Search for astrophysical sources of neutrinos using cascade events in IceCube
IceCube Collaboration, M. G. Aartsen, M. Ackermann, J. Adams, J. A. Aguilar, M. Ahlers, M. Ahrens, I. Al Samarai, D. Altmann, K. Andeen, T. Anderson, I. Ansseau, G. Anton, C. Argüelles, J. Auffenberg, S. Axani, H. Bagherpour, X. Bai, J. P. Barron, S. W. Barwick, V. Baum, R. Bay, J. J. Beatty, J. Becker Tjus, K.-H. Becker, S. BenZvi, D. Berley, E. Bernardini, D. Z. Besson, G. Binder, D. Bindig, E. Blaufuss, S. Blot, C. Bohm, M. Börner, F. Bos, D. Bose, S. Böser, O. Botner, J. Bourbeau, F. Bradascio, J. Braun, L. Brayeur, M. Brenzke, H.-P. Bretz, S. Bron, A. Burgman, T. Carver, J. Casey, M. Casier, E. Cheung, D. Chirkin, A. Christov, K. Clark, L. Classen, S. Coenders, G. H. Collin, J. M. Conrad, D. F. Cowen, R. Cross, M. Day, J. P. A. M. de André, C. De Clercq, J. J. DeLaunay, H. Dembinski, S. De Ridder, P. Desiati, K. D. de Vries, G. de Wasseige, M. de With, T. DeYoung, J. C. Díaz-Vélez, V. di Lorenzo, H. Dujmovic, J. P. Dumm, M. Dunkman, B. Eberhardt, T. Ehrhardt, B. Eichmann, P. Eller, P. A. Evenson, S. Fahey, A. R. Fazely, J. Felde, K. Filimonov, C. Finley, S. Flis, A. Franckowiak, E. Friedman, T. Fuchs, T. K. Gaisser, J. Gallagher, L. Gerhardt, K. Ghorbani, W. Giang, T. Glauch, T. Glüsenkamp, A. Goldschmidt, J. G. Gonzalez, D. Grant, et al. (214 additional authors not shown)
SSubmitted to The Astrophysical Journal
Preprint typeset using L A TEX style AASTeX6 v. 1.0
SEARCH FOR ASTROPHYSICAL SOURCES OF NEUTRINOS USING CASCADE EVENTS IN ICECUBE
IceCube Collaboration: M. G. Aartsen , M. Ackermann , J. Adams , J. A. Aguilar , M. Ahlers , M. Ahrens ,I. Al Samarai , D. Altmann , K. Andeen , T. Anderson , I. Ansseau , G. Anton , C. Arg¨uelles ,J. Auffenberg , S. Axani , H. Bagherpour , X. Bai , S. W. Barwick , V. Baum , R. Bay , J. J. Beatty ,J. Becker Tjus , K.-H. Becker , S. BenZvi , D. Berley , E. Bernardini , D. Z. Besson , G. Binder ,D. Bindig , E. Blaufuss , S. Blot , C. Bohm , M. B¨orner , F. Bos , D. Bose , S. B¨oser , O. Botner ,J. Bourbeau , F. Bradascio , J. Braun , L. Brayeur , M. Brenzke , H.-P. Bretz , S. Bron , A. Burgman ,T. Carver , J. Casey , M. Casier , E. Cheung , D. Chirkin , A. Christov , K. Clark , L. Classen ,S. Coenders , G. H. Collin , J. M. Conrad , D. F. Cowen , R. Cross , M. Day , J. P. A. M. de Andr´e ,C. De Clercq , J. J. DeLaunay , H. Dembinski , S. De Ridder , P. Desiati , K. D. de Vries ,G. de Wasseige , M. de With , T. DeYoung , J. C. D´ıaz-V´elez , V. di Lorenzo , H. Dujmovic , J. P. Dumm ,M. Dunkman , B. Eberhardt , T. Ehrhardt , B. Eichmann , P. Eller , P. A. Evenson , S. Fahey ,A. R. Fazely , J. Felde , K. Filimonov , C. Finley , S. Flis , A. Franckowiak , E. Friedman , T. Fuchs ,T. K. Gaisser , J. Gallagher , L. Gerhardt , K. Ghorbani , W. Giang , T. Glauch , T. Gl¨usenkamp ,A. Goldschmidt , J. G. Gonzalez , D. Grant , Z. Griffith , C. Haack , A. Hallgren , F. Halzen ,K. Hanson , D. Hebecker , D. Heereman , K. Helbing , R. Hellauer , S. Hickford , J. Hignight ,G. C. Hill , K. D. Hoffman , R. Hoffmann , B. Hokanson-Fasig , K. Hoshina , F. Huang , M. Huber ,K. Hultqvist , S. In , A. Ishihara , E. Jacobi , G. S. Japaridze , M. Jeong , K. Jero , B. J. P. Jones ,P. Kalacynski , W. Kang , A. Kappes , T. Karg , A. Karle , U. Katz , M. Kauer , A. Keivani ,J. L. Kelley , A. Kheirandish , J. Kim , M. Kim , T. Kintscher , J. Kiryluk , T. Kittler , S. R. Klein ,G. Kohnen , R. Koirala , H. Kolanoski , L. K¨opke , C. Kopper , S. Kopper , J. P. Koschinsky ,D. J. Koskinen , M. Kowalski , K. Krings , M. Kroll , G. Kr¨uckl , J. Kunnen , S. Kunwar ,N. Kurahashi , T. Kuwabara , A. Kyriacou , M. Labare , J. L. Lanfranchi , M. J. Larson , F. Lauber ,D. Lennarz , M. Lesiak-Bzdak , M. Leuermann , Q. R. Liu , L. Lu , J. L¨unemann , W. Luszczak ,J. Madsen , G. Maggi , K. B. M. Mahn , S. Mancina , R. Maruyama , K. Mase , R. Maunu , F. McNally ,K. Meagher , M. Medici , M. Meier , T. Menne , G. Merino , T. Meures , S. Miarecki , J. Micallef ,G. Moment´e , T. Montaruli , M. Moulai , R. Nahnhauer , P. Nakarmi , U. Naumann , G. Neer ,H. Niederhausen , S. C. Nowicki , D. R. Nygren , A. Obertacke Pollmann , A. Olivas , A. O’Murchadha ,T. Palczewski , H. Pandya , D. V. Pankova , P. Peiffer , J. A. Pepper , C. P´erez de los Heros ,D. Pieloth , E. Pinat , M. Plum , P. B. Price , G. T. Przybylski , C. Raab , L. R¨adel , M. Rameez ,K. Rawlins , R. Reimann , B. Relethford , M. Relich , E. Resconi , W. Rhode , M. Richman , B. Riedel ,S. Robertson , M. Rongen , C. Rott , T. Ruhe , D. Ryckbosch , D. Rysewyk , T. S¨alzer ,S. E. Sanchez Herrera , A. Sandrock , J. Sandroos , S. Sarkar , S. Sarkar , K. Satalecka ,P. Schlunder , T. Schmidt , A. Schneider , S. Schoenen , S. Sch¨oneberg , L. Schumacher , D. Seckel ,S. Seunarine , D. Soldin , M. Song , G. M. Spiczak , C. Spiering , J. Stachurska , T. Stanev , A. Stasik ,J. Stettner , A. Steuer , T. Stezelberger , R. G. Stokstad , A. St¨oßl , N. L. Strotjohann ,G. W. Sullivan , M. Sutherland , I. Taboada , J. Tatar , F. Tenholt , S. Ter-Antonyan , A. Terliuk ,G. Teˇsi´c , S. Tilav , P. A. Toale , M. N. Tobin , S. Toscano , D. Tosi , M. Tselengidou , C. F. Tung ,A. Turcati , C. F. Turley , B. Ty , E. Unger , M. Usner , J. Vandenbroucke , W. Van Driessche ,N. van Eijndhoven , S. Vanheule , J. van Santen , M. Vehring , E. Vogel , M. Vraeghe , C. Walck ,A. Wallace , M. Wallraff , N. Wandkowsky , A. Waza , C. Weaver , M. J. Weiss , C. Wendt ,S. Westerhoff , B. J. Whelan , S. Wickmann , K. Wiebe , C. H. Wiebusch , L. Wille , D. R. Williams ,L. Wills , M. Wolf , J. Wood , T. R. Wood , E. Woolsey , K. Woschnagg , D. L. Xu , X. W. Xu , Y. Xu ,J. P. Yanez , G. Yodh , S. Yoshida , T. Yuan , and M. Zoll (Received July 28, 2017) Department of Physics, University of Adelaide, Adelaide, 5005, Australia DESY, D-15735 Zeuthen, Germany Dept. of Physics and Astronomy, University of Canterbury, Private Bag 4800, Christchurch, New Zealand Universit´e Libre de Bruxelles, Science Faculty CP230, B-1050 Brussels, Belgium Niels Bohr Institute, University of Copenhagen, DK-2100 Copenhagen, Denmark Oskar Klein Centre and Dept. of Physics, Stockholm University, SE-10691 Stockholm, Sweden D´epartement de physique nucl´eaire et corpusculaire, Universit´e de Gen`eve, CH-1211 Gen`eve, Switzerland Erlangen Centre for Astroparticle Physics, Friedrich-Alexander-Universit¨at Erlangen-N¨urnberg, D-91058 Erlangen, Germany Department of Physics, Marquette University, Milwaukee, WI, 53201, USA Dept. of Physics, Pennsylvania State University, University Park, PA 16802, USA Dept. of Physics, Massachusetts Institute of Technology, Cambridge, MA 02139, USA III. Physikalisches Institut, RWTH Aachen University, D-52056 Aachen, Germany Physics Department, South Dakota School of Mines and Technology, Rapid City, SD 57701, USA a r X i v : . [ a s t r o - ph . H E ] A ug M. G. Aartsen et al. Dept. of Physics and Astronomy, University of California, Irvine, CA 92697, USA Institute of Physics, University of Mainz, Staudinger Weg 7, D-55099 Mainz, Germany Dept. of Physics, University of California, Berkeley, CA 94720, USA Dept. of Physics and Center for Cosmology and Astro-Particle Physics, Ohio State University, Columbus, OH 43210, USA Dept. of Astronomy, Ohio State University, Columbus, OH 43210, USA Fakult¨at f¨ur Physik & Astronomie, Ruhr-Universit¨at Bochum, D-44780 Bochum, Germany Dept. of Physics, University of Wuppertal, D-42119 Wuppertal, Germany Dept. of Physics and Astronomy, University of Rochester, Rochester, NY 14627, USA Dept. of Physics, University of Maryland, College Park, MD 20742, USA Dept. of Physics and Astronomy, University of Kansas, Lawrence, KS 66045, USA Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA Dept. of Physics, TU Dortmund University, D-44221 Dortmund, Germany Dept. of Physics, Sungkyunkwan University, Suwon 440-746, Korea Dept. of Physics and Astronomy, Uppsala University, Box 516, S-75120 Uppsala, Sweden Dept. of Physics and Wisconsin IceCube Particle Astrophysics Center, University of Wisconsin, Madison, WI 53706, USA Vrije Universiteit Brussel (VUB), Dienst ELEM, B-1050 Brussels, Belgium SNOLAB, 1039 Regional Road 24, Creighton Mine 9, Lively, ON, Canada P3Y 1N2 Institut f¨ur Kernphysik, Westf¨alische Wilhelms-Universit¨at M¨unster, D-48149 M¨unster, Germany Physik-department, Technische Universit¨at M¨unchen, D-85748 Garching, Germany Dept. of Astronomy and Astrophysics, Pennsylvania State University, University Park, PA 16802, USA Dept. of Physics and Astronomy, Michigan State University, East Lansing, MI 48824, USA Bartol Research Institute and Dept. of Physics and Astronomy, University of Delaware, Newark, DE 19716, USA Dept. of Physics and Astronomy, University of Gent, B-9000 Gent, Belgium Institut f¨ur Physik, Humboldt-Universit¨at zu Berlin, D-12489 Berlin, Germany Dept. of Physics, Southern University, Baton Rouge, LA 70813, USA Dept. of Astronomy, University of Wisconsin, Madison, WI 53706, USA Dept. of Physics, University of Alberta, Edmonton, Alberta, Canada T6G 2E1 Dept. of Physics and Institute for Global Prominent Research, Chiba University, Chiba 263-8522, Japan CTSPS, Clark-Atlanta University, Atlanta, GA 30314, USA Dept. of Physics, University of Texas at Arlington, 502 Yates St., Science Hall Rm 108, Box 19059, Arlington, TX 76019, USA Dept. of Physics and Astronomy, Stony Brook University, Stony Brook, NY 11794-3800, USA Universit´e de Mons, 7000 Mons, Belgium Dept. of Physics and Astronomy, University of Alabama, Tuscaloosa, AL 35487, USA Dept. of Physics, Drexel University, 3141 Chestnut Street, Philadelphia, PA 19104, USA Dept. of Physics, University of Wisconsin, River Falls, WI 54022, USA Dept. of Physics, Yale University, New Haven, CT 06520, USA Dept. of Physics and Astronomy, University of Alaska Anchorage, 3211 Providence Dr., Anchorage, AK 99508, USA Dept. of Physics, University of Oxford, 1 Keble Road, Oxford OX1 3NP, UK School of Physics and Center for Relativistic Astrophysics, Georgia Institute of Technology, Atlanta, GA 30332, USA Earthquake Research Institute, University of Tokyo, Bunkyo, Tokyo 113-0032, Japan
ABSTRACTThe IceCube neutrino observatory has established the existence of a flux of high-energy astrophysicalneutrinos inconsistent with the expectation from atmospheric backgrounds at a significance greaterthan 5 σ . This flux has been observed in analyses of both track events from muon neutrino interactionsand cascade events from interactions of all neutrino flavors. Searches for astrophysical neutrino sourceshave focused on track events due to the significantly better angular resolution of track reconstructions.To date, no such sources have been confirmed. Here we present the first search for astrophysicalneutrino sources using cascades interacting in IceCube with deposited energies as small as 1 TeV. Nosignificant clustering was observed in a selection of 263 cascades collected from May 2010 to May2012. We show that compared to the classic approach using tracks, this statistically-independentsearch offers improved sensitivity to sources in the southern sky, especially if the emission is spatiallyextended or follows a soft energy spectrum. This enhancement is due to the low background fromatmospheric neutrinos forming cascade events and the additional veto of atmospheric neutrinos atdeclinations (cid:46) − ◦ . Keywords: astroparticle physics — neutrinos INTRODUCTIONNeutrinos are promising messenger particles for astro-physical observations due to their extremely small inter-action cross-sections and lack of electric charge. Theycan travel enormous distances largely unimpeded by in-tervening matter and undeflected by magnetic fields.These properties make it possible to associate neutrinosfrom distant sources with each other and with knownsources of electromagnetic radiation. Furthermore, be-cause neutrinos are produced in high-energy hadronicinteractions, observations of astrophysical neutrinos willshed light on the still-elusive origins of the highest-energy cosmic rays (Gaisser et al. 1995; Learned &Mannheim 2000; Becker 2008).IceCube is the first km -scale neutrino detector(Achterberg et al. 2006). Using an array of photo-multiplier tubes (PMTs) deployed deep in the antarc-tic glacial ice near the South Pole, it can detect neu-trinos of all flavors by collecting the Cherenkov lightemitted by the relativistic charged particles producedwhen neutrinos interact with atomic nuclei in the ice.Neutrinos produce one of two topologically distinct sig-natures: tracks and cascades . Charged current (CC)muon neutrino interactions yield long-lived muons thatcan travel several kilometers through the ice (Chirkin& Rhode 2004), producing an elongated track signaturein the detector. Charged current interactions of otherneutrino flavors, and all neutral current (NC) interac-tions, yield hadronic and electromagnetic showers thattypically range less than 20 m (Aartsen et al. 2014a),with 90% of the light emitted within 4 m of the showermaximum (Radel & Wiebusch 2013) — a short distancecompared to the scattering and absorption lengths oflight in the ice (Aartsen et al. 2013b) as well as thespacing of the PMTs. These showers produce a nearlyspherically symmetric cascade signature in light.A flux of astrophysical neutrinos above ∼
60 TeV in-consistent with the expectation from atmospheric back-grounds at greater than 5 σ was first established usingneutrinos interacting within the instrumented volume ofIceCube (Aartsen et al. 2014c, 2015a). The majority ofevents contributing to this measurement were cascades.More recently, this flux has been confirmed in an anal-ysis of tracks from muon neutrinos above ∼
300 TeVoriginating in the northern sky (Aartsen et al. 2015c,2016c). No significant anisotropy has yet been observed,and the neutrino flavor ratio at Earth is consistent with1:1:1 (Aartsen et al. 2015d). Searches for astrophysical neutrino sources have tra-ditionally focused on track events because the elongatedsignature gives much better angular resolution than canbe obtained for cascades. While ANTARES recently re-ported the addition of a cascade selection to their all-skysearch for sources of steady neutrino emission (Adrian-Martinez et al. 2015), IceCube has so far excluded cas-cades from its all-sky search (Aartsen et al. 2017a) ex-cept in the simplified analysis applied only to very high-energy contained events (Aartsen et al. 2015a).In this paper, we present the first all-sky search forastrophysical neutrino sources producing cascades inIceCube with deposited energies as small as 1 TeV. Thisanalysis includes 263 cascades observed from May 2010to May 2012. We find that, due to the relatively low rateof atmospheric backgrounds in this sample, this searchreduces the energy threshold in the southern sky relativeto previous IceCube work with tracks. The sensitivityof this search is much less dependent on the declina-tion, spatial extension, and emission spectrum of a pos-sible source. In the following sections, we begin with anoverview of the detector, experimental dataset, and sta-tistical methods used in this analysis before reportingresults from the two-year sample and discussing direc-tions for future work. ICECUBEThe IceCube detector (described in detail in Aartsenet al. (2017b)) consists of 5160 Digital Optical Modules(DOMs) buried in the glacial ice near the South Pole.The DOMs are mounted on 86 vertical “strings”, with60 DOMs on each string. Each string is connected to acentral lab on the surface by a cable that provides powerand communication with the data acquisition (DAQ)system (Abbasi et al. 2009). Seventy-eight of the stringsare arranged in a hexagonal grid with a spacing of 125 m;on these strings, DOMs are distributed uniformly from1450 m to 2450 m below the surface of the ice. The re-maining 8 strings make up the denser DeepCore in-fillarray (Abbasi et al. 2012), with inter-string spacing of 30to 60 m. The in-fill strings include 50 DOMs in the par-ticularly clear ice at depths of 2100 m to 2450 m and anadditional 10 DOMs evenly spaced at depths of 1750 mto 1850 m. Construction was performed during Australsummers starting in 2004. A nearly complete 79-stringconfiguration began taking data in May 2010, and thefirst year of data from the complete 86-string detectorwas taken from May 2011 to May 2012.
M. G. Aartsen et al.
Each DOM includes a 25 cm diameter PMT (Abbasiet al. 2010) and supporting electronics. A local coinci-dence condition occurs when a DOM and one of its near-est neighbors exceed a threshold of 1/4 of the mean ex-pected voltage from a single photoelectron (PE). Whenat least eight DOMs observe local coincidence within6 . µ s, the DAQ produces an event consisting of 400 nsdigitized waveforms from all DOMs observing local co-incidence and 75 ns waveforms from all other DOMs ex-ceeding the threshold. The waveforms are then decom-posed into series of pulse arrival times and PE countswhich are used to reconstruct the trajectory and de-posited energy of the relativistic particles in the detector(Ahrens et al. 2004; Aartsen et al. 2014a).The simple eight-DOM trigger accepts neutrino-induced events with very high efficiency. Unfortunately,even deep in the glacial ice, cosmic ray-induced atmo-spheric muons trigger the detector at an average rate of2.7 kHz, overwhelming the trigger rate of atmosphericneutrinos ( ∼
20 mHz) and rare astrophysical neutrinos.An initial data reduction step performed at the SouthPole reduces the event rate by a factor of 100 by rejectinglower-energy events that are consistent with downgoingtracks. The remaining dataset, still dominated by down-going muons, is transmitted to the northern hemispherevia satellite for further analysis. NEUTRINO SELECTIONIceCube searches for moderate to high energy neutri-nos generally exploit one of two methods to reject the at-mospheric muon background. The largest effective vol-ume and best angular resolution are available when in-coming muon tracks are accepted. This approach offersgood performance for muon neutrinos from the north-ern celestial hemisphere because only neutrinos can sur-vive passage through the intervening earth before pro-ducing upgoing muons in the ice. However, neutrino-and cosmic ray-induced downgoing muon tracks enteringthe detector from above produce nearly indistinguish-able event topologies. Astrophysical neutrinos from thesouthern sky can be identified on a statistical basis if theneutrino spectrum is harder than the atmospheric muonspectrum, but this strategy increases the energy thresh-old to ∼
100 TeV in the southern sky, compared to only ∼ ν µ -induced muon tracks,resulting in a smaller final sample that is dominatedby cascades. The angular uncertainty of cascade recon-structions ( (cid:38) ◦ ) is large compared to that of track re-constructions ( (cid:46) ◦ ). However, the requirement that theneutrino interaction vertex is located within the instru-mented volume results in good energy resolution (within ∼ (cid:38)
60 TeV (Aart-sen et al. 2014c). In a follow-up analysis, the energythreshold was reduced to ∼ ν µ point source searches (Aartsen et al. 2014d) and asmall fraction of the cascades are included in the earlierhigh energy starting event analysis (Aartsen et al. 2013a,2014c), the majority of these cascades have not yet beenstudied in the context of spatial clustering. In this pa-per, we turn our attention to 263 of these cascades withdeposited energies of 1 TeV − E − . spectrum and contributes an expected 71 +9 . − . cascadesin 641 days — a far larger fraction of the total eventrate than in previous source searches with tracks (Aart-sen et al. 2017a). The neutrino energy distribution isshown in Figure 2 for the best-fit spectrum as well as thehard ( E − ) and soft ( E − ) source spectrum hypotheses E [GeV]10 e v e n t s i n d a y s Conventional ν Penetrating µ Astrophysical ν Observed Events − . − . . . . δ )10 e v e n t s i n d a y s Figure 1 . Reconstructed energy (left) and declination (right) distributions for the best-fit atmospheric and astrophysical spectra(shaded histograms) obtained in Aartsen et al. (2015b) compared to the distributions for the 263 cascades (black crosses)depositing at least 1 TeV observed in that analysis. Atmospheric muons misidentified as cascades after passing undetectedthrough the veto layer are concentrated at sin( δ ) < − .
3, while in the same range some atmospheric neutrinos are rejectedbecause they are accompanied by incoming muons. tested directly in this paper. For an E − spectrum, weexpect 90% of events to have energies between 2 TeVand 90 TeV; for an E − spectrum this range shifts to6 TeV − ∼ few PeV. The reconstructed energy agrees withthe neutrino energy within ∼
10% for 68% of CC ν e in-teractions and is on average proportional to neutrino en-ergy for other interaction flavors (Aartsen et al. 2015b).Agreement between reconstructed and true neutrino en-ergy is shown in Figure 3. The primary challenge forsource searches with cascades is the angular reconstruc-tion, for which the performance is shown as a functionof energy in Figure 4 and averaged over all energiesin Figure 5. At low energies, the reconstruction ben-efits to some degree from the preferential selection ofinteractions in or near the more densely instrumentedDeepCore. At high energies, performance is somewhatpoorer than optimal — compare with, e.g., Aartsen et al.(2014a) — likely due to the specific reconstruction set-tings used for this sample, which are less computation-ally intensive but which employ a coarser description ofthe expected light yield and a less rigorous scan of thedirectional likelihood landscape. METHODS AND PERFORMANCEWe use an unbinned maximum likelihood method toquantify the extent to which the observed events aremore consistent with a spatially localized astrophysi-cal signal hypothesis than a randomly distributed back-ground hypothesis. This method exploits the spatialdistribution of events as well as the distribution of per-event deposited energies, where the latter improves thesensitivity to sources with harder spectra than atmo-spheric backgrounds. While we largely follow the ap-proach used in traditional track analyses (most recentlyAartsen et al. 2017a), the specific signal and backgroundmodels are modified to accommodate the large angularuncertainties and overall low statistics of the cascadeevent selection. In Section 4.1 we review the likelihoodconstruction, including explanations for changes withrespect to previous work with tracks. In Section 4.2we introduce the specific hypothesis tests considered inthis work. Systematic uncertainties are discussed in Sec-tion 4.3 and the performance of the cascade analysis ispresented in Section 4.4.
M. G. Aartsen et al. E [GeV] r e l a t i v e a bund a n ce [ a r b i tr a r y un i t s ] E − E − . E − Figure 2 . Neutrino energy distributions expected fromsources emitting a hard ( E − , blue) or soft ( E − , red) spec-trum compared with the best-fit all-sky astrophysical com-ponent following an E − . spectrum (green). Each distribu-tion includes all standard model neutrino flavors, assuming a1:1:1 flavor ratio at Earth with equal fluxes of ν and ¯ ν . Ver-tical lines indicate intervals containing 90% of events. Whileno such events have yet been observed, an enhanced accep-tance is expected for ν e at 6 . Maximum Likelihood Method
The likelihood is expressed as a product over events i : L ( n s , γ ) = (cid:89) i (cid:104) n s N S i + (cid:16) − n s N (cid:17) B i (cid:105) , (1)where n s is the number of signal events, γ is the spectralindex of the source, N = 263 is the total number ofevents, S i is the likelihood of event i contributing to thesource, and B i is the likelihood of event i contributingto atmospheric or unresolved astrophysical backgrounds. S i depends on the properties of both event i and thesource hypothesis (including spectral index γ ), while B i depends only on the properties of the events. ˆ n s andˆ γ are the values that give the maximum likelihood ˆ L ,subject to the constraint that ˆ n s ≥
0. Events that aremore correlated spatially or energetically with the sourcehypothesis obtain larger values for S i , driving the fittowards larger values of ˆ n s and ˆ L .We approximate the signal and background likeli-hoods S i and B i as products of space and energy fac-tors: S space i · S energy i and B space i · B energy i , respectively.Each factor is obtained by convolving the properties ofthe event origin — either astrophysical source, or atmo-spheric or unresolved astrophysical background — first − E reco /E true . . . . . p r o b a b ili t y d e n s i t y Figure 3 . Ratio of reconstructed to true neutrino energyfor signal MC following an E − . spectrum. Reconstructedenergy is on average proportional to true neutrino energy forall interaction flavors, with agreement within ∼
10% for 68%of CC ν e interactions. with the detector response and then with the event re-construction resolution. For B space i , this is done using anormalized histogram of reconstructed declination δ foran ensemble of background-like events, accounting fordetector effects and smearing from finite angular resolu-tion simultaneously. Similarly, for S energy i and B energy i ,we use normalized histograms of the logarithm of the de-posited energy log E for ensembles of signal-like andbackground-like events, respectively, accounting for thedeclination dependence with separately-normalized his-tograms in each of ten bins in sin δ . S energy i is computedfrom signal Monte Carlo (MC) on a grid of spectral in-dices γ ranging from 1 to 4. For a given event, B space i , S energy i and B energy i are equal to the values of the his-tograms for the bin containing the event. The locationof IceCube at the geographic South Pole allows us toexpress these factors as functions of declination, ratherthan zenith angle with respect to the detector, withoutloss of information. A small additional dependence onazimuth angle is neglected.In the classic track analysis, the background per-eventlikelihoods are constructed from the full experimentaldataset. With a large sample of well-reconstructed muontracks dominated by atmospheric backgrounds, both B space i and B energy i are well constrained statistically evenfor dense binning in both sin δ and log E . By contrast,our sample of only 263 cascade events is only sufficientto constrain B space i . Thus our first modification to the E [GeV]0102030405060 a n g u l a r e rr o r [ ◦ ] Figure 4 . Expected angular reconstruction performance asa function of neutrino energy. Shaded regions indicate theradii of error circles covering 20%, 50%, and 80% of events.Below 20 PeV, the median angular error, highlighted by thedark blue curve, ranges from 11 ◦ to 20 ◦ . method is to construct B energy i from neutrino and at-mospheric muon MC simulations weighted to the best-fit atmospheric and astrophysical spectra found by theall-sky flux analysis using these events (Aartsen et al.2015b). In this way we obtain a detailed estimate ofthe energy distribution throughout the sky, even at en-ergies not yet observed at all declinations in two yearsof experimental livetime.The signal space factor S space i is obtained by convolv-ing a source hypothesis with an analytical estimate ofthe spatial probability density distribution for event i originating at reconstructed right ascension and decli-nation ( α i , δ i ). In track analyses, it is a good approxi-mation to model this distribution as a 2D Gaussian withwidth σ i estimated event-by-event using a dedicated re-construction. We modify this treatment for cascadesboth because the angular uncertainties are much largerand because it is too computationally expensive to esti-mate them directly for each event.In this analysis we parameterize the angular resolutionas a function of reconstructed declination δ i and energy E i . In parts of this parameter space, either the declina-tion or right ascension errors tend to be systematicallylarger, so these are treated independently. For each of10 bins in sin δ and 12 in log E , we find the values σ α and σ δ such that | α i − α true i | < σ α , and separately | δ i − δ true i | < σ δ , for 68.27% of simulated events in thebin. The spatial probability density distribution for ob- , true)0 . . . . . . p r o b a b ili t y d e n s i t y Figure 5 . Expected distribution of angular separation be-tween reconstructed and true neutrino direction for signalMC following an E − . spectrum. While the distributionincludes a tail extending all the way to 180 ◦ , 50% (90%) ofevents are reconstructed within 13 ◦ (45 ◦ ). served event i is the product of 1D Gaussians with thesewidths, normalized such that the distribution integratesto unity on the sphere.We consider two types of source hypothesis: pointsources and the galactic plane — an extremely extendedsource. A point source is modeled as a 2D delta distribu-tion centered at the source coordinates. The expectedemission from the galactic plane is in general model-dependent. Here we represent the galactic plane as asimple line source at galactic latitude b = 0. In eithercase, S space i is obtained by convolving the source hy-pothesis with the per-event spatial probability densitydistributions described above. For point sources, theconvolution is trivial; for the galactic plane, it is evalu-ated numerically on a grid with 1 ◦ spacing.4.2. Hypothesis Tests
In this work we consider three search categories: (1) ascan for point-like sources anywhere in the sky, (2) asearch for neutrinos correlated with an a priori catalogof promising source candidates, and (3) a search for neu-trinos correlated with the galactic plane. Each searchentails multiple specific hypothesis tests. The all-skyscan tests for point-like sources on a dense grid of coor-dinates throughout the sky. The catalog search tests thecoordinates of each source candidate individually. Thegalactic plane search includes partially correlated testsfor a hypothesis including the entire galactic plane and a
M. G. Aartsen et al. hypothesis including only the part of the galactic planein southern sky.The test statistic used to compute significances is thelikelihood ratio: T = − (cid:20) L ( n s = 0) L (ˆ n s , ˆ γ ) (cid:21) , (2)where L ( n s = 0) is the background-only likelihood andis independent of γ . For an individual hypothesis, thepre-trials significance p pre of an observation yielding atest statistic T obs is the probability of observing T > T obs if the background-only hypothesis were true. Thebackground-only T distribution is found by performingthe likelihood test on a large number of ensembles withrandomized α i , which removes any clustering that maybe present in the true event ensemble. At declinationsclose to the poles, | δ i | > ◦ , randomizing α i alone isinsufficient to remove a possible cluster of cascades. Thisis addressed by additionally randomizing sin δ i for the 15events within these regions.The pre-trials results, p pre , do not account for multi-ple and partially correlated hypothesis tests conductedin each search category. The post-trials significance isdetermined by the most significant p pre for any hypoth-esis in the category. Specifically, for each search cate-gory we find the post-trials probability p post of observ-ing any min( p pre ) < min( p pre ) obs if the background-onlyhypothesis were true. The background-only min( p pre )distribution is found by generating additional random-ized event ensembles and noting the most significant p pre in each one. This construction leads to one fi-nal significance p post for each type of search; a furtherlook-elsewhere effect between the all-sky, source candi-date catalog, and galactic plane searches is not explic-itly accounted for. This method is conservative in thatit strictly controls only the false positive, but not thetrue positive, error rate.We use the classical statistical approach (Neyman1937; Lehmann & Romano 2005) to calculate the sen-sitivity, discovery potential, and flux upper limits. Theflux level is determined using randomized trials in whichsignal MC events are injected at a Poisson rate n sig and distributed according to the spatial and energeticproperties of the signal hypothesis. The remaining N − n sig events are injected according to the backgroundmodeling procedure described above. The sensitivityflux is that which gives a 90% probability of obtaining T > T med , where T med is the median of the background-only T distribution. The discovery potential flux is ob-tained by the same procedure, but for a 50% probabilityof yielding a 5 σ pre-trials significance. The 90% confi-dence level upper limit is the larger of either the sen-sitivity or that flux which gives a 90% probability ofobtaining T > T obs . 4.3. Systematic Uncertainties
The randomization procedures described in the pre-vious section yield background models and significancesthat are robust against systematic uncertainties. How-ever, flux calculations in this analysis are based on de-tailed neutrino signal MC as described in Aartsen et al.(2016c) and are subject to systematic uncertainties. Weestimate the impact of these uncertainties on our resultsvia their impact on the cascade angular resolution andsignal acceptance. Of these, uncertainties related to theangular resolution are the dominant effect. Reconstruc-tion performance estimates from the baseline MC arelimited by statistical uncertainties in the observed lightas well as any practical computational tradeoffs madein data processing. These estimates do not account forpossible systematic errors in the modeling of light ab-sorption and scattering in either the bulk of South Poleglacial ice or the narrow columns of refrozen ice sur-rounding the DOMs. Uncertainties in the light yieldfrom showers and the optical efficiency of the DOMs arealso neglected in the baseline MC. Taken together, weestimate that these effects introduce an angular resolu-tion uncertainty that can be approximated as a Gaus-sian smearing of the baseline point spread function withwidth σ sys ∼ ◦ (compare, e.g., the typical per-event er-rors in Aartsen et al. (2014c) with the median expectedpure-statistical errors in Aartsen et al. (2014a)). Ap-plying this smearing weakens the sensitivity by ∼ ∼ E − ( E − ) spectrum,approximately independent of source declination.The uncertainties described above also have a smallimpact on the estimated signal acceptance of the eventselection. Uncertainties in the DOM efficiency are on av-erage inversely correlated with uncertainties in the scat-tering and absorption coefficients, so we can safely esti-mate the impact of these uncertainties using a parame-terization from available MC datasets which only varythe DOM efficiency explicitly. We consider a reducedDOM efficiency of −
10% relative to the baseline MC,which decreases both the number of accepted events fora given flux and the reconstructed deposited energy ofeach simulated event. Under this change most signalevents are assigned slightly smaller weights (
S/B ) energy i and some fall below the detection threshold, weakeningthe sensitivity by ∼ / − .
4% below 100 PeV (Cooper-Sarkaret al. 2011). The resulting impact on this analysis is ingeneral dependent on declination and neutrino energy,as an increased (decreased) cross section would simulta-neously increase (decrease) the probability of detecting − . − . . . . δ )10 − − − − E · ( E / T e V ) γ − · d N / d E [ T e V c m − s − ] NorthSouth E − E − E − E − E − Figure 6 . Per-flavor sensitivity of the present 2-year cascadeanalysis and previous 7-year IceCube (Aartsen et al. 2017a)and 1338-day ANTARES (Adrian-Martinez et al. 2014) trackanalyses as a function of declination for a hard spectrum( γ = 2) and soft spectrum ( γ = 3). a neutrino upon arrival in the instrumented volume butdecrease (increase) the probability of a neutrino reachingthe detector after passing through the intervening earthand ice. We take ∼
4% as a conservative estimate ofthe acceptance uncertainty due to neutrino interactioncross section uncertainties.While the signal acceptance depends largely on thetotal amount of light recorded by the DOMs, the angu-lar resolution depends most strongly on the spatial andtemporal distribution of light in the detector. Therefore,we take these effects to be approximately independentand add the above values in quadrature to obtain a totalsystematic uncertainty of 21% (24%) for sources follow-ing an E − ( E − ) spectrum. All following sensitivities,discovery potentials, and flux upper limits include thisfactor. 4.4. Performance
The per-flavor sensitivity flux as a function of sourcedeclination for this work and the most recently pub-lished IceCube (Aartsen et al. 2017a) and ANTARES(Adrian-Martinez et al. 2014) track analyses are com-pared in Figure 6. The cascade sensitivity shows onlyweak declination dependence and, for an E − spec-trum, roughly traces the sensitivity of ANTARES. Nearthe South Pole, the sensitivity is enhanced by the vetoof atmospheric neutrinos accompanied by muons fromthe same cosmic ray-induced shower. The sensitivity E [GeV]10 − − − − E · d N / d E [ T e V c m − s − ] δ = − ◦ Figure 7 . Per-flavor differential sensitivity for a source at δ = − ◦ for track analyses of throughgoing (Aartsen et al.2014d) and starting (Aartsen et al. 2016b) tracks, comparedto this cascade analysis using the event selection from Aart-sen et al. (2015b). The sensitivity in cascades is enhancedat 6 . is weaker near the horizon, where this veto of atmo-spheric neutrinos is not possible. From the horizon tothe North Pole, the sensitivity then improves for a soft E − spectrum but continues to weaken for a hard E − spectrum because high-energy neutrinos are subject tosignificant absorption in transit through the Earth. Thesensitivity of the classic track search, by contrast, isstrongly declination-dependent, with best performancein the northern sky. For a southern source with a softspectrum, the sensitivity flux is better with just twoyears of cascades than with seven years of tracks.We further explore the sensitivity to a southern sourceat δ = − ◦ in Figure 7, which shows the per-flavorsensitivity flux for an E − signal spectrum injected inquarter-decade bins in neutrino energy. Here we directlycompare the cascade and track channels by scaling eachanalysis to an equal three year livetime — the sameexposure as in the first IceCube point source search tomake use of starting tracks (Aartsen et al. 2016b). Atthis declination, the low background cascade search ismore sensitive to such a southern source than IceCubetrack-based searches up to ∼ M. G. Aartsen et al. ◦ )01234 E · d N / d E [ T e V c m − s − ] × − δ = − ◦ δ = +30 ◦ δ = +0 ◦ δ = − ◦ Figure 8 . Per-flavor sensitivity as a function of angular ex-tension of the source. For cascades, a point source hypoth-esis is used in the likelihood regardless of injected sourceextension. For tracks, the sensitivity is found for an ex-tended source hypothesis matching the injected signal usingthe throughgoing track dataset from Aartsen et al. (2017a). source. The source extension is modeled as a Gaussiansmearing of a point source hypothesis. For a smear-ing of up to 10 ◦ , the sensitivity of this search is only30% weaker than for a point source. In the classic tracksearches with angular resolution (cid:46) ◦ , the sensitivityflux increases much more rapidly with source extension— even when a matching extended source hypothesis isused in the likelihood. As shown in Figure 8, the per-flavor sensitivity flux for a source with extension ≥ ◦ in the southern sky at δ ≤ − ◦ is lower with just twoyears of cascades than with seven years of tracks. Thecascade analysis performance is sufficiently independentof source extension that we need not apply dedicatedextended source hypothesis tests in this work. RESULTSThe result of the all-sky scan is shown in Figure 9. Themost significant deviation from the isotropic expectationis found in the southern sky at ( α, δ ) = (277 . ◦ , − . ◦ ).The pre-trials significance is p pre = 0 . n s =7 . γ = 2 .
2, respectively. Accounting for the largenumber of partially correlated hypothesis tests in thisscan, as described in 4.2, the post-trials significance is p post = 66%.For the source candidate catalog search, an ensembleof 74 promising source candidates was selected a pri- Equatorial − − − log p Figure 9 . Two-year starting cascade skymap in equatorialcoordinates (J2000). The skymap shows pre-trial p -values forall locations in the sky. The grey curve indicates the galacticplane, and the grey dot indicates the galactic center. ori by merging previously studied catalogs of interest-ing galactic and extra-galactic objects (Aartsen et al.2017a; Adrian-Martinez et al. 2016b). The result of thesearch is shown in Table 1. The most significant sourceis BL Lac, located at ( α, δ ) = (330 . ◦ , . ◦ ). Thepre-trials significance is p pre = 1 . n s = 6 . γ = 3 .
0, respectively. The post-trials significance is p post = 36%. Flux upper limits for each object in thecatalog are shown in Figure 10 along with the sensitivityand 5 σ discovery potential as functions of declination.Of the galactic plane searches, the southern-sky-onlyhypothesis test was more significant, with a pre-trials p pre = 50%. The fit obtained n s = 2 . γ = 2. Thistest is strongly correlated with the all-sky search; thepost-trials significance is p post = 65%. CONCLUSION AND OUTLOOKIn this first search for sources of astrophysical neutri-nos using cascades with energies as low as 1 TeV in twoyears of IceCube data, no significant source was found.This result is consistent with previous ν µ searches (Aart-sen et al. 2017a; Adrian-Martinez et al. 2012, 2016b)which already find stringent constraints on emissionfrom astrophysical point sources of neutrinos. Never-theless, this analysis shows that despite large angularuncertainties, all-flavor source searches with cascadesare surprisingly sensitive, particularly to emission fromsouthern sources that follow a soft energy spectrum orare spatially extended. This type of analysis is thereforecomplementary to standard ν µ searches, which are mostsensitive to point-like and northern sources.Future source searches with cascades will benefit fromseveral improvements. Most importantly, the adaptiveveto method will soon be applied to at least four moreyears of IceCube data. Because of the low backgroundin this event selection, the sensitivity strengthens faster1 − . − . . . . δ )10 − − E · d N / d E [ T e V c m − s − ] dNdE ∝ E − σ )2 Year SensitivityUpper Limits (90% CL) − . − . . . . δ )10 − − E · ( E / T e V ) · d N / d E [ T e V c m − s − ] dNdE ∝ E − σ )2 Year SensitivityUpper Limits (90% CL) Figure 10 . Sensitivity and 5 σ discovery potential as functions of declination, with flux upper limits for each object in the sourcecatalog. Left: hard spectral assumption ( γ = 2). Right: soft spectral assumption ( γ = 3). than [detector livetime] − / , as shown in Figure 11. On-going work on the optimization of cascade angular re-constructions, including increasingly detailed studies ofCherenkov light propagation in South Pole glacial ice,may lead to angular resolution improvements that in-crease the cascade channel signal-to-background ratiofurther still.In this work, we searched for neutrino emission froma catalog of source candidates previously studied intrack analyses (Aartsen et al. 2017a; Adrian-Martinezet al. 2016b). The catalog was optimized in light ofthe strengths of those analyses, and thus includes manynorthern sources which would almost certainly be visiblefirst in throughgoing tracks. We may be able to improvethe discovery potential for future catalog analyses withcascades by considering a catalog of source candidatesfor which this analysis is best-suited, such as extendedobjects in the southern sky.We have considered only very simple models for ex-tended emission from the galactic plane, which we havetreated here as a uniform line source. However, detailedmodels (Ackermann et al. 2012; Gaggero et al. 2015)have been constructed to account for the measured dis-tribution of γ emission from poorly resolved sources andcosmic ray interactions with galactic dust clouds. Futurecascade analyses will test these models directly, leadingto clearer statements on neutrino emission within ourown galaxy.Here we have searched only for steady, time-independent neutrino emission, but the conclusions of this paper apply equally well to transient sources. Whilea cascade event selection has been added to IceCube’sgamma-ray burst analysis (Aartsen et al. 2016a), othertime-dependent analyses (e.g. Aartsen et al. 2015e)have not yet made use of this channel. In the future,searches for emission from objects such as flaring AGNcould benefit from the inclusion of neutrino-inducedcascades. Proposed next-generation detectors (Aartsenet al. 2014b; Adrian-Martinez et al. 2016a) may alsobenefit by considering source searches with the cascadechannel in the optimization of their optical sensors andarray geometry.We acknowledge the support from the followingagencies: U.S. National Science Foundation-Office ofPolar Programs, U.S. National Science Foundation-Physics Division, University of Wisconsin Alumni Re-search Foundation, the Grid Laboratory Of Wiscon-sin (GLOW) grid infrastructure at the University ofWisconsin - Madison, the Open Science Grid (OSG)grid infrastructure; U.S. Department of Energy, andNational Energy Research Scientific Computing Cen-ter, the Louisiana Optical Network Initiative (LONI)grid computing resources; Natural Sciences and En-gineering Research Council of Canada, WestGrid andCompute/Calcul Canada; Swedish Research Council,Swedish Polar Research Secretariat, Swedish NationalInfrastructure for Computing (SNIC), and Knut andAlice Wallenberg Foundation, Sweden; German Min-istry for Education and Research (BMBF), Deutsche2 M. G. Aartsen et al.
Forschungsgemeinschaft (DFG), Helmholtz Alliance forAstroparticle Physics (HAP), Initiative and Network-ing Fund of the Helmholtz Association, Germany; Fundfor Scientific Research (FNRS-FWO), FWO Odysseusprogramme, Flanders Institute to encourage scientificand technological research in industry (IWT), BelgianFederal Science Policy Office (Belspo); Marsden Fund,New Zealand; Australian Research Council; Japan Soci-ety for Promotion of Science (JSPS); the Swiss NationalScience Foundation (SNSF), Switzerland; National Re-search Foundation of Korea (NRF); Villum Fonden,Danish National Research Foundation (DNRF), Den-mark . . . . . . E · d N / d E [ T e V c m − s − ] × − δ = − ◦ Projected E − SensitivityBackground-limited (1 / √ T )Background-free (1 /T ) Figure 11 . Projected sensitivity as a function of detectorlivetime for a source at δ = − ◦ . Time evolution scal-ing with 1 /T (background-free case) and 1 / √ T (background-limited case) are shown in thick and thin grey dashed curves,respectively. REFERENCES
Aartsen, M. G., et al. 2013a, Science, 342, 1242856—. 2013b, Nucl. Instrum. Meth., A711, 73—. 2014a, JINST, 9, P03009—. 2014b, arXiv:1412.5106—. 2014c, Phys. Rev. Lett., 113, 101101—. 2014d, Astrophys. J., 796, 109—. 2015a, ICRC, 34, 1081—. 2015b, Phys. Rev., D91, 022001—. 2015c, Phys. Rev. Lett., 115, 081102—. 2015d, Phys. Rev. Lett., 114, 171102—. 2015e, Astrophys. J., 807, 46—. 2016a, Astrophys. J., 824, 115—. 2016b, Astrophys. J., 824, L28—. 2016c, Astrophys. J., 833, 3—. 2017a, Astrophys. J., 835, 151—. 2017b, JINST, 12, P03012Abbasi, R., et al. 2009, Nucl. Instrum. Meth., A601, 294—. 2010, Nucl. Instrum. Meth., A618, 139—. 2012, Astropart. Phys., 35, 615Achterberg, A., et al. 2006, Astropart. Phys., 26, 155Ackermann, M., et al. 2012, Astrophys. J., 750, 3 Adrian-Martinez, S., et al. 2012, Astrophys. J., 760, 53—. 2014, Astrophys. J., 786, L5—. 2015, ICRC, 34, 1078—. 2016a, J. Phys., G43, 084001—. 2016b, Astrophys. J., 823, 65Ahrens, J., et al. 2004, Nucl. Instrum. Meth., A524, 169Becker, J. K. 2008, Phys. Rept., 458, 173Chirkin, D., & Rhode, W. 2004, arXiv:hep-ph/0407075Cooper-Sarkar, A., Mertsch, P., & Sarkar, S. 2011, JHEP, 08, 042Gaggero, D., et al. 2015, Astrophys. J., 815, L25Gaisser, T. K., Halzen, F., & Stanev, T. 1995, Phys.Rept., 258,173Glashow, S. L. 1960, Phys. Rev., 118, 316Learned, J., & Mannheim, K. 2000, Ann. Rev. Nucl. Part. Sci.,50, 679Lehmann, E. L., & Romano, J. P. 2005, Testing statisticalhypotheses, 3rd edn., Springer Texts in Statistics (New York:Springer), xiv+784Neyman, J. 1937, Philos. Tr. R. Soc. A, 236, 333Radel, L., & Wiebusch, C. 2013, Astropart. Phys., 44, 102 Table 1 . Summary of the source catalog search. The objects are grouped by type, and withineach type are sorted by increasing declination. The type, common name, and equatorialcoordinates (J2000) are shown for each object. Where non-null (ˆ n s >
0) results are found, thepre-trials significance p pre and best-fit ˆ n s and ˆ γ are given.Type Source α ( ◦ ) δ ( ◦ ) p pre ˆ n s ˆ γ BL Lac PKS 2005-489 302.37 − .
82 0.252 2.4 2.2PKS 0537-441 84.71 − .
09 0.256 1.7 1.8PKS 0426-380 67.17 − .
93 0.597 1.0 1.8PKS 0548-322 87.67 − .
27 0.634 1.2 2.2H 2356-309 359.78 − .
63 0.809 0.2 2.4PKS 2155-304 329.72 − .
23 0.642 1.2 2.41ES 1101-232 165.91 − .
49 0.390 3.3 2.81ES 0347-121 57.35 − .
99 0.543 2.5 3.8PKS 0235+164 39.66 16 . · · · · · · . · · · · · · W Comae 185.38 28 .
23 0.618 0.6 3.8Mrk 421 166.11 38 . · · · · · · Mrk 501 253.47 39 .
76 0.404 1.5 2.6 † BL Lac 330.68 42 .
28 0.010 6.9 3.0H 1426+428 217.14 42 .
67 0.566 0.5 3.83C66A 35.67 43 .
04 0.482 0.9 3.81ES 2344+514 356.77 51 .
70 0.189 2.9 3.21ES 1959+650 300.00 65 .
15 0.519 0.6 3.0S5 0716+71 110.47 71 . · · · · · · Flat spectrum radio quasar PKS 1454-354 224.36 − .
65 0.612 1.6 2.2PKS 1622-297 246.53 − .
86 0.286 3.6 2.2PKS 0454-234 74.27 − . · · · · · · QSO 1730-130 263.26 − .
08 0.365 4.5 3.8PKS 0727-11 112.58 − . · · · · · · PKS 1406-076 212.24 − .
87 0.375 5.6 3.8QSO 2022-077 306.42 − . · · · · · · HESS J1837-069 279.41 − .
95 0.121 8.9 3.83C279 194.05 − .
79 0.754 0.9 3.83C 273 187.28 2 .
05 0.718 0.9 2.8PKS 1502+106 226.10 10 .
49 0.057 9.1 3.8PKS 0528+134 82.73 13 . · · · · · ·
3C 454.3 343.49 16 .
15 0.066 7.4 3.84C 38.41 248.81 38 .
13 0.391 1.6 2.4Galactic center Sgr A* 266.42 − .
01 0.080 5.6 2.2Not identified HESS J1507-622 226.72 − .
34 0.473 0.7 1.0HESS J1503-582 226.46 − .
74 0.438 0.7 1.0HESS J1741-302 265.25 − .
20 0.072 5.7 2.2
Table 1 continued M. G. Aartsen et al.
Table 1 (continued)
Type Source α ( ◦ ) δ ( ◦ ) p pre ˆ n s ˆ γ HESS J1834-087 278.69 − .
76 0.180 7.5 3.8MGRO J1908+06 286.98 6 .
27 0.078 8.5 3.8Pulsar wind nebula HESS J1356-645 209.00 − .
50 0.795 0.1 3.8PSR B1259-63 197.55 − . · · · · · · HESS J1303-631 195.74 − . · · · · · · MSH 15-52 228.53 − .
16 0.408 0.7 1.0HESS J1023-575 155.83 − . · · · · · · HESS J1616-508 243.78 − .
40 0.166 2.4 2.0HESS J1632-478 248.04 − .
82 0.108 3.0 2.0Vela X 128.75 − . · · · · · · Geminga 98.48 17 . · · · · · · Crab Nebula 83.63 22 .
01 0.556 1.1 2.8MGRO J2019+37 305.22 36 .
83 0.224 3.5 3.6Star formation region Cyg OB2 308.08 41 .
51 0.135 4.2 3.4Supernova remnant RCW 86 220.68 − .
48 0.582 0.5 1.0RX J0852.0-4622 133.00 − . · · · · · · RX J1713.7-3946 258.25 − .
75 0.042 5.3 2.2W28 270.43 − .
34 0.159 4.3 2.2IC443 94.18 22 . · · · · · · Cas A 350.85 58 .
81 0.261 2.0 3.4TYCHO 6.36 64 . · · · · · · Starburst/radio galaxy Cen A 201.37 − .
02 0.629 1.0 2.6M87 187.71 12 .
39 0.438 1.8 2.63C 123.0 69.27 29 .
67 0.379 2.2 3.0Cyg A 299.87 40 .
73 0.276 2.6 3.4NGC 1275 49.95 41 .
51 0.479 1.0 3.8M82 148.97 69 .
68 0.251 0.8 2.0Seyfert galaxy ESO 139-G12 264.41 − .
94 0.096 3.0 2.0HMXB/mqso Cir X-1 230.17 − .
17 0.372 0.8 1.0GX 339-4 255.70 − .
79 0.052 4.3 2.2LS 5039 276.56 − .
83 0.444 1.7 2.2SS433 287.96 4 .
98 0.086 8.7 3.8HESS J0632+057 98.25 5 . · · · · · · Cyg X-1 299.59 35 .
20 0.382 2.2 3.6Cyg X-3 308.11 40 .
96 0.137 4.2 3.4LSI 303 40.13 61 . · · · · · · Massive star cluster HESS J1614-518 63.58 − .
82 0.330 1.3 1.6 † Most significant source in the catalog, yielding p postpost