Search for annihilating dark matter in the Sun with 3 years of IceCube data

12 Pages·2017·1.64 MB·English

Eur.Phys.J.C(2017)77:146
DOI10.1140/epjc/s10052-017-4689-9
RegularArticle-ExperimentalPhysics
Search for annihilating dark matter in the Sun with 3years of
IceCube data
IceCubeCollaboration
M.G.Aartsen2,M.Ackermann52,J.Adams16,J.A.Aguilar12,M.Ahlers30,M.Ahrens42,D.Altmann24,
K.Andeen32,T.Anderson48,I.Ansseau12,G.Anton24,M.Archinger31,C.Argüelles14,J.Auffenberg1,S.Axani14,
X.Bai40,S.W.Barwick27,V.Baum31,R.Bay7,J.J.Beatty18,19,J.BeckerTjus10,K.-H.Becker51,S.BenZvi49,
D.Berley17,E.Bernardini52,A.Bernhard34,D.Z.Besson28,G.Binder7,8,D.Bindig51,M.Bissok1,E.Blaufuss17,
S.Blot52,C.Bohm42,M.Börner21,F.Bos10,D.Bose44,S.Böser31,O.Botner50,J.Braun30,L.Brayeur13,
H.-P.Bretz52,S.Bron25,A.Burgman50,T.Carver25,M.Casier13,E.Cheung17,D.Chirkin30,A.Christov25,
K.Clark45,L.Classen35,S.Coenders34,G.H.Collin14,J.M.Conrad14,D.F.Cowen47,48,R.Cross49,M.Day30,
J.P.A.M.deAndré22,C.DeClercq13,E.delPinoRosendo31,H.Dembinski36,S.DeRidder26,P.Desiati30,
K.D.deVries13,G.deWasseige13,M.deWith9,T.DeYoung22,J.C.Díaz-Vélez30,V.diLorenzo31,H.Dujmovic44,
J.P.Dumm42,M.Dunkman48,B.Eberhardt31,T.Ehrhardt31,B.Eichmann10,P.Eller48,S.Euler50,
P.A.Evenson36,S.Fahey30,A.R.Fazely6,J.Feintzeig30,J.Felde17,K.Filimonov7,C.Finley42,S.Flis42,
C.-C.Fösig31,A.Franckowiak52,E.Friedman17,T.Fuchs21,T.K.Gaisser36,J.Gallagher29,L.Gerhardt7,8,
K.Ghorbani30,W.Giang23,L.Gladstone30,T.Glauch1,T.Glüsenkamp24,A.Goldschmidt8,J.G.Gonzalez36,
D.Grant23,Z.Griffith30,C.Haack1,A.Hallgren50,F.Halzen30,E.Hansen20,T.Hansmann1,K.Hanson30,
D.Hebecker9,D.Heereman12,K.Helbing51,R.Hellauer17,S.Hickford51,J.Hignight22,G.C.Hill2,
K.D.Hoffman17,R.Hoffmann51,K.Hoshina30,53,F.Huang48,M.Huber34,K.Hultqvist42,S.In44,A.Ishihara15,
E.Jacobi52,G.S.Japaridze4,M.Jeong44,K.Jero30,B.J.P.Jones14,W.Kang44,A.Kappes35,T.Karg52,A.Karle30,
U.Katz24,M.Kauer30,A.Keivani48,J.L.Kelley30,A.Kheirandish30,J.Kim44,M.Kim44,T.Kintscher52,
J.Kiryluk43,T.Kittler24,S.R.Klein7,8,G.Kohnen33,R.Koirala36,H.Kolanoski9,R.Konietz1,L.Köpke31,
C.Kopper23,S.Kopper51,D.J.Koskinen20,M.Kowalski9,52,K.Krings34,M.Kroll10,G.Krückl31,C.Krüger30,
J.Kunnen13,S.Kunwar52,N.Kurahashi39,T.Kuwabara15,M.Labare26,J.L.Lanfranchi48,M.J.Larson20,
F.Lauber51,D.Lennarz22,M.Lesiak-Bzdak43,M.Leuermann1,L.Lu15,J.Lünemann13,J.Madsen41,G.Maggi13,
K.B.M.Mahn22,S.Mancina30,M.Mandelartz10,R.Maruyama37,K.Mase15,R.Maunu17,F.McNally30,
K.Meagher12,M.Medici20,M.Meier21,A.Meli26,T.Menne21,G.Merino30,T.Meures12,S.Miarecki7,8,
T.Montaruli25,M.Moulai14,R.Nahnhauer52,U.Naumann51,G.Neer22,H.Niederhausen43,S.C.Nowicki23,
D.R.Nygren8,A.ObertackePollmann51,A.Olivas17,A.O’Murchadha12,T.Palczewski7,8,H.Pandya36,
D.V.Pankova48,P.Peiffer31,Ö.Penek1,J.A.Pepper46,C.PérezdelosHeros50,D.Pieloth21,E.Pinat12,
P.B.Price7,G.T.Przybylski8,M.Quinnan48,C.Raab12,L.Rädel1,M.Rameez20,25,a,K.Rawlins3,R.Reimann1,
B.Relethford39,M.Relich15,E.Resconi34,W.Rhode21,M.Richman39,B.Riedel23,S.Robertson2,M.Rongen1,
C.Rott44,T.Ruhe21,D.Ryckbosch26,D.Rysewyk22,L.Sabbatini30,S.E.SanchezHerrera23,A.Sandrock21,
J.Sandroos31,S.Sarkar20,38,K.Satalecka52,P.Schlunder21,T.Schmidt17,S.Schoenen1,S.Schöneberg10,
L.Schumacher1,D.Seckel36,S.Seunarine41,D.Soldin51,M.Song17,G.M.Spiczak41,C.Spiering52,T.Stanev36,
A.Stasik52,J.Stettner1,A.Steuer31,T.Stezelberger8,R.G.Stokstad8,A.Stößl15,R.Ström50,N.L.Strotjohann52,
G.W.Sullivan17,M.Sutherland18,H.Taavola50,I.Taboada5,J.Tatar7,8,F.Tenholt10,S.Ter-Antonyan6,
A.Terliuk52,G.Tešic´48,S.Tilav36,P.A.Toale46,M.N.Tobin30,S.Toscano13,D.Tosi30,M.Tselengidou24,
A.Turcati34,E.Unger50,M.Usner52,J.Vandenbroucke30,N.vanEijndhoven13,S.Vanheule26,M.vanRossem30,
J.vanSanten52,M.Vehring1,M.Voge11,E.Vogel1,M.Vraeghe26,C.Walck42,A.Wallace2,M.Wallraff1,
N.Wandkowsky30,Ch.Weaver23,M.J.Weiss48,C.Wendt30,S.Westerhoff30,B.J.Whelan2,S.Wickmann1,
K.Wiebe31,C.H.Wiebusch1,L.Wille30,D.R.Williams46,L.Wills39,M.Wolf42,T.R.Wood23,E.Woolsey23,
K.Woschnagg7,D.L.Xu30,X.W.Xu6,Y.Xu43,J.P.Yanez23,G.Yodh27,S.Yoshida15,M.Zoll42,b
1III.PhysikalischesInstitut,RWTHAachenUniversity,52056Aachen,Germany
2DepartmentofPhysics,UniversityofAdelaide,Adelaide5005,Australia
123
146 Page 2 of 12 Eur.Phys.J.C(2017)77 :146
3DepartmentofPhysicsandAstronomy,UniversityofAlaskaAnchorage,3211ProvidenceDr.,Anchorage,AK99508,USA
4CTSPS,Clark-AtlantaUniversity,Atlanta,GA30314,USA
5SchoolofPhysicsandCenterforRelativisticAstrophysics,GeorgiaInstituteofTechnology,Atlanta,GA30332,USA
6DepartmentofPhysics,SouthernUniversity,BatonRouge,LA70813,USA
7DepartmentofPhysics,UniversityofCalifornia,Berkeley,CA94720,USA
8LawrenceBerkeleyNationalLaboratory,Berkeley,CA94720,USA
9InstitutfürPhysik,Humboldt-UniversitätzuBerlin,12489Berlin,Germany
10FakultätfürPhysik&Astronomie,Ruhr-UniversitätBochum,44780Bochum,Germany
11PhysikalischesInstitut,UniversitätBonn,Nussallee12,53115Bonn,Germany
12ScienceFacultyCP230,UniversitéLibredeBruxelles,1050Brussels,Belgium
13DienstELEM,VrijeUniversiteitBrussel(VUB),1050Brussels,Belgium
14DepartmentofPhysics,MassachusettsInstituteofTechnology,Cambridge,MA02139,USA
15DepartmentofPhysicsandInstituteforGlobalProminentResearch,ChibaUniversity,Chiba263-8522,Japan
16DepartmentofPhysicsandAstronomy,UniversityofCanterbury,PrivateBag4800,Christchurch,NewZealand
17DepartmentofPhysics,UniversityofMaryland,CollegePark,MD20742,USA
18DepartmentofPhysicsandCenterforCosmologyandAstro-ParticlePhysics,OhioStateUniversity,Columbus,OH43210,USA
19DepartmentofAstronomy,OhioStateUniversity,Columbus,OH43210,USA
20PresentAddress:NielsBohrInstitute,UniversityofCopenhagen,2100Copenhagen,Denmark
21DepartmentofPhysics,TUDortmundUniversity,44221Dortmund,Germany
22DepartmentofPhysicsandAstronomy,MichiganStateUniversity,EastLansing,MI48824,USA
23DepartmentofPhysics,UniversityofAlberta,Edmonton,ABT6G2E1,Canada
24ErlangenCentreforAstroparticlePhysics,Friedrich-Alexander-UniversitätErlangen-Nürnberg,91058Erlangen,Germany
25Départementdephysiquenucléaireetcorpusculaire,UniversitédeGenève,1211Geneva,Switzerland
26DepartmentofPhysicsandAstronomy,UniversityofGent,9000Ghent,Belgium
27DepartmentofPhysicsandAstronomy,UniversityofCalifornia,Irvine,CA92697,USA
28DepartmentofPhysicsandAstronomy,UniversityofKansas,Lawrence,KS66045,USA
29DepartmentofAstronomy,UniversityofWisconsin,Madison,WI53706,USA
30DepartmentofPhysicsandWisconsinIceCubeParticleAstrophysicsCenter,UniversityofWisconsin,Madison,WI53706,USA
31InstituteofPhysics,UniversityofMainz,StaudingerWeg7,55099Mainz,Germany
32DepartmentofPhysics,MarquetteUniversity,Milwaukee,WI53201,USA
33UniversitédeMons,7000Mons,Belgium
34Physik-department,TechnischeUniversitätMünchen,85748Garching,Germany
35InstitutfürKernphysik,WestfälischeWilhelms-UniversitätMünster,48149Münster,Germany
36DepartmentofPhysicsandAstronomy,BartolResearchInstitute,UniversityofDelaware,Newark,DE19716,USA
37DepartmentofPhysics,YaleUniversity,NewHaven,CT06520,USA
38DepartmentofPhysics,UniversityofOxford,1KebleRoad,OxfordOX13NP,UK
39DepartmentofPhysics,DrexelUniversity,3141ChestnutStreet,Philadelphia,PA19104,USA
40PhysicsDepartment,SouthDakotaSchoolofMinesandTechnology,RapidCity,SD57701,USA
41DepartmentofPhysics,UniversityofWisconsin,RiverFalls,WI54022,USA
42DepartmentofPhysics,OskarKleinCentre,StockholmUniversity,10691Stockholm,Sweden
43DepartmentofPhysicsandAstronomy,StonyBrookUniversity,StonyBrook,NY11794-3800,USA
44DepartmentofPhysics,SungkyunkwanUniversity,Suwon440-746,Korea
45DepartmentofPhysics,UniversityofToronto,Toronto,ONM5S1A7,Canada
46DepartmentofPhysicsandAstronomy,UniversityofAlabama,Tuscaloosa,AL35487,USA
47DepartmentofAstronomyandAstrophysics,PennsylvaniaStateUniversity,UniversityPark,PA16802,USA
48DepartmentofPhysics,PennsylvaniaStateUniversity,UniversityPark,PA16802,USA
49DepartmentofPhysicsandAstronomy,UniversityofRochester,Rochester,NY14627,USA
50DepartmentofPhysicsandAstronomy,UppsalaUniversity,Box516,75120Uppsala,Sweden
51DepartmentofPhysics,UniversityofWuppertal,42119Wuppertal,Germany
52DESY,15735Zeuthen,Germany
53EarthquakeResearchInstitute,UniversityofTokyo,Bunkyo,Tokyo113-0032,Japan
Received:20December2016/Accepted:8February2017/Publishedonline:8March2017
©TheAuthor(s)2017.ThisarticleispublishedwithopenaccessatSpringerlink.com
Abstract Wepresentresultsfromananalysislookingfor theSunasourceof GeVneutrinos.IceCubeisabletodetect
darkmatterannihilationintheSunwiththeIceCubeneutrino neutrinoswithenergies>100GeVwhileitslow-energyinfill
telescope. Gravitationally trapped dark matter in the Sun’s arrayDeepCoreextendsthisto>10GeV.Thisanalysisuses
core can annihilate into Standard Model particles making datagatheredintheaustralwintersbetweenMay2011and
May2014,correspondingto532daysoflivetimewhenthe
Sun,beingbelowthehorizon,isasourceofup-goingneu-
ae-mail:[email protected]
be-mail:[email protected] trino events, easiest to discriminate against the dominant
123
Eur.Phys.J.C(2017)77 :146 Page 3 of 12 146
backgroundofatmosphericmuons.Thesensitivityisafac- ofthemuonservesasaproxyforthedirectionoftheinitial
toroftwotofourbetterthanprevioussearchesduetoaddi- neutrinoandallowsustoidentifyadirectionalexcessfrom
tionalstatisticsandimprovedanalysismethodsinvolvingbet- theSuninreconstructedevents.
ter background rejection and reconstructions. The resultant We exploit this fact in the event selection (Sect. 4) for
upperlimitsonthespin-dependentdarkmatter-protonscat- this analysis, using only seasons where the Sun is a source
teringcrosssectionreachdownto1.46×10−5pbforadark of up-going signal events. Furthermore we devise an event
matterparticleofmass500GeVannihilatingexclusivelyinto selection which minimizes atmospheric muon background
τ+τ−particles.Thesearecurrentlythemoststringentlimits contamination and limits the impact of mis-reconstructed
on the spin-dependent dark matter-proton scattering cross events. The remaining samples of events are then analyzed
sectionforWIMPmassesabove50GeV. using an unbinned maximum likelihood ratio method [7],
lookingforanexcessofeventsfromthedirectionoftheSun.
Thismethodcomparestheobservedanglesandenergyspec-
1 Introduction trumtosignalexpectationsfromdifferentsimulatedWIMP
masses and annihilation channels (Sect. 5). Sections 6 and
Astrophysical observations provide strong evidence for the 7 present the results of this analysis as well as their inter-
existence of dark matter (DM). However its nature and pretationintheframeworkofthelargerefforttodetectdark
possible particle constituents remain unknown. Interesting matter.
and experimentally accessible candidates are the so called
‘weaklyinteractingmassiveparticles(WIMPs)’–expected
toexistinthemassrangeofafewGeVstoafewTeVs(see[1] 2 Thedetector
foracomprehensivereview).IfDMconsistsofWIMPs,they
can be gravitationally captured by the Sun [2–5], eventu- IceCubeisacubic-kilometerneutrinodetectorinstalledinthe
allysinkingtoitscore,wheretheymaypair-annihilateinto ice [8] at the geographic South Pole [9] between depths of
standardmodelparticlesproducingneutrinos.Givenenough 1450and2450m.Neutrinoreconstructionreliesontheopti-
time,thecaptureandannihilationprocesseswouldreachan cal detection of Cherenkov radiation emitted by secondary
equilibrium[6]with,onaverage,onlyasmanyDMparticles particles produced in neutrino interactions in the ice or the
annihilatingasarecapturedperunittime.ThisDM-generated nearby bedrock. The photons are detected by photomulti-
neutrinofluxmaybedetectedatterrestrialneutrinodetectors plier tubes (PMT) [10] housed in digital optical modules
suchasIceCube.AstheregionatthecenteroftheSunwhere (DOM) [11]. Construction of the detector started in 2005
mostoftheannihilationswilloccurisverysmall,thesearch and the detector has been running in its complete configu-
isequivalenttolookingforapoint-likesourceofneutrinos. ration since May 2011, with a total of 86 strings deployed,
Neutrinosabove1TeVhaveinteractionlengthssignificantly eachequippedwith60DOMs.
smallerthantheradiusoftheSunandaremostlyabsorbed. TheprincipalIceCubearrayconsistsof78stringsordered
As a result all the signal is expected in the range of a few in a hexagonal grid with a string spacing of approximately
GeVsto∼1TeV. 125m,aninter-DOMspacingof17malongeachstring,and
IceCube (Sect. 2) detects neutrinos by looking for the candetecteventswithenergiesaslowas∼100GeV.Eight
Cherenkovlightfromchargedparticlesproducedintheneu- infill strings are deployed in the central region of IceCube
trino interactions. While charged-current (CC) interactions to form DeepCore, optimized in geometry and instrumen-
ofνμ(andν¯μ)producemuonsthattraversethedetectorpro- tation for the detection of neutrinos at further lower ener-
ducingcleartrack-likesignatures,thevastmajorityofsuch gies,downto∼10GeV.Alayerofdust,causingaregionof
eventsobservedbyIceCubearemuonsproducedwhencos- increased scattering and absorption, intersects the detector
micraysinteractintheupperatmosphere(Sect.3).Although atdepthsbetween1860and2100m.Sincetheicebecomes
they are observed only in the downgoing direction as they more transparent at increasing depth, the main part of the
donotcrosstheEarth,theirdominance innumbers byfive DeepCoreinstrumentationisdeployedbelowthedustlayer
orders of magnitude with respect to the atmospheric neu- withaninter-DOMspacingofonly7m.Avetocapofaddi-
trinofluxrequirestrongmeasuresfortheirrejection.Similar tional 10 DOMs deployed above the dust-layer completes
eventscreatedbytheinteractionsofatmosphericneutrinosin the DeepCore strings. A majority of the DeepCore DOMs
iceare,exceptfortheirspectralcomposition,indistinguish- are equipped with PMTs of higher quantum efficiency to
able from neutrino events of extra-terrestrial origin and so increaselightcollection.TheseDeepCorestrings,alongwith
remainanirreduciblebackground.Acorrectlyreconstructed the seven adjacent standard IceCube strings, constitute the
up-goingeventthusmustcomefromaneutrinointeraction. fiducialregionoftheDeepCoresubarrayforthepurposeof
Thisanalysisfocusesexclusivelyonthesetrack-likeupgoing this analysis [12]. For DM annihilations producing neutri-
events.Attheenergiesrelevanttothisanalysis,thedirection nos above ∼100GeV, the full instrumented volume of the
123
146 Page 4 of 12 Eur.Phys.J.C(2017)77 :146
Fig. 1 Differential νμ (solid) and ν¯μ (dashed) fluxes at Earth from ibleaswigglesintheplotontheleft),aspredictedbyWimpSim[13].
theannihilationsof1TeV(left)and50GeV(right)WIMPsintheSun Theν¯μfluxesarehigherthantheνμfluxesatlowerenergiessincetheir
respectively,includingabsorptionandneutrinooscillationeffects(vis- interactionswiththematteroftheSunarehelicitysuppressed
principalIceCubearraycontributestothesensitivity,while The principal background of muons generated in the
for lower DM masses when the signal neutrinos are below interactions of cosmic rays with the Earth’s atmosphere is
the IceCube threshold, only the DeepCore fiducial volume simulated using the CORSIKA package [17]. Atmospheric
is relevant. The IceCube array nevertheless plays a role in neutrino interactions with the ice and the bedrock sur-
identifying and rejecting background events at these lower roundingthedetectoraresimulatedusingneutrino-generator
energies. (NuGen)[18]above150GeVwiththecrosssectionsof[19]
and GENIE [20] below 150GeV. The atmospheric neu-
trinofluxpredictionsof[22]areusedtoweightNuGenand
3 Signalandbackgroundsimulations GENIEsimulateddatasetstovalidatethedataprocessingand
eventselection.
Neutrino flux predictions at Earth from WIMP annihila-
tions in the Sun have been widely studied, for example in
Ref.[13].WeusethefluxpredictionsfromDarkSUSY[14] 4 Eventselection
andWimpSim[13]tosimulatesignalsfortheIceCubedetec-
toraccordingtospecificannihilationscenarios,incorporating Theenergyrangeoftheexpectedsignal(afewTeVatmax-
effectsfromabsorptionintheSunaswellasneutrinooscil- imum)andtheeventtopologiesinthedetectorattheseener-
lations[15].Eventsfromallthreeflavoursofsignalneutri- giesdictatetheeventselectionstrategies.ForWIMPmasses
+ −
nosaresimulated.WhenWIMPsannihilateintoW W (see less than 200GeV, which produce signal neutrinos mostly
Fig.1),theW bosonsdecaypromptlyandneutrinoemission withenergiesbelowtheIceCubethreshold,onlyDeepCore
fromtheleptonicdecaychannelspeaksatenergiescloseto will contribute significantly towards the effective volume.
themassoftheWIMP.Theτ+τ−channelproducesasimilar However,forhigherWIMPmasses,whereasignificantfrac-
distributionofneutrinosinenergywithahigheroverallnor- tionoftheresultantneutrinosareabovetheIceCubethresh-
malization. These are referred to as ‘hard’ channels. When old, the full instrumented volume of IceCube comes into
theWIMPannihilatespredominantlytoa‘soft’channelsuch play.Consequently weselecttwononoverlapping samples
¯
as bb, the neutrino emission peaks at energies much below ofeventsasillustratedinFig.2.
themassoftheWIMP,sincethebquarkshadronizebefore Tooptimizetheeventselectionsfortheanalysis,wecon-
they can decay to produce neutrinos. WimpSim does not sider two scenarios: WIMPs annihilating completely into
+ − ¯
account for modifications to the spectrum originating from W W and WIMPs annihilating completely into bb. For
theradiationofelectroweakgaugebosonsbytheintermedi- WIMP masses below 80.4GeV, the mass of the W boson,
ateandfinalstatesofthedecayprocess.Theseeffectshave weconsidertheWIMPannihilatingintoτ+τ−,sinceanni-
± + −
been studied in [16]. Since both W and Z bosons decay hilationstoW W arenotkinematicallyallowed.Sincethe
promptlytoproducehighenergyneutrinos,theneteffectof detectoracceptanceisenergydependent,cutshavetobeopti-
theseelectroweakcorrectionsistohardenthefluxesfromthe mized for the spectral composition of the expected signal
softerchannelsandenhancesignalrateexpectations. flux.
123
Eur.Phys.J.C(2017)77 :146 Page 5 of 12 146
(a) (b)
Fig. 2 The two event selection strategies for this analysis. The Ice-
Cubedominatedhighenergysample(a)issensitivetoneutrinosabove
∼100GeV.Mostofthesensitivityforneutrinosignalsbelow100GeV
comesfromtheDeepCore(DC)dominatedlowenergysample(b).This
approachissimilartothatofearlierIceCubeanalyses[21]
WithinIceCube,astandardsetoffilterspre-selectsignal-
Fig. 3 Zenith distributions for simulation, indicated for their simu-
likeeventsandreducetherateofthedominantbackground latedparticledirection(MC,dashedlines)andreconstructeddirection
ofatmosphericmuons,subsequenttowhichreconstructions (reco,solidlines),anddata,withonlyreconstructeddirections(circles),
specifictotheeventtopologyarecarriedout,atwhatisknown at filter level (L2) and analysis level. At filter level the down-going
atmosphericμ-background(reddashed),dominateseventheup-going
as the filter level or level 2 (L2). We focus on a stream of
regionfortherecordeddata(solidcircles),becauseoffalsedirection
datafromthreeofthesefilters,alow-energyeventfilteron reconstructions (solid grey). The flux expectation of atmospheric νμ
the topological region of DeepCore and two further filters (greendashed)areindicated[22].Afterremovalofbackgroundevents
selectingmuon-likeeventsinthebiggerIceCubearray.One intheeventselectionreconstructedtrack-likeatmosphericνμ-events
(greensolid)dominatetheremainingexp.data(opencircles)atfinal
ofthesefiltersfavoursshortlowenergyupwardgoingtracks.
level.TheplotalsoshowstheobtainedlimitsonthesolarWIMPνμ
The other selects general bright track-like events, both up signalfluxobtainedbythisanalysisfortwodifferentWIMPmodels,
anddown-going,wherethelatterclassisrestrictedtoevents whicharereconstructedinDeepCore(50GeVτ+τ−,lightblue)and
starting within the detector. After these filters the data rate IceCube(1TeVW+W−,darkblue)atanalysislevel(solid)andscaled
bytheirselectionefficiencyatfilterlevel(dashed)
is reduced from 3kHz to about 100Hz. Still, atmospheric
muons constitute the overwhelming majority of events. At
This analysis makes use of a new approach for the nec-
this stage, about 30% of the neutrino events recorded by
essarynoisecleaningandseparationofcoincidenteventsby
IceCubeincludeacoincidentatmosphericmuonevent.The
anagglomerativehitclusteringalgorithm[23,24].Itoperates
goalistofurtherreducethedatawithaseriesofreconstruc-
progressively on the IceCube data stream described by the
tionsandcutstoasampleofsignal-likeneutrinoeventsThis
time-distributionofhits.Withinthealgorithm,whichtakes
sample will be, however comprised almost exclusively of
intoaccountthehexagonaldesignofthedetectorandthedif-
atmospheric neutrino events, an irreducible background to
ferenceininstrumentationdensitybetweenitscomponents,
the analysis. Figure 3 provides a comparative summary of
thephysicalcausalrelationbetweenconsecutivehitsisana-
theeventratesatfilterandanalysislevel.
lyzed.Iffoundtobecausallyconnected,hitsareconsidered
toformacluster.Clustersgrowbyfurtheradditionofmore
4.1 Datatreatment connected hits, while unconnected hits are rejected. Each
such identified cluster can later be attributed to a particle
TheprocessingofIceCubedataproceedsinsequentialsteps, (sub)eventwithinthedetector.Persistenterrors,suchasthe
referredtoasselectionlevels.Itinvolvestheabstractionofthe splittingofasingleeventintotwoseparatesubeventsarecor-
recordedanalogtodigitalconverterdataasphotonsimpact- rected by a subsequent algorithm described in [23], which
ingonsinglePMTs(hits),theremovalofnuisancehitscaused probes the recombination of subevents back into a single
by detector noise and coincident events,1 event reconstruc- event. The combination of these algorithms performs 50%
tionsofincreasingcomplexityandeventselectioncuts.The betterthanpreviousapproaches,inbothselectingthecorrect
reconstructionsassumesingleeventtopologiesbuiltuponly hits created by the radiating particle as well as the correct
byhitsthatarecausedbytheradiatingparticle.Theycaneas- separationofeventsarrivingincoincidence.
ilybemisledbynuisancehits,makinghitcleaningapriority
foranyIceCubeanalysis.
4.2 IceCubeeventselection
1 Twoormoreeventsbeingpresentinthedetectoratthesametimeand
endingupinthesamereadoutwindow.Aneffectobservedin∼10%of Fromthe∼100Hz ofdatafromthethreefiltersatL2,cuts
recordedevents,upto30%dependingonfilterstreamselection. favoring horizontal, well reconstructed events are used to
123
146 Page 6 of 12 Eur.Phys.J.C(2017)77 :146
select ∼3Hz of data (L3). The position of the Sun varies requiring the zenith angle to be reconstructed within 10◦
between ∼66◦ and 104◦ in zenith angle. Consequently the oftheactualSunposition.AsecondBDT,whichistrained
signaleventsareexpectedtobehorizontalwithinthedetector. ontheexactsignalpropertiesbyadditionalenergy-sensitive
Subsequently, events thathave morehits outsidethe Deep- eventvariables,isusedtofurtherrefinetheeventclassifica-
Corefiducialvolumeoratleast7hitsintheIceCubestrings tion(L6).AloosecutonthesecondBDT-scoresufficiently
areselected.Moresophisticatedandcomputationallyinten- reducesthesample,sothatthecomputationallydemanding
sivereconstructionsareperformedatthisstage.ABayesian energyreconstructiondescribedin[28]canbeappliedtoall
likelihood-based reconstruction that uses the prior knowl- remainingevents(L7).Thisreconstructionestimatesthetotal
edgethatthedataarestilldominatedbydown-goingmuons energyoftheincomingneutrinofromthelengthofthemuon
is used, along with consistency tests between the various track as well as the photons from hadronic debris from the
trackreconstructionsperformedsofar.Thisreducesthedata charged current interaction, when the interaction has taken
rateto∼140mHz(L4).Subsequently,aBoosted[25]Deci- placewithintheinstrumenteddetectorvolume.Thesample
sionTree(BDT)isusedtoquantifyeacheventassignalor nowcontainsallvariablesneededforthelikelihoodanalysis
background-like using a score, based on a set of variables procedure.TheBDT-scorecutcanbefurtheroptimizedfor
describing the event topology and direction, as well as rel- thebestsensitivityforabroadrangeofWIMPmodelsatthe
ative positions and arrival times of the various photon hits lowWIMPmassendandobtainthesampleatL8.
withinthedetector.TheBDTistrainedonsimulatedsignal Theselectioncriteriadescribedinthetwosectionsabove
+ −
eventsoftheW W -annihilationchannelof1TeVWIMPs. are applied to data from the austral winters between May
TheoptimumthresholdontheBDTscorewasdetermined 2011 and March 2014. This produces two non overlapping
usingtheModelRejectionFactormethoddescribedin[26] sampleswiththeνμ+ν¯μeffectiveareasandangularresolu-
forthesamesignalhypothesis.Theremaining∼2.9mHzof tionsshowninFig.4,correspondingto532daysofoperation
data (L5) are dominated by up-going muons from charged ofIceCube-DeepCore.WhilethisTables1and2summarize
current interactions of atmosphericνμ (and ν¯μ).Theangu- theratesandneutrinopuritiesofthetwostreamsatvarious
lar resolution of this sample is further improved using a levelsoftheeventselection.
reconstruction which utilizes tabulated photon arrival time During the austral summer, when the Sun is above the
distributionsobtainedfromsimulationasdescribedin[27]. horizon and a source of down-going neutrinos, an addi-
Themedianneutrinoangularresolutionforthisfinalsample tionalbackgroundofdown-goingatmosphericmuons,∼105
rangesfrom∼6◦fora100GeVneutrinoto<1◦fora1TeV higherinratethanatmosphericneutrinosatfilterlevel,dom-
neutrino. inatesoverthesignal.Forthisdatatakingperiod,inorderto
reachasampleofsuitableeventsforanalysis,considerably
4.3 DeepCoreeventselection hardercutsarerequired,diminishingtheacceptanceofneu-
trinoevents.Samplesisolatedfromtheseperiodsofoperation
ThefiducialregionofDeepCoreisalreadyembeddeddeep of IceCube-DeepCore [23,29] have been found to not con-
within the detector. In consequence, using a selection of tributesignificantlytothesensitivityandarethusnotfurther
DeepCoredominatedeventswithlessthan7hitsonregular considered.
IceCube strings (the compliment of the selection criterion
for the IceCube event selection) already provides a certain
degree of background rejection via containment and start- 5 Analysismethod
ingrequirementforevents.Thesetwopropertiesarefurther
exploited for the identification of events originating from
An unbinned maximum likelihood ratio method [7] is sub-
neutrinointeractions.
sequentlyusedtolookforastatisticallysignificantexcessof
TheDeepCoreeventselectionstartswitheventsselected
eventsfromthedirectionoftheSun.Thesignalprobability
byanyfilteratL2.Straightcutswhichenforceminimalevent
densityfunction(p.d.f.),explicitlydependentontheevent’s
qualityandalooseselectionforlowenergyhorizontalevents reconstructeddirection,x(cid:4),energy,E ,andobservationtime,
i i
are applied. To reject down-going events still contained in t ,isgivenby:
i
thesample,hit-basedvetosrequiringnohitsintheouterand
top-mostDOMsareapplied.Subsequently,eventsarerecon- Si(x(cid:4)i,ti,Ei,mχ,cχ)=K(|x(cid:4)i −x(cid:4)(cid:5)(ti)|,κi)
structedwiththereconstructiondescribedin[27]followedby ×Emχ,cχ(Ei), (1)
furtherstraightcuts,whichreducethecontentofatmospheric
muonswithinthesample(L3+L4).AfterthisaBDT,which where K stands for the spatial and E for the spectral parts
istrainedontheselectionoftrack-likeνμ-eventsbyvariables ofthep.d.f.andmχ andcχ standforthemassandannihila-
expressingtheposition,incomingdirectionandreconstruc- tionchannel of the WIMPrespectively. Here K isapproxi-
tion quality of events, is used. Thereafter a cut is applied matedbythemonovariateFisher-Binghamdistribution[30]
123
Eur.Phys.J.C(2017)77 :146 Page 7 of 12 146
Fig. 4 Event selection performance. Left νμ + ν¯μ effective area, DeepCoreselection.ThosethatareincludedintheDeepCoreselection
derivedfromMonteCarlosimulationsperformedusingGENIE[20]for aretheonesoflowerenergyandthoseinwhichalargerfractionofthe
theDeepCoreselectionandNuGen[18]employingthecrosssections neutrinoenergyhasbeentransferredtothehadroniccascade.Forthe
ascalculatedby[19]fortheIceCubeselection.Rightmedianangular latterevents,alargerfractionoftheobservedphotonyieldcomesfrom
resolutionasafunctionoftrueneutrinoenergy.Thedashedlinesindi- thehadroniccascade,affectingtheperformanceofthetrackdirection
catethemediankinematicanglebetweentheincomingneutrinoandthe reconstruction.Inadditionthekinematicscatteringanglealsoishigher
muon.EventsfromCCinteractionsofneutrinoswithenergieshigher forsuchevents.Consequently,asaturationeffectcanbeseeninboth
than∼100GeVarepreferentiallyincludedintheIceCubeselectionas angularresolutionandkinematicanglelinesfortheDeepCoreSelec-
therangeofthemuonishigherthanthecontainmentrequirementsofthe tion,atenergiesof∼100GeV
Table 1 Rate summary for the IceCube event selection. The signal totalMonteCarlorateisduetodeficienciesinCORSIKA.Ascutsreject
efficiencies are with respect to L2 for the 1TeV→ W+W− signal. mostoftheatmosphericmuonbackground,thediscrepancybecomes
Theatmosphericmuonneutrinoratesindicatethesumofνμandν¯μin smaller.Thefinalanalysismethodusesrandomizeddatatoestimatethe
theexpectedratio.ThediscrepancybetweenthedatarateatL2andthe background,andisnotaffectedbythisdiscrepancy
Cutlevel Data(Hz) CORSIKA(Hz) Atmosνμ(Hz) Sigeff(%)
L2 98.4 72.8 18.6×10−3 100
L3 2.81 3.32 8.4×10−3 82
L4 0.14 0.15 5.2×10−3 63
L5 2.9×10−3 0.8×10−3 2.1×10−3 37
Table2 RatesummaryfortheDeepCoreeventselection.ThesignalefficienciesarewithrespecttoL2forthe50GeV→τ+τ−–signal
Cutlevel Data(Hz) CORSIKA(Hz) Atmosνμ(Hz) Sigeff(%)
L2 25.6 25.1 5.90×10−3 100
L3 1.37 1.03 3.36×10−3 58
L4 0.74 0.66 3.24×10−3 57
L5 0.55 0.48 2.48×10−3 42
L6 66.9×10−3 59.6×10−3 2.140×10−3 38
L7 1.82×10−3 2.07×10−3 0.423×10−3 16
L8 0.334×10−3 0.143×10−3 0.220×10−3 10
123
146 Page 8 of 12 Eur.Phys.J.C(2017)77 :146
Fig. 6 DistributionofcosineoftheopeninganglestowardstheSun
observedineventsoftheIceCube(top)andDeepCore(bottom)sam-
Fig. 5 Distributionofthereconstructedenergy(Reco)oftheevents ples.Theblackdots representthenumberofeventsreconstructedat
fortheDeepCoreselection.Forbackgroundandsignalsimulations,the thecorrespondingdirection,theredlinesaretheaveragebackground
trueenergydistribution(MC)isshownindashedlines.Thedistribu- expectationswiththegrayshadedregionscorrespondingtostatistical
tionsforthesignalarescaleddownbyafactorof1/2forimproved uncertaintiesonthebackgroundexpectations,whilethebluelinesindi-
visualization.eps catetheeventsexpectedfromWIMPsofmasses1TeVannihilatinginto
W+W−at2.84×1019s−1and50GeVannihilatingtoτ+τ−attherate
of3.46×1022s−1respectively,thepresentupperlimits.Theanalysis
fromdirectionalstatistics,dependentontheopeningangle,θ, methodemployedisunbinnedindirection,consequentlythebinning
betweentheeventandthedirectionoftheSunatobservation employedinthisfigureisforindicativepurposesonly
time,x(cid:4)(cid:5)(ti),givenby
K(|x(cid:4)i −x(cid:4)(cid:5)(ti)|,κi)= κi2eπκi(ceoκsi(θ−|x(cid:4)i−ex(cid:4)−(cid:5)κ(iti))|) (2) tihneRyecfa.n[b3e2c].oCmobninfieddesntcaetisintitcearlvlaylussoinngtthheemnuemthboedrdoefscsriigbneadl
events present within the sample are constructed using the
The concentration factor κ of the event i is obtained from
i methodof[33].
thelikelihood-basedestimateoftheangularresolutionofthe
trackreconstruction[31].Theenergypartofthesignalp.d.f.
5.1 Systematicuncertainties
isconstructedfromsignalsimulations.
Thebackgroundp.d.f.is:
Background levels are estimated in this analysis method
B (x(cid:4),E )= D(δ )× P(E |φ ) (3) using data randomized in right ascension (see Fig. 6) and
i i i i i atm
soare,byconstruction,freeofsignificantsystematicuncer-
where D(δ ) is the declination dependence and P(E|φ )
i atm tainties.Apreviousstudyofthesignaluncertaintiesondata
indicates the distribution of the energy estimator E in the
fromthe79-stringconfigurationofIceCube[21]concluded
eventsamplewhichisconstructedfromthedatadominated
thatthefollowingsourcesofuncertaintyareintrinsictothe
byatmosphericneutrinos.
signalsimulation(percentageimpactonsensitivityinparen-
Theenergypartofthesignalandbackgroundp.d.f.sare
thesis):
thedistributionsofreconstructedenergyobtainedfromsignal
simulations and observed data, respectively, and are used
1. Neutrino-nucleon cross sections (7% at mχ <35GeV
onlyfortheDeepCoresample.TheyareillustratedinFig.5.
downto3.5%formχ >100GeV)
For a sample of N events consisting of n signal events
s 2. Uncertaintiesinneutrinooscillationparameters(6%)
fromtheSunand N −n backgroundevents,thelikelihood
s 3. Uncertaintiesinmuonpropagationinice(<1%),
canthenbewrittenas:
(cid:2)(cid:3) (cid:3) (cid:4) (cid:4)
n n
L(ns)= sSi + 1− s Bi (4) whilethefollowingsourcesdominatethedetectionprocess
N N
N
Thebestestimateforthenumberofsignaleventsinthesam- 1. AbsoluteDOMefficiency
pleisobtainedbymaximizingthelikelihoodratioasdefined 2. Photonpropagationinice(absorptionandscattering).
inRef.[7].Thesignificanceoftheobservationcanbeesti-
mated without depending on Monte Carlo simulations by Sincethefirstclassofsourcesofuncertainties,directinputs
repeatingtheprocessondatasetsrandomizedinrightascen- totheWimpSimsignalgeneratorthatmostlyaffectthesig-
sion.Asthetwoeventselectionshavenoeventsincommon, nalfluxnormalisations,havenotsignificantlychangedwith
123
Eur.Phys.J.C(2017)77 :146 Page 9 of 12 146
Table3 Systematicssummarystatinguncertaintiesinthedetectionprocess.Thelastcolumnstatesthetotaluncertaintyestimate,whichisthe
largestsuminquadratureoftheindividualuncertainties
mχ (GeV) Annih.channel AbsoluteDOMeff(%) Photonprop.inice(%) Total(%)
20 τ+τ− −11/+29 −13/+18 35
50 τ+τ− −8/+23 −9/+13 29
100 W+W− −9/+19 −9/+11 23
500 bb¯ −7/+11 −8/+7 15
1000 W+W− −6/+9 −6/+4 12
respect to the study in Ref. [21], we assume them to be direct detection. Even though these tend to be significantly
similar. The second class of sources of uncertainties also stronger,theyaresubjecttodifferentuncertaintiesfromthe
includeeffectsalteringthesignal’sapparentspectralcompo- nuclearscatteringprocessandfromastrophysics.Limitshave
sition in the detection process and are thus more important improved by a factor of ∼2 to 4 with respect to previous
tothisanalysiswhichissensitivetothereconstructedevent IceCubeanalyses[21,42].Whiletheseconstraintsexplicitly
energy. assumeequilibriumbetweencaptureandannihilationinthe
TostudytheeffectoftheabsoluteDOMefficiencyaswell Sun,forthenaturalscaleof(cid:7)σ v(cid:8)∼3×10−26cm3s−1[43],
A
as the absorption and scattering properties of the ice, a set the time required for equilibrium to be achieved becomes
of signal simulations were generated for one year of data as large as the age of the Sun only for σχS−Dp as low as
by individually varying each quantities by ±10% from the 10−43cm2,makingthisaveryreasonableassumption[44].
baselinevalueforcertainbenchmarksignalsofinterest. Theuncertaintiesontheselimitsduetouncertaintiesinveloc-
The percentage impact of these variations on the muon itydistributionsofDMhavebeenquantifiedinRef.[45]and
flux(cid:10)¯μ+μ¯,whichcanbeconvertedtotheotherquantitiesin do not exceed ∼50%. The study also concludes that these
Table4,aresummarizedinTable3.Thepercentageimpactof limits are conservative with respect to the possible exis-
theuncertaintiesonneutrino-nucleoncrosssections,neutrino tence of a dark-disk [46], since a population of DM parti-
oscillationsandmuonpropagationinicearetakenfrom[21] cles with lower velocities in the disk will enhance the cap-
andsummedinquadraturetotheonesfromuncertaintiesin ture rate. For a dark disk contributing an additional 25%
DOMefficiencyandiceopticalpropertiestoobtainthetotal to the local DM density, co-rotating with the visible stel-
systematicuncertainty. lardiskwithnolaginvelocityandavelocitydispersionof
σ = 50km/s,thespin-dependentcapturerateisboostedby
afactorof∼20athighDMmasses,improvingtheconstraint
6 Results on the spin-dependent cross section by a corresponding
amount.
No significant excess of events over the expected back- AsdemonstratedinFig.7,theseconstraintsexcludesome
ground was found in the direction of the Sun, allowing us models corresponding to neutralinos from a scan of ∼500
to setlimitson theneutrino flux fromthe Sun inthe GeV– million points in the 19 parameter realization of the phe-
TeV range. Assuming a conservative local DM density of nomenological minimally super-symmetric standard model
0.3GeV/cm3 [41],astandardMaxwellianhalovelocitydis- (pMSSM)[47,48]performedusingmicrOMEGAs[49]with
tributionandtheStandardSolarModel,thislimitcanalsobe logarithmically distributed priors (chosen to preferentially
interpreted as a limit on the WIMP-proton scattering cross populatelowmass,highσχS−Dp models)onthemassparam-
section. Table 4 summarizes the best fit number of signal eters typically in the range 50–10,000GeV/c2. The points
eventsandtheupperlimitonthemuonandneutrinofluxes,as were required to yield a relic dark matter density consis-
wellasspin-dependentandspin-independentWIMP-proton tent with PLANCK measurements [50] and a Higgs mass
scatteringcrosssections. withinthecurrentlyknownuncertaintyrange[51],inaddi-
tion to being consistent with recent measurements of the
Bs → μ+μ− branching ratio [52] and the CKM matrix
0
7 Conclusionsandinterpretations elementVub [53].Furtherdetailsaregivenin[54].
BeyondtheWIMPparadigm,thissearchissensitivetoany
For spin-dependent WIMP-proton scattering, IceCube lim- scenario with a DM particle in the 20GeV to 10TeV mass
its are the most competitive in the region above ∼80GeV rangethatcanscatteroffnucleisufficientlystronglytocause
(Fig. 7). The constraints on spin-independent scattering anover-densityatthecenteroftheSun,andcanannihilateto
(Fig.8)fromthissearcharecomplementarytothelimitsfrom produce neutrinos asprimaryorsecondary products.Some
123
146 Page 10 of 12 Eur.Phys.J.C(2017)77 :146
einitsendent .L(pb) 06 05 07 05 07 06 07 08 06 08 08 06 08 08 06 08 08 06 07 08 05 07 07 05 06 07
e-DeepCorspin-indep .σ90%CSI −4.06e −4.77e−6.95e −2.44e−3.02e −7.38e−2.13e−6.48e −3.50e−6.62e−3.52e −2.82e−3.49e−1.35e −2.00e−5.28e−1.60e −4.65e−3.70e−8.25e −1.06e−9.14e−2.19e −3.46e−3.88e−9.11e
ubnd b)
Ca p
perationofIcepin-dependent ..σ90%CL(SD −4.85e04 −9.25e03−1.35e04 −6.39e03−7.90e05 −3.29e03−9.52e05−2.91e05 −2.80e03−5.30e05−2.82e05 −3.06e03−3.76e05−1.46e05 −2.59e03−6.80e05−2.07e05 −6.76e03−5.42e04−1.21e04 −1.58e02−1.37e03−3.28e04 −5.27e02−5.96e03−1.40e03
os )
ofhe −1
threeyearsnrate,andt ..(cid:12)90%CL(sχχ→SM 9.19e+23 7.39e+24 1.08e+23 2.79e+24 3.46e+22 4.09e+23 1.18e+22 3.60e+21 5.96e+22 1.13e+21 5.99e+20 1.66e+22 2.04e+20 7.96e+19 3.56e+21 9.34e+19 2.84e+19 1.04e+21 8.33e+19 1.85e+19 8.74e+20 7.59e+19 1.82e+19 7.31e+20 8.26e+19 1.94e+19
oo
orrespondingtflux,annihilati −−21myear)
c
e,on (k
oflivetimonthemu ..(cid:10)90%CLμ+¯μ 1.52e+05 2.38e+05 3.22e+04 1.27e+05 1.45e+04 3.22e+04 3.73e+03 2.75e+03 8.57e+03 6.02e+02 8.64e+02 3.37e+03 5.53e+01 5.93e+01 1.55e+02 3.31e+01 3.46e+01 7.56e+01 3.13e+01 2.90e+01 7.24e+01 2.84e+01 2.93e+01 6.87e+01 3.11e+01 3.21e+01
daysmits )
∼532perli −1ear
p y
esinasu −2m
wosampld,aswell ¯(cid:10)(kμ+¯μ 1.58e+05 2.44e+05 3.33e+04 1.29e+05 1.39e+04 3.05e+04 2.81e+03 2.13e+03 6.79e+03 5.03e+02 7.74e+02 3.04e+03 4.04e+01 4.71e+01 1.30e+02 3.06e+01 3.30e+01 7.29e+01 3.07e+01 2.85e+01 7.11e+01 2.78e+01 2.86e+01 6.74e+01 3.08e+01 3.18e+01
te
ed
ventswithintharealsoprovi 3V(km)eff −4.40e04 −2.79e04−1.26e03 −4.71e04−2.31e03 −1.39e03−6.64e03−9.40e03 −4.42e03−7.38e02−7.20e02 −1.54e02−1.87e01−1.95e01 −3.24e02−2.67e01−2.86e01 −6.62e02−2.86e01−2.92e01 −7.72e02−3.09e01−3.10e01 −8.26e02−3.18e01−3.19e01
berofsignalehethreeyears ..90%CLns 97.2 96.8 59.1 87.3 48.9 65.2 36.1 37.6 55.1 64.7 90.6 75.6 36.0 45.1 43.1 24.6 28.6 32.1 23.1 21.1 33.7 22.4 22.3 32.5 25.2 25.0
umert %
nv e
ontheumeso pvalu >50 >50>50 >50 48.4 46.1 34.7 31.3 28.2 39.8 42.1 46.1 39.3 38.7 37.2 48.9 46.5 48.2 49.6 49.4 49.1 49.8 49.8 49.8>50>50
upperlimitseffectivevolections Dataset DC DC DC DC DC DC DC DC DC+IC DC+IC DC+IC DC+IC IC IC IC IC IC IC IC IC IC IC IC IC IC IC
valuesand90%C.L.uration.Theaverageonscatteringcrosss Annih.channel +−ττ ¯bb+−ττ ¯bb+−ττ ¯bb+−WW+−ττ ¯bb+−WW+−ττ ¯bb+−WW+−ττ ¯bb+−WW+−ττ ¯bb+−WW+−ττ ¯bb+−WW+−ττ ¯bb+−WW+−ττ
Pnfigprot V)
Table4finalcoWIMP- m(Geχ 20 35 35 50 50 100 100 100 250 250 250 500 500 500 1000 1000 1000 3000 3000 3000 5000 5000 5000 10,000 10,000 10,000
123