . ◦ . This procedure is repeated untilthe significance distribution of the non-excluded pixelshas a normal shape with | µ | < .
05 and w < . µ and The pixel size in the maps is 0 . ◦ × . ◦ . w being the mean and the width of the distribution re-spectively). The resulting excluded regions are visualizedby the dark areas in Fig. 1. In addition, the complete re-gion along the Galactic Plane with a latitude range of − . ◦ < b < . ◦ is excluded (visualized by the horizon-tal dashed lines in Fig. 1). The choice of the latituderange is a compromise between a desired large excludedregion in order to avoid contamination of the backgroundestimate on the one hand and the need for statistics andreduction of systematics in the background measurementon the other hand. An adaptive ring background subtrac-tion method has been chosen [12] to allow for optimalchoices of background regions.A consequence of the applied background subtraction isthat the method used is rather insensitive to large-scaleemission with modest variation in latitudinal intensitybecause such signals are subtracted along with the back-ground. The observed signal therefore needs to be inter-preted as excess relative to the γ -ray emission at absolutelatitudes exceeding | b | = 1 . ◦ . D. Generation of flux maps
For the region of − ◦ < l < ◦ and − ◦ < b < ◦ a map of the differential flux normalization at 1 TeVis obtained from the background-subtracted γ -ray ex-cess map by division by the integrated exposure map: φ = n γ / P A int t obs . The exposure is summed over indi-vidual observation positions, with integrated acceptance A int and dead-time corrected observation time t obs . Theintegrated acceptance is obtained from simulations andrequires a spectral assumption, which is a powerlaw withspectral index of 2 .
2. The result turns out to be onlyweakly sensitive to the choice of spectral index (with de-viations in regions off known γ -ray sources of less than5% when altering the spectral index assumption to 2 . E. Definition of the Analysis Regions
In the following sections total flux distributions arecompared with those of regions that do not contain sig-nificantly detected γ -ray sources. These regions are la-belled diffuse analysis region (DAR) and are defined inthe same way as regions suitable for background mea-surements. The DAR is shown in Fig. 1.As the Galactic Plane contains a large number of ex-tended sources (including those with complex morphol-ogy), the percentage of regions excluded from the DARamounts to 20%, whereas in the latitudinal region − . ◦ < b < . ◦ this percentage increases to more than40%. F. Profile Generation
For an investigation of the distribution of γ -ray fluxprofiles in Galactic longitude and latitude are generated.These profiles are obtained by integrating the flux mapover either longitude or latitude and by normalizing tothe covered area, thus resulting in an average flux pro-file for the latitudinal and longitudinal region considered.This procedure is done once for the complete data setand once for the DAR. The resulting profiles including1 σ uncertainties are shown in Figs. 1 and 2. III. RESULTSA. Spatial characteristics of the signal
The longitudinal profile in Fig. 1 shows a spiky distri-bution of the Galactic γ -ray sources for the complete re-gion (middle panel). For the DAR (bottom panel) fluxesare on average positive (although hardly significant inmost individual bins). For the signal, a clear correlationwith the distribution of the excluded regions in the DARcan be seen: excess is observed only in longitude rangeswith sparse exclusion of regions at small latitudes. Zeroor even mildly negative fluxes are found when large re-gions close to the Galactic Equator are excluded fromthe DAR. The reason for this is an over-subtraction ofthe background determined from signal-contaminated re-gions. The reflection of the shape of the DAR in the lon-gitudinal profile strongly limits its potential in terms ofa physics interpretation. However, it can be seen thatthe signal does not originate from left-over contributionsof excluded sources but it rather accumulates over longi-tude.The latitudinal profiles of both the complete data set andthe DAR, shown in Fig. 2, exhibit a clear excess overzero. The significance of the detected signal has beenevaluated by comparing the observed latitudinal profileto a zero-flux baseline hypothesis as a function of Galacticlatitude. Whereas the full latitudinal profile ( | b | ≤ ◦ )has been found to deviate by more than 6 σ from thenull hypothesis, the significance increases to above 20 σ in those latitudinal ranges that are close to the GalacticEquator.The latitudinal profile exhibits a maximum of around3 × − TeV − s − cm − sr − , at a latitude slightlyshifted from the Galactic Equator towards negative val-ues ( b max ≈ − . ◦ ). This is similarly observed for thetotal flux, which contains all γ -ray sources in addition tothe diffuse emission. The distribution falls off towardshigher latitude values and reaches zero flux at latitude of b ≈ ± ◦ . As a consequence of the applied backgroundsubtraction, slightly negative fluxes can be observed at b ≈ ± . ◦ (see previous discussion). In comparison withthe total flux, the signal in the DAR makes up ∼
28% inthe central 2 ◦ region. A large-scale signal prevails in thelongitudinal profile regardless of the details in the defini- l [deg]-420.02 -284.98 b [ d e g ] -2-1012 -60-40-200204060-420 -400 -380 -360 -340 -320 -300 ] - s r - T e V - s - F l u x [ c m -9 × ] - s r - T e V - s - F l u x [ c m -2024 -9 × FIG. 1. Top panel: The white regions depict the diffuse analysis region (DAR). Black are regions of significant γ -ray emission.Horizontal dashed lines mark the region − . ◦ < b < . ◦ that is excluded from background subtraction. Middle panel: Thelongitudinal profile of the Galactic Plane over a latitude range of − ◦ < b < ◦ . Shown is the differential flux at 1 TeV includingsources. H.E.S.S. TeV data, which include known sources, are indicated by black crosses. The minimal 1 TeV γ -ray emissionfrom hadronic interactions, estimated using HI and H data (traced by CO data) and a solar-like cosmic-ray spectrum (seetext), is shown as model curve. The dashed line includes a nuclear enhancement factor of 2.1. Model curves do not comprise areduction due to background subtraction. Bottom panel: The same as the middle panel, except only the DAR is considered.The distribution is strongly influenced by the shape of the DAR ( cf. top panel). Model curves correspond to the minimalhadronic γ -ray emission expected in the same region. tion of the DAR.Systematic uncertainties enter at several stages in thepresented analysis. A comparison with an independentcross-check analysis with separate calibration proceduresof the data resulted in consistency within ∼
30% in fluxnormalisation. This uncertainty, however, does not ac-count for the effect of a reduced signal due to the appliedbackground subtraction, which is present in both anal-ysis chains. The influence of the background removaltechnique can be determined under a model assumptionfor the γ -ray emission: a Gaussian of width 2 ◦ in lati-tude results in a reduction of 30% of the original signal,a Gaussian of width 20 ◦ in a reduction of 95%. B. Assessment of the Detection
The observed signal can originate from hadronic emis-sion of cosmic-ray interactions with matter via π pro-duction and decay, inverse Compton scattering of cosmic- ray electrons off radiation fields and unresolved γ -raysources. The contributions of these possible origins arediscussed in the following sections.
1. Hadronic emission
The component of hadronic emission is constrained bythe level of cosmic rays and the total target material.For an estimation of the minimum, guaranteed contribu-tion to be present in our observed signal, the emissionfrom the sea of cosmic rays (assumed to resemble thelocally measured cosmic-ray spectrum) interacting withgas content (indicated from respective spectral line ob-servations) is calculated. This minimum γ -ray emissionrelated to hadronic gas interactions undergoes the samespatial selection as H.E.S.S. flux maps for the produc-tion of profiles. Results of these calculations are shownin Figs. 1 and 2, together with the H.E.S.S. data, as redmodel curves. b [deg]-2 -1.5 -1 -0.5 0 0.5 1 1.5 2 ] - s r - T e V - s - F l u x [ c m -2024681012 -9 × b [deg]-2 -1.5 -1 -0.5 0 0.5 1 1.5 2 ] - s r - T e V - s - F l u x [ c m -2-1012345 -9 × FIG. 2. Top panel: The latitudinal profile of the GalacticPlane over a longitude range of − ◦ < l < ◦ . Shown isthe differential flux at 1 TeV including sources. H.E.S.S. TeVdata, which include known sources, are indicated by blackcrosses. The minimal 1 TeV γ -ray from hadronic interactions,estimated using HI and H data (traced by CO data) and asolar-like cosmic-ray spectrum (see text), is shown as modelcurve. The dashed line includes a nuclear enhancement factorof 2.1. Model curves do not comprise a reduction due tobackground subtraction. Bottom panel: The same as the toppanel, except only the DAR (for the definition see Fig. 1 toppanel) is considered. Model curves correspond to the minimalhadronic γ -ray emission expected in the same region. Gas templates of HI and H column densities are usedfor the calculation: HI data originate from the Lei-den/Argentine/Bonn Survey [16], a column density is ob-tained assuming a spin temperature of T S = 125 K. TheH column density is traced by CO (1-0) measured bythe NANTEN telescope. The conversion factor is chosento be X CO = 2 · cm − K − km − s [11]. Since thedegeneracy between HI, H and dust-related tracers forenergetic γ -ray emission is not yet satisfactorily resolvedat lower energies - where the majority of all observedphotons is attributed to diffuse Galactic emission [2, 4]- an additional dust-related (dark gas) component is notconsidered here.The minimum expected γ -ray flux is obtained from in-tegrating the product of the gas column density n ( l, b ),the interaction cross section dσ CR −→ γ dE CR , and the cosmic-rayenergy spectrum J ( E CR ) [9] over energy: dF ( l, b ) dE γ = Z dσ CR −→ γ dE CR n ( l, b ) J ( E CR ) dE CR . The parametrization of the interaction cross section fol- lows Kelner et al. [17]. H is treated as two individualprotons. For a conservative minimum in the calculated γ -ray emission, the proton cross section is applied also forheavier cosmic-ray nuclei. A nuclear enhancement fac-tor accounting for contributions of nucleonic cosmic-rayinteractions (beyond proton-proton) to the diffuse γ -rayemission is model-dependent but typically considered inthe range of 1.5 to 2 (see [20] and references therein). InFigs. 1 and 2 the corresponding flux according to a morerecent estimate of ≈ . ◦ . The H.E.S.S. data exhibits a narrower width of 1 ◦ for the total flux including γ -ray sources, while the pro-file of the DAR has a FWHM of 1 . ◦ - slightly broader,which could hint at a composite origin of the DAR sig-nal, consisting of both γ -ray sources and hadronic diffuseemission. Considering the fraction of the hadronic contri-bution, the minimum estimated from p-gas interactionsin the range of − ◦ < b < ◦ is 9% for the total flux and26% for the DAR. These values increase to 19% (total)and 55% (DAR) when considering the nuclear enhance-ment factor. The background subtraction that is appliedto the H.E.S.S. data reduces the detectable γ -ray emis-sion by around a third, yielding fractions of 14% (total)and 36% (DAR) for the hadronic contribution in the re-spective signal.
2. Large-scale inverse Compton emission
Another major contribution to the diffuse emission sig-nal at very high energies is predictably related to contin-uous cosmic-ray electron and positron energy losses viainverse Compton (IC) scattering. Both existence and rel-evance of an IC-emission contributing to an observablediffuse emission signal can be deduced from the imme-diately neighboring energy band, the Galactic diffuse γ -ray emission at GeV energies. Studies of the Galacticdiffuse emission in the Fermi-LAT energy range [3] indi-cated contributions by IC-scattering to the total observeddiffuse emission with an intensity up to the same orderof the pionic emission component. More specifically, IC-related γ -ray emission was reported at similar intensity tothe hadronic γ -ray emission produced from gas traced byHI for high Galactic latitudes, and dominant above tensof GeV [3]. Spectral extrapolation is suggestive of bothhadronic and IC-related emission components extendingtowards even higher energies before either energy lossessoften or even cut-off the IC-spectrum, or the neutralpion production spectrum might indicate the imprint ofthe maximum energy reached by particle acceleration inour Galaxy. At first glance, the IC-emission componentused to interpret the Fermi-LAT detected diffuse Galac-tic emission might serve as a reasonable template for suchan extrapolation. Respective predictions were derived onthe basis of a 2D slab model by the GALPROP cosmic-ray propagation code and an interstellar radiation fieldmodel [3] available alongside. The intensity distributionis generally smooth, with mild gradients along the Galac-tic Plane and comparably steep gradients towards higherGalactic latitudes. The latitudinal intensity profiles aresignificantly more extended compared to those derivedfrom the gas template. The present analysis frameworkallows only for partial recovery of such gradients. Forthe case of a GALPROP-based prediction, the aforemen-tioned background subtraction would yield a reductionof ∼
95% of the celestial IC flux.More realistic predictions for an IC-component at TeVenergies are not expected to resemble the smoothnessfrom lower GeV energies since energy losses increasinglyconfine the emission to local sources or source regions.Accordingly, an imprint from sources as well as Galacticstructure is anticipated [3]. Developments beyond limita-tions of present diffuse data interpretation and modelling(e.g. as indicated in [3] and [24]) are ongoing.
3. Contribution from unresolved sources
A third major contribution to the detected highGeV to low TeV signal is related to the existenceof VHE γ -ray sources below instrumental detectionthreshold, namely unresolved sources. The H.E.S.S.Galactic survey region is clearly dominated by emissionof individual sources [12], and there is every reasonto assume that this source wealth continues below thecurrent H.E.S.S. detection threshold. The sensitivity ac-complished with the HGPS does not comprise the depthof the whole Milky Way ( cf. Fig. 4 in [12]), accordinglyunresolved sources will contribute to the large-scaleemission signal that is discussed here. This is not anunexpected situation. Contributions from unresolvedsources are expected to increasingly contribute to thedetected emission signal towards higher energies. Also,the γ -ray flux from unresolved sources will not suffer aheavy suppression from the background removal, sincethe population of unresolved γ -ray sources is likely tofollow the distribution of the resolved sources, narrowlylocalized along the Galactic Plane.Refined analysis of the number of detected sourcesvs. cumulative flux distribution (log N( > S) - log S)of the Galactic VHE source population [22] on thebasis of the upcoming H.E.S.S. legacy source catalogproject [12] will allow for a quantitative assessment of thecontribution of unresolved sources to the observed signal.
IV. CONCLUSION
This paper presents the first detection of large-scale γ -ray emission along the Galactic Plane using imaging at- mospheric Cherenkov telescopes. A significant flux alongthe Galactic Plane is detected, which is not attributed toresolved and significantly detected γ -ray sources. The de-tection can be interpreted as diffuse Galactic γ -ray emis-sion and contributions from unresolved sources. Owingto limitations of the applied background removal tech-nique, modest variations in the emission on the scale ofthe H.E.S.S. field of view are suppressed in such measure-ments. As a consequence, the reported signal is consid-ered to represent a lower limit compared to what mightbe detected with improved analysis strategies at theseenergies.The observed signal is comprised of contributions fromcosmic-ray interactions via neutral pion decay-induced γ -ray emission and inverse Compton scattering as wellas from unresolved γ -ray sources. The flux of the γ -rayemission related to π decay is estimated via line-of-sightcolumn densities in HI and CO with the correspondingnarrow latitudinal profile. Such low-scale-height com-ponents are not severely impacted by the applied back-ground subtraction. The same can be expected for thecontribution from unresolved sources. In contrast, in-verse Compton emission is expected to have a distinctlylarger scale height and is, as such, only partly recover-able. While a guaranteed contribution of γ -ray emis-sion from cosmic-ray interactions with the interstellarmedium already makes up a sizable fraction of the signal,the nature of the remaining excess flux and its divisionamong the different emission components remains to beconclusively identified. ACKNOWLEDGMENTS
The support of the Namibian authorities and of theUniversity of Namibia in facilitating the construction andoperation of H.E.S.S. is gratefully acknowledged, as isthe support by the German Ministry for Education andResearch (BMBF), the Max Planck Society, the GermanResearch Foundation (DFG), the French Ministry for Re-search, the CNRS-IN2P3 and the Astroparticle Interdis-ciplinary Programme of the CNRS, the U.K. Science andTechnology Facilities Council (STFC), the IPNP of theCharles University, the Czech Science Foundation, thePolish Ministry of Science and Higher Education, theSouth African Department of Science and Technologyand National Research Foundation, and by the Univer-sity of Namibia. We appreciate the excellent work ofthe technical support staff in Berlin, Durham, Hamburg,Heidelberg, Palaiseau, Paris, Saclay, and in Namibia inthe construction and operation of the equipment. http://galprop.sourceforge.net/, http://galprop.stanford.edu [1] A. A. Abdo et al. , Astrophys. J. , 1078 (2008).[2] M. Ackermann et al. , Astrophys. J. , 22 (2012).[3] M. Ackermann et al. , Astrophys. J. , 3 (2012).[4] M. Ackermann et al. , Astrophys. J. , 81 (2011).[5] F. Aharonian et al. , Astron. Astrophys. , 899 (2006).[6] F. Aharonian et al. , Nature , 695 (2006).[7] F. A. Aharonian, Very High Energy Cosmic Gamma Ra-diation, A Crucial Window on the Extreme Universe (World Scientific Publishing Co. Pte. Ltd., Singapore,2004).[8] D. Berge, S. Funk, and J. Hinton, Astron. Astrophys. , 1219 (2007).[9] J. Beringer et al. (Particle Data Group), Phys. Rev. D , 010001 (2012).[10] G. F. Bignami, C. E. Fichtel, D. A. Kniffen, and D. J.Thompson, Astrophys. J. , 54 (1975).[11] A. D. Bolatto, M. Wolfire, and A. K. Leroy, Annual Re-view of Astronomy and Astrophysics , 207 (2013).[12] S. Carrigan et al. for the H.E.S.S. Collaboration, in Pro-ceedings of the 33rd International Cosmic Ray Confer-ence, Rio de Janeiro (2013).[13] M. de Naurois and L. Rolland, Astroparticle Physics ,231 (2009).[14] S. D. Hunter et al. , Astrophys. J. , 205 (1997). [15] M. Kachelriess, I. V. Moskalenko, and S. S. Ostapchenko,Astrophys. J. , 136 (2014).[16] P. M. W. Kalberla, W. B. Burton, D. Hartmann, E. M.Arnal, E. Bajaja, R. Morras, and W. G. L. P¨oppel, As-tron. Astrophys. , 775 (2005).[17] S. R. Kelner, F. A. Aharonian, and V. V. Bugayov, Phys.Rev. D , 034018 (2006).[18] L. Ma for the ARGO-YBJ Collaboration, in Proceedingsof the 32nd International Cosmic Ray Conference, Bei-jing, , 2 (2011).[19] H. A. Mayer-Hasselwander, G. Kanbach, K. Bennett, G.G. Lichti, G. F. Bignami, P. A. Caraveo, R. Buccheri, F.Lebrun, J. L. Masnou, W. Hermsen, Astron. Astrophys. , 164 (1982).[20] M. Mori, Astroparticle Physics , 341 (2009).[21] S. Ohm, C. van Eldik, and K. Egberts, AstroparticlePhysics , 383 (2009).[22] M. Renaud, in Proceedings of 44th Recontres de Moriond(2009).[23] A. W. Strong, Astrophysics and Space Science , 35(2007).[24] H. J. V¨olk and E. G. Berezhko, Astrophys. J.777