arXiv:1702.06131v2

23 Pages·2017·2.22 MB·German

SubmittedtoAstronomy&Astrophysics (cid:13)cESO2017
February22,2017
Multiwavelength follow-up of a rare IceCube neutrino multiplet
IceCube:M.G.Aartsen2,M.Ackermann114,J.Adams28,J.A.Aguilar16,M.Ahlers66,M.Ahrens99,I.AlSamarai42,D.Altmann40,K.Andeen68,
T.Anderson108,I.Ansseau16,G.Anton40,M.Archinger67,C.Argüelles18,J.Auffenberg1,S.Axani18,X.Bai88,S.W.Barwick58,V.Baum67,R.Bay11,
J.J.Beatty30,31,J.BeckerTjus14,K.-H.Becker113,S.BenZvi91,D.Berley29,E.Bernardini114,A.Bernhard75,D.Z.Besson61,G.Binder12,11,D.Bindig113,
E.Blaufuss29,S.Blot114,C.Bohm99,M.Börner35,F.Bos14,D.Bose101,S.Böser67,O.Botner111,J.Braun66,L.Brayeur17,H.-P.Bretz114,S.Bron42,
A.Burgman111,T.Carver42,M.Casier17,E.Cheung29,D.Chirkin66,A.Christov42,K.Clark105,L.Classen76,S.Coenders75,G.H.Collin18,J.M.Conrad18,
D.F.Cowen108,107,R.Cross91,M.Day66,J.P.A.M.deAndré37,C.DeClercq17,E.delPinoRosendo67,H.Dembinski77,S.DeRidder43,P.Desiati66,
K.D.deVries17,G.deWasseige17,M.deWith13,T.DeYoung37,V.diLorenzo67,H.Dujmovic101,J.P.Dumm99,M.Dunkman108,B.Eberhardt67,T.Ehrhardt67,
B.Eichmann14,P.Eller108,S.Euler111,P.A.Evenson77,S.Fahey66,A.R.Fazely9,J.Feintzeig66,J.Felde29,K.Filimonov11,C.Finley99,S.Flis99,C.-C.Fösig67,
A.Franckowiak114,E.Friedman29,T.Fuchs35,T.K.Gaisser77,J.Gallagher65,L.Gerhardt12,11,K.Ghorbani66,W.Giang38,L.Gladstone66,T.Glauch1,
T.Glüsenkamp40,A.Goldschmidt12,J.G.Gonzalez77,D.Grant38,Z.Griffith66,C.Haack1,A.Hallgren111,F.Halzen66,E.Hansen32,T.Hansmann1,K.Hanson66,
D.Hebecker13,D.Heereman16,K.Helbing113,R.Hellauer29,S.Hickford113,J.Hignight37,G.C.Hill2,K.D.Hoffman29,R.Hoffmann113,K.Hoshina66,104,
F.Huang108,M.Huber75,K.Hultqvist99,S.In101,A.Ishihara26,E.Jacobi114,G.S.Japaridze7,M.Jeong101,K.Jero66,B.J.P.Jones18,W.Kang101,A.Kappes76,
7 T.Karg114,A.Karle66,U.Katz40,M.Kauer66,A.Keivani108,J.L.Kelley66,A.Kheirandish66,J.Kim101,M.Kim101,T.Kintscher114,J.Kiryluk100,T.Kittler40,
1 S.R.Klein12,11,G.Kohnen70,R.Koirala77,H.Kolanoski13,R.Konietz1,L.Köpke67,C.Kopper38,S.Kopper113,D.J.Koskinen32,M.Kowalski13,114,
0 K.Krings75,M.Kroll14,G.Krückl67,C.Krüger66,J.Kunnen17,S.Kunwar114,N.Kurahashi84,T.Kuwabara26,A.Kyriacou2,M.Labare43,J.L.Lanfranchi108,
2 M.J.Larson32,F.Lauber113,M.Lesiak-Bzdak100,M.Leuermann1,L.Lu26,J.Lünemann17,J.Madsen90,G.Maggi17,K.B.M.Mahn37,S.Mancina66,
M.Mandelartz14,R.Maruyama78,K.Mase26,R.Maunu29,F.McNally66,K.Meagher16,M.Medici32,M.Meier35,T.Menne35,G.Merino66,T.Meures16,
b S.Miarecki12,11,J.Micallef37,G.Momenté67,T.Montaruli42,M.Moulai18,R.Nahnhauer114,U.Naumann113,G.Neer37,H.Niederhausen100,S.C.Nowicki38,
e D.R.Nygren12,A.ObertackePollmann113,A.Olivas29,A.O’Murchadha16,T.Palczewski12,11,H.Pandya77,D.V.Pankova108,P.Peiffer67,Ö.Penek1,
F J.A.Pepper106,C.PérezdelosHeros111,D.Pieloth35,E.Pinat16,P.B.Price11,G.T.Przybylski12,M.Quinnan108,C.Raab16,L.Rädel1,M.Rameez32,
K.Rawlins6,R.Reimann1,B.Relethford84,M.Relich26,E.Resconi75,W.Rhode35,M.Richman84,B.Riedel38,S.Robertson2,M.Rongen1,C.Rott101,
0
T.Ruhe35,D.Ryckbosch43,D.Rysewyk37,L.Sabbatini66,S.E.SanchezHerrera38,A.Sandrock35,J.Sandroos67,S.Sarkar32,81,K.Satalecka114,P.Schlunder35,
2 T.Schmidt29,S.Schoenen1,S.Schöneberg14,L.Schumacher1,D.Seckel77,S.Seunarine90,D.Soldin113,M.Song29,G.M.Spiczak90,C.Spiering114,
J.Stachurska114,T.Stanev77,A.Stasik114,J.Stettner1,A.Steuer67,T.Stezelberger12,R.G.Stokstad12,A.Stößl26,R.Ström111,N.L.Strotjohann114,
] G.W.Sullivan29,M.Sutherland30,H.Taavola111,I.Taboada8,J.Tatar12,11,F.Tenholt14,S.Ter-Antonyan9,A.Terliuk114,G.Tešic´108,S.Tilav77,P.A.Toale106,
E M.N.Tobin66,S.Toscano17,D.Tosi66,M.Tselengidou40,C.F.Tung8,A.Turcati75,E.Unger111,M.Usner114,J.Vandenbroucke66,N.vanEijndhoven17,
HS.Vanheule43,M.vanRossem66,J.vanSanten114,M.Vehring1,M.Voge15,E.Vogel1,M.Vraeghe43,C.Walck99,A.Wallace2,M.Wallraff1,N.Wandkowsky66,
A.Waza1,Ch.Weaver38,M.J.Weiss108,C.Wendt66,S.Westerhoff66,B.J.Whelan2,S.Wickmann1,K.Wiebe67,C.H.Wiebusch1,L.Wille66,D.R.Williams106,
h. L.Wills84,M.Wolf99,T.R.Wood38,E.Woolsey38,K.Woschnagg11,D.L.Xu66,X.W.Xu9,Y.Xu100,J.P.Yanez38,G.Yodh58,S.Yoshida26,M.Zoll99
p ASAS-SN:K.Z.Stanek31,30,B.J.Shappee82,54,C.S.Kochanek31,30,T.W.-S.Holoien31,30,J.L.Prieto96,97
- TheAstrophysicalMultimessengerObservatoryNetwork:D.B.Fox107,109,110,J.J.DeLaunay108,109,C.F.Turley108,109,S.D.Barthelmy46,A.Y.Lien46,
o
P.Mészáros108,107,109,110,K.Murase108,107,109,110
r
t Fermi:D.Kocevski46,R.Buehler114,M.Giomi114,J.L.Racusin46
s
aHAWC:A.Albert63,R.Alfaro20,C.Alvarez25,J.D.Álvarez72,R.Arceo25,J.C.Arteaga-Velázquez72,H.A.AyalaSolares53,E.Belmont-Moreno20,A.Bernal19,
[ C.Brisbois53,K.S.Caballero-Mora25,T.Capistrán86,A.Carramiñana86,S.Casanova60,M.Castillo72,U.Cotti72,E.delaFuente48,S.CoutiñodeLeón86,
C.DeLeón87,R.DiazHernandez86,B.L.Dingus63,M.A.DuVernois66,J.C.Díaz-Vélez48,66,D.W.Fiorino29,N.Fraija19,J.A.García-González20,

1 M.Gerhardt53,A.GonzálezMuñoz20,M.M.González19,J.A.Goodman29,Z.Hampel-Arias66,J.P.Harding63,S.Hernandez20,C.M.Hui55,P.Hüntemeyer53,
v A.Iriarte19,A.Jardin-Blicq50,V.Joshi50,S.Kaufmann25,A.Lara21,R.J.Lauer3,W.H.Lee19,D.Lennarz7,H.LeónVargas20,J.T.Linnemann37,
1 G.LuisRaya51,R.Luna-García22,R.López-Coto50,K.Malone108,S.S.Marinelli37,O.Martinez87,I.Martinez-Castellanos29,J.Martínez-Castro22,
3 H.Martínez-Huerta23,J.A.Matthews3,E.Moreno87,M.Mostafá108,L.Nellen24,M.Newbold92,M.U.Nisa91,R.Pelayo22,J.Pretz109,E.G.Pérez-Pérez51,
Z.Ren3,C.D.Rho91,C.Rivière29,D.Rosa-González86,M.Rosenberg109,F.SalesaGreus60,A.Sandoval20,M.Schneider95,H.Schoorlemmer50,G.Sinnis63,
1
A.J.Smith29,R.W.Springer92,P.Surajbali50,O.Tibolla25,K.Tollefson37,I.Torres86,L.Villaseñor72,T.Weisgarber66,I.G.Wisher66,J.Wood66,T.Yapici37,
6
A.Zepeda23,H.Zhou63
0
LCO:I.Arcavi44,94,93,39,G.Hosseinzadeh44,93,D.A.Howell44,93,S.Valenti34,C.McCully44,93
.
2 MASTER:V.M.Lipunov73,74,E.S.Gorbovskoy74,N.V.Tiurina74,P.V.Balanutsa74,A.S.Kuznetsov74,V.G.Kornilov73,74,V.Chazov74,N.M.Budnev57,
0 O.A.Gress57,K.I.Ivanov57,A.G.Tlatov59,R.ReboloLopez103,M.Serra-Ricart103
7 Swift:P.A.Evans62,J.A.Kennea107,N.Gehrels46(cid:63),J.P.Osborne62,K.L.Page62
1
VERITAS:A.U.Abeysekara92,A.Archer98,W.Benbow4,R.Bird64,T.Brantseg5,V.Bugaev98,J.VCardenzana5,M.P.Connolly41,W.Cui112,10,
:
v A.Falcone107,Q.Feng71,J.P.Finley112,H.Fleischhack114,L.Fortson69,A.Furniss49,S.Griffin71,98,J.Grube52,M.Hütten114,O.Hervet95,J.Holder77,
i G.Hughes4,T.B.Humensky79,C.A.Johnson95,P.Kaaret56,P.Kar92,N.Kelley-Hoskins114,M.Kertzman47,M.Krause114,S.Kumar77,M.J.Lang41,
X T.T.Y.Lin71,S.McArthur112,P.Moriarty41,R.Mukherjee80,D.Nieto79,R.A.Ong64,A.N.Otte8,M.Pohl85,114,A.Popkow64,E.Pueschel36,J.Quinn36,
K.Ragan71,P.T.Reynolds33,G.T.Richards8,E.Roache4,C.Rulten69,I.Sadeh114,M.Santander80,G.H.Sembroski112,D.Staszak27,S.Trépanier71,J.Tyler71,
r
a S.P.Wakely27,A.Weinstein5,P.Wilcox56,A.Wilhelm85,114,D.A.Williams95,B.Zitzer71
E.Bellm83,Z.Cano45,A.Gal-Yam89,D.A.Kann102,E.O.Ofek89,M.Rigault13,M.Soumagnac89
(Affiliationscanbefoundafterthereferences)
February22,2017
Sendoffprintrequeststo:[email protected]
(cid:63) Deceased:6Feb2017
Articlenumber,page1of23
SubmittedtoAstronomy&Astrophysics
Abstract
OnFebruary172016,theIceCubereal-timeneutrinosearchidentified,forthefirsttime,threemuonneutrinocandidatesarrivingwithin100s
ofeachotherwhichareconsistentwithapointsourceorigin.Suchatripletisexpectedonlyonceevery13.7yearsasarandomcoincidenceof
background events. However, considering the lifetime of the follow-up program the probability to detect at least one triplet from atmospheric
backgroundsis32%.Follow-upobservatorieswerenotifiedinordertosearchforanelectromagneticcounterpart.Observationswereobtainedby
Swift’sX-raytelescope,byASAS-SN,LCOandMASTERatopticalwavelengths,andbyVERITASintheveryhighenergygamma-rayregime.
Moreover,theSwiftBATserendipitouslyobservedthelocation100safterthefirstneutrinowasdetected,anddatafromtheFermiLATandHAWC
wereanalyzed.Wepresentdetailsoftheneutrinotripletandthefollow-upobservations.Nolikelyelectromagneticcounterpartwasdetected,and
we discuss the implications of these constraints on candidate neutrino sources such as gamma-ray bursts, core-collapse supernovae and active
galacticnucleusflares.Thisstudyillustratesthepotentialofandchallengesforfuturefollow-upcampaigns.
Keywords.astroparticlephysics—neutrinos—Gamma-rayburst:general—supernovae:general—Galaxies:active—X-rays:bursts
1. Introduction gamma-rayoutburstofablazarwhichwasalignedwithamulti
PeV neutrino event. However, all of these associations have a
In2013,theIceCubeneutrinoobservatorypresentedthefirstev-
chance-coincidence probability of a few percent and are hence
idenceforahigh-energyfluxofcosmicneutrinos(Aartsenetal.
notsignificantdetections.
2013, 2015a). While the evidence for their existence continues
tomount,noexplicitsourceshavebeenidentified(seee.g.Aart- Themostenergeticneutrinocandidatedetectedsofar,witha
sen et al. 2014, 2017). The arrival directions of the events are depositedenergyof2.6PeV,wasobservedinJune2014(Schoe-
distributed isotropically which likely implies that many events nen & Raedel 2015; Aartsen et al. 2016a). The probability that
areofextragalacticorigin. thiseventwasproducedintheEarth’satmosphereissmallerthan
High-energyneutrinosareproducedwhencosmicraysinter- 1% and the angular uncertainty is 0.27◦ (at 50% confidence)
act with ambient matter (pp interactions) or photon fields (pγ which makes it one of the best localized events observed with
interactions).Theseinteractionsareexpectedtohappenmainly IceCube.However,notimelyfollow-upobservationsweretrig-
within cosmic-ray sources where the target photon and/or mat- gered and a transient counterpart could have gone unnoticed.
terdensitiesarehigh.Thedetectionofaneutrinosourcewould Since mid-2016, such events are identified, reconstructed and
implythatthissourcealsoacceleratescosmicrays. published within minutes (Aartsen et al. 2016f) to allow quick
Cosmicrayscanbeacceleratedatcollisionlessshockfronts follow-upobservations(seeBlaufuss2016asanexampleforthe
which are expected in a wide variety of astrophysical objects. firstpublishedevent).
Among those are gamma-ray bursts (GRBs; see e.g.: Baerwald
In addition to the publicly announced high-energy neutrino
etal.2015;Bustamanteetal.2015;Mészáros2015),aswellas
alerts, IceCube has a real-time program that searches for mul-
the related class of low-luminosity GRBs (LLGRBs) or core-
tiple neutrinos from a similar direction (Abbasi et al. 2012b;
collapse supernovae (CCSNe) containing a choked jet (Senno
Aartsen et al. 2016f). When two or more muon neutrino can-
etal.2016;Tamborra&Ando2016;Fraija2014).CCSNecould
didatesaredetectedwithin100sofeachotheropticalandX-ray
in addition produce cosmic rays when their ejecta interact with
observations can be triggered automatically (Evans et al. 2015;
circumstellarmediumemittedbythestarpriortotheexplosion
Aartsenetal.2015b).Real-timefollow-upobservationsarealso
(Murase et al. 2014; Murase & Ioka 2013; Katz et al. 2011).
triggered by the ANTARES neutrino telescope, but have not
Otherpotentialneutrinosourcesareactivegalacticnuclei(AGN;
leadtothediscoveryofanelectromagneticcounterpart(Adrián-
seeMurase2015forareview),tidaldisruptionevents(Wang&
Liu2016;Pfefferetal.2017;Farrar&Piran2014)andstarburst Martínezetal.2016;Ageronetal.2012).
galaxies(Waxman2015;Tamborraetal.2014).
InFebruary2016,wefound–forthefirsttime–threeevents
Thus far dedicated searches for correlations with specific
within this 100s time window. The detection of such a triplet
source classes have not yielded a significant detection. At 90%
from atmospheric backrounds is not unlikely considering that
confidence level GRBs can at most account for 1% of the de-
the search has been running since December 2008 (compare
tected flux (Aartsen et al. 2015c) and the contribution from
Sect.3.2).However,sinceitisthemostsignificantneutrinomul-
blazarshasbeenlimitedtoatmost30%(Aartsenetal.2016b).
tiplet detected so far, multiwavelength follow-up observations
The non-detection of any neutrino sources implies that the as-
weretriggeredtosearchforapotentialelectromagneticcounter-
trophysicalfluxmustoriginatefromalargenumberofrelatively
part.
faint neutrino sources (Ahlers & Halzen 2014; Kowalski 2015;
Murase&Waxman2016). In this paper we present details of the neutrino triplet and
Several coincidences of neutrino events with astrophysical results of the follow-up observations. In Sect. 2 we introduce
sources have been reported in the literature. For example a su- the follow-up program. The properties of the triplet are given
pernovaofTypeIInwasdetectedinfollow-upobservationsofa in Sect. 3. The follow-up observations, covering from optical
neutrinodoublet(Aartsenetal.2015b).Itishoweverlikelyunre- wavelengthsuptovery-high-energy(VHE)gammarays,arepre-
latedgiventhelargeimpliedneutrinoluminosity.Padovanietal. sented in Sect. 4. Finally, in Sect. 5 we draw conclusions from
(2016) observe a correlation between extreme blazars and high the various observations and discuss the sensitivity of our pro-
energy neutrino events and Kadler et al. (2016) found a bright gramtocandidateneutrinosourceclasses.
Articlenumber,page2of23
IceCubeetal.:Follow-upofaneutrinomultiplet
2. TheIceCubeFollow-upProgram inatmosphericshowersandoutof∼105detectedeventsperyear
only several hundreds are expected to be of cosmic origin (see
2.1. TheIceCubeNeutrinoTelescope Sect.5.1).Toovercomethisbackgroundwerestrictoursearchto
shorttransientsourceswhicharedetectedwithseveralneutrinos.
IceCubeisacubic-kilometer-sizedneutrinodetectorinstalledin
theiceatthegeographicSouthPolebetweenadepthof1,450m
and 2,450m (Aartsen et al. 2016e). An array of 5,160 digital 2.3. Alertgeneration
opticalmodules(DOMs;Abbasietal.2009,2010a),whichare
deployed in the ice, detects the Cherenkov radiation from sec- TheIceCubeopticalfollow-upprogramhasbeenrunningsince
ondary particles produced in neutrino interactions (Achterberg December2008 (Abbasietal.2012b).Afterselectingastream
etal.2006).BasedonthepatternoftheCherenkovlight,boththe dominatedbyupward-goingneutrinoevents,itsearchesforco-
directionandenergyoftheneutrinoscanbemeasured.Thede- incident events. A multiplet alert is generated whenever two or
tectorhasbeenrunninginitsfullconfigurationsinceMay2011. more tracks arrive within 100s with an angular separation of
lessthan3.5◦1.Thelengthofthetimewindowwaschosensuch
Neutrinos can interact and produce secondary particles
thatitcoversthetypicaldurationofaSNcore-collapseandthe
through neutral current (NC) interactions or through charged
lifetime of a jet in a GRB (compare Abbasi et al. 2012b). To
current (CC) interactions. CC interactions induced by electron
measurethesignificanceofaneutrinodoublet,aqualityparam-
ortauneutrinos,aswellasNCinteractionsinducedbyanyneu-
eter is calculated using Eq. 1 in Aartsen et al. (2015b). Based
trino flavor, produce localized, almost spherical, light patterns
onthisparameter,weselectthedoubletsthataretheleastlikely
insidethedetector(seeAartsenetal.2013forexamples),which
tobechancecoincidencesofbackgroundevents(i.e.,therecon-
makesdirectionalreconstructionschallenging.Muonsproduced
structed directions of the two events are consistent within the
inν CCinteractions,ontheotherhand,cantraveluptoseveral
µ
errors,theyaredetectedwithinashorttimeandbotheventsare
kilometersintheiceandemitCherenkovlightalongtheirtrajec-
well localized). Follow-up observations are triggered automat-
tories.Theseeventsarecalledtracksandtheirsourcedirections
ically for doublets above a fixed significance threshold. Multi-
canbereconstructedtobetterthanonedegreeiftheirenergyis
pletsconsistingofmorethantwoeventsarerare(compareSect.
>1TeV(Aartsenetal.2017).Trackeventsoftenextendbeyond
3.2)andnoadditionalsignificancecutisapplied.
the detector volume which means that the detected energy is a
We use simulated neutrino events following an E−2.5 spec-
lowerlimitontheneutrinoenergy.Duetotheirsuperiorangular
trumtoquantifytheefficiencyofthemultipletselectionprocess.
resolution,trackeventsarepreferredforneutrinoastronomyand
Ifthreeneutrinosfromatransientsourcepasstheeventselection
thereal-timesystemonlyusesν CCevents.
µ
withinlessthan100s,atripletortwodoubletswithonecommon
eventaredetectedin79%ofthecases.Onedoubletwouldbede-
2.2. Real-timeEventSelection tected if one of the three events is separated by more than 3.5◦
from the two other events which happens with a probability of
IceCube has several real-time follow-up programs which se- 18%. There is a 3% chance that the reconstructed directions of
lect events and generate alerts in different ways (Aartsen et al. allthreeneutrinoswouldbeseparatedbymorethan3.5◦ andno
2016f). The neutrino alert described in this paper was found alertwouldbeissued.
bytheopticalfollow-upprogram(seealsoAbbasietal.2012b;
Evansetal.2015;Aartsenetal.2015b)whichsearchesforshort
transientneutrinosourcesandtriggersopticaltelescopesaswell 3. Thealert
astheSwiftX-raytelescope.
Twoneutrinodoublets,whichhaveoneeventincommon,were
Event selection starts from the online Muon Filter selec-
found on 2016-02-17 19:21:31.65 (detection time of the first
tionthatidentifieshigh-qualitymuontrackswitharateofabout
neutrino event, refered to as T0 in the following; all dates are
40Hz. This rate is dominated by muons produced in cosmic-
in UTC). All three events arrived within less than 100s. They
ray air showers. To increase the neutrino purity of the sam-
werenotautomaticallyidentifiedasatripletbecausethesecond
ple,moreadvancedandtime-consumingreconstructionsarere- andthirdeventswereseparatedby3.6◦,whileourcutisanan-
quired.SincecomputingpowerattheSouthPoleislimited,these gulardistanceof3.5◦.However,forconveniencewewillcallthe
reconstructionscanonlybeappliedtoasubsetofevents.Atthe
alertatripletinthefollowing.
SouthPole,theOnlineLevel2Filterusestheoutcomeofamax-
Neither doublet passed the required significance cut for in-
imumlikelihoodreconstructiontofurtherreducecontamination
dividualdoubletstobeautomaticallyforwardedtothePalomar
fromatmosphericmuons.Thisreconstructiontakesintoaccount
TransientFactory(PTF;Lawetal.2009;Rauetal.2009)orto
how photons propagate to the optical modules in the detector.
theSwiftsatellite(Gehrelsetal.2004).Moredetailsontheindi-
Selectioncriteriaare,forexample,thequalityofthelikelihood
vidualeventsaregiveninTable1andtheprojectionoftheevents
fit and the total number of modules which detected a photon.
ontheskyisshowninFig.1.
After application of these criteria, the event rate is reduced to ThecombinedaverageneutrinodirectionisRA=26.1◦ and
5Hz,whichislowenoughtoapplymoresophisticatedandtime- Dec =39.5◦ J2000 with a 50% error circle of 1.0◦ and a 90%
consuming reconstruction algorithms (see Aartsen et al. 2015b error circle of 3.6◦. This direction corresponds to the weighted
for a more detailed description). Based on the results of these
arithmeticmeanpositiontakingintoaccounttheangularuncer-
reconstructions,themostsignal-likeeventsareselectedusinga
tainties of the individualevents, σ. The error on the combined
multivariateclassifier(seeAartsenetal.2016fformoredetails direction is defined as σ =((cid:80)N σi −2)−1/2, where N =3 is the
ontheeventselectionanddatatransmission). w i=1 i
numberofevents.
Toavoidthebackgroundofatmosphericmuonsenteringthe
All quoted directions were obtained with the multi-
detector from above the follow-up program only uses events
photoelectron(MPE)fit(seeAhrensetal.2004)whichwasused
comingfrombelowandishenceonlysensitivetosourcesinthe
Northern sky. The final event rate is 3mHz and has a neutrino 1 WhileIceCubewasrunninginthe40and59stringconfigurationthe
purityof∼80%.Mostselectedneutrinocandidatesareproduced requiredangularseparationwas4◦(2008-12-16to2009-12-31).
Articlenumber,page3of23
SubmittedtoAstronomy&Astrophysics
Table1:DetailsonIceCubeevents
ID IceCubeEventID AlertID Time R.A. Dec. Error DepositedEnergy
(s) (◦) (◦) (◦) (TeV)
1 62474825 7,8 0 26.0[30.2] 39.9[43.2] 4.5[3.6] 0.26
2 62636100 7 +55.4 24.4[24.2] 37.8[38.4] 1.6[0.9] 1.1
3 62729180 8 +87.3 27.2[26.8] 40.7[40.7] 1.4[0.9] 0.52
Notes.Thedirectionsaretheresultofthereconstructionalgorithmthatwasusedinthefollow-upprogramatthetimeofthealert(MPEfit),while
thevaluesinbracketsresultfromanalternativereconstructionalgorithmwithanimprovedicemodel(SplineMPEfit).Theerroronthedirection
istheradiusofthe50%errorcircle.Thelastcolumnshowsanestimateoftheenergydepositedbythemuonsinthedetector,whichisalower
limitontheneutrinoenergy.Alltimesarerelativeto2016-02-1719:21:31.65UTC.
forthefollow-upprogramatthetimeofthealert.Animproved 48
versionofthisalgorithm,calledSplineMPE,usesamorereal-
istic model of light propagation in ice and on average reaches
a more precise reconstruction of the direction (Aartsen et al. 46
2014). The Spline MPE reconstruction has been used for the
follow-up program since May 2016. The Spline MPE fit yields
44
shifted coordinates which are shown in brackets in Table 1.
)
Thereconstructeddirectionchangesthemostforthefirstevent, g
e
whichdepositedlightinarelativelysmallnumberofDOMsdue d 42
(
to its low energy. Based on the Spline MPE fit the average di- n 3
rectionofallthreeeventsisRA=25.6◦,Dec=39.6◦ witherror o
circlesof0.7◦(50%)and1.9◦(90%). ati40 1
n
BasedontheSplineMPEreconstruction,events1and2(see cli
Table 1) would no longer form a doublet, while events 2 and e 38 2
D
3 would have formed a doublet. We expect the detection of 66
doubletsperyearduetobackground,andthe∼5mostsignificant
doubletsarefollowedup.Thedoubletconsistingofevents2and 36
3 does not pass the significance threshold (compare Sect. 2.3).
Hence,thealertwouldnothavebeenconsideredinterestingand
34
nofollow-upobservationswouldhavebeentriggeredevenifour
program had been running with the Spline MPE reconstruction
34 32 30 28 26 24 22 20
atthetimeofthealert.
Right Ascension (deg)
We used simulated neutrino events following an E−2.5 neu-
trino spectrum (compare Sect. 5.1) to calculate the probabil-
ity that three events from a point source form a triplet based Fig.1:Locationofthethreeneutrinocandidatesinthetripletwiththeir
50%errorcircles.Theplussignshowsthecombineddirectionandthe
on the MPE reconstruction, which is not recovered when using
shadedcircleisthecombined50%errorcircle.Thesolidcirclesshow
theSplineMPEalgorithm.Theresultingprobabilityis8%.For
theresultsoftheMPEreconstructionwhichisasthedefaultreconstruc-
background triplets (i.e., events that are aligned by chance but
tioninthefollowingandthethindashedcirclescorrespondtotheresults
do not stem from a point source) we evaluate scrambled data
oftheSplineMPEreconstruction(compareTable1).
(compare Sect. 3.2) and find that the probability is 36%. The
factthatthetripletisnotre-detectedwhenusingtheSplineMPE
algorithm is therefore a slight indication that it might not be of
astrophysical origin, but a coincidence of aligned background ingatleastfourpairsinclosetemporalandspatialcoincidence
events. whichtriggerwithin5µs.
All following analyses are based on the MPE position and Thesequantitiesaresensitivetodisturbancesinthedatatak-
error estimate which are shown in Fig. 1. Compared to the an- ingprocess(Aartsenetal.2016f).Thesedisturbancesareclassi-
gularseparationsbetweentheneutrinocandidatesthemeanpo- fiedaseitherinternal,suchasinterruptedconnectionstoaseg-
sitiononlychangesslightlyandthe50%errorcircleoftheMPE mentofthedetector,orexternal,suchasinterferencefromother
reconstruction fully contains the 50% error circle of the Spline experiments at the South Pole. Periods of bad operating condi-
MPEfit. tions can be flagged by monitoring the moving average of the
rates and comparing it to expected statistical fluctuations. This
systemhasoperatedforseveralyearsandhasreliablyidentified
3.1. DetectorStability
occasionalinternalandexternaldisturbancesduringthatperiod.
Before triggering follow-up observations we examined the sta- Nosignificantdeviationfromnormaldetectorbehaviorwasob-
tus of the detector carefully. A set of selected trigger and filter servedforatimeperiodspanningseveralhoursaroundtheevents
rates related to the analysis are monitored in real-time. Fig. 2 inthetriplet.
showstherateoftheSimpleMultiplicityTrigger,theMuonFil- In addition we generated test alerts which consisted of two
terandtheOnlineLevel2Filter(seeSect.2.2)nearthetimeof eventswithin100sthatareseparatedbymorethan3.5◦,butless
theevents.ASimpleMultiplicityconsistsofeightDOMsform- than7.5◦.Thetestalertratedidnotshowanyanomaliesaround
Articlenumber,page4of23
IceCubeetal.:Follow-upofaneutrinomultiplet
the follow-up program were improved yielding different sensi-
tivities. Moreover, we consider the down time of the follow-up
2350 program. Adding up the different contributions since 2008, the
total number of expected triplets from background was 0.38 at
2300
thearrivaltimeofthefirsttriplet.Theprobabilitytodetectone
2250
or more triplets from background is hence 32%. The detected
neutrinotripletmighthencebecausedbyachancealignmentof
40 backgroundevents.
z) 30 SMT
H
( MuonFilter 4. Follow-upobservations
te 20
a OnlineL2
R The neutrino triplet was not automatically forwarded to any
10 follow-upobservatorybecauseitdidnotpasstherequiredcrite-
riaforatriplet(alleventswithin3.5◦)andneitheroftheindivid-
0 ualdoubletsreachedtherequiredsignificancethresholdfortrig-
180 120 60 30 0 30 60
− − − − gering follow-up observations. However, as shown in Sect. 3.2
Minutesrelativetofirstevent
thecoincidenceoftwodoubletsisarareandinterestingeventso
Fig. 2: Temporal behavior of different filter rates: The Simple Mul- theIceCubeCollaborationdecidedtonotifythepartnersprovid-
tiplicity Trigger, Muon Filter and Online Level 2 rate. No significant ingelectromagneticfollow-upobservations.Ourfollow-uppart-
deviationfromnormaldetectorbehaviorwasobservedaroundthetime nerswereinformed22hoursafterthedetectionofthetriplet.In
ofthealert. caseofautomaticforwarding,themedianlatencyfortriggering
follow-upobservatoriesis∼1min.
The triplet direction was ∼70◦ from the Sun and difficult
toobservefromground-basedobservatoriessinceitwaslocated
the time of the alert. We hence conclude that the detector was
close to the horizon during night time and a large air mass im-
stablewhentheneutrinotripletwasdetected.
pairedtheimagequality.
Severalsourceclasseshavebeensuggestedaspotentialtran-
3.2. SignificanceCalculation sient neutrino sources. We therefore obtained multiwavelength
observations at different times after the neutrino detection. We
Toquantifythesignificanceoftheneutrinodetection,wecalcu- specifically search for GRBs, CCSNe (which might contain
latehowoftentripletsareexpectedfromchancecoincidencesof choked jets) and AGN flares. In this section we present re-
background events. We use the data obtained during the previ- portsontheobservationsobtainedwithoptical(Sect.4.1),X-ray
ous IceCube season from 2014-05-06 to 2015-05-18 when the (Sect.4.2)andgamma-ray(Sect.4.3)telescopes.Theresultsare
follow-upprogramwasrunninginthesameconfiguration.Con- summarizedandevaluatedinSect.5.
sideringonlythetimewhenthefollow-upprogramwasrunning
stably, the uptime of this season was 359days, during which
100,799 neutrino candidates passed the event selection of the 4.1. OpticalObservations
follow-upprogram.
Optical follow-up observations were obtained with ASAS-SN,
Toestimatethemultipletfalsepositiveratefromatmospheric
MASTER and LCO. No observations could be obtained with
backgrounds,werandomlyexchangedthedetectiontimesofall
thePTFP48telescopewhichwasundergoingengineeringwork.
events during this data-taking season. The event directions in
In addition to these follow-up observations, we also analyze
detector coordinates remained the same, but the equatorial co-
archivaldataobtainedwithin30daysbeforetheneutrinotriplet.
ordinates were recalculated using the newly assigned detection
time.Thismethodpreservesboththetemporalvariationsinthe
data(e.g.seasonalvariations;seeAbbasietal.2010b)anddirec- 4.1.1. ASAS-SN
tionaleffectscausedbythedetectorgeometry.Atthesametime
The All-Sky Automated Survey for SuperNovae (ASAS-SN or
anypotentialsignalfromatransientorsteadysourceissmeared
“Assassin”; Shappee et al. 2014) monitors the whole sky down
out.
to a limiting magnitude of V ∼17 mag. The focus of the sur-
To the generated background data we applied our a priori
veyistofindnearbysupernovae(SNe)andotherbrighttransient
cuts and searched for neutrino doublets (two events arriving
within 100s and with an angular separation of at most 3.5◦). sources.Currently,ASAS-SNconsistsoftwofullyroboticunits
with four telescopes each on Mount Haleakala in Hawaii and
We then counted how many doublets had at least one neutrino
Cerro Tololo in Chile. These eight telescopes allow ASAS-SN
event in common and found that such overlapping doublets or
tosurvey20,000deg2 pernight,coveringtheentirevisiblesky
tripletsareexpected0.0732±0.0009timesperfullyearoflive
everytwodays.Thepipelineisfullyautomaticanddiscoveries
time, hence one every 13.7years assuming the configuration in
areannouncedwithinhoursofthedatabeingtaken.Thedataare
whichtheprogramwasrunningatthetimeofthealert2.Theex-
photometrically calibrated using the AAVSO Photometric All-
pectednumberofbackgroundalertsiscalculatedforeverysea-
SkySurvey(APASS;Hendenetal.2015).
sonsincethestartofthefollow-upprograminDecember2008.
TheASAS-SN“Brutus”stationinHawaiihasregularlyob-
Withinthistimeboththeeventselectionandalertgenerationof
served the field containing the triplet position since 2013-10-
2 Weemphasizethatourdefinitionofatripletonlyrequiresthatoneof 27,obtaining408ninety-secondV-bandimageson178separate
thethreeeventsformsadoubletwiththetwootherones.Thetwoother nights. Before the neutrino trigger, this field was last observed
eventscanthereforebeseparatedbymorethan3.5◦anddonothaveto two weeks earlier, on 2016-02-03, as the observability of this
arrivewithin100s. field was limited due to the Sun angle. In Table B.1 we list the
Articlenumber,page5of23
SubmittedtoAstronomy&Astrophysics
Table2:ObservingconditionsattheMASTERtelescopesatthetime2016-02-1817:15:58UTC
MASTERnode Objectaltitude Sunaltitude Notes
(◦) (◦)
MASTER-Amur 3.98 −47.01 tooclosetothehorizonforgoodobservations
MASTER-Tunka 13.45 −49.91 cloudyandtooclosetothehorizonforgoodobservations
MASTER-Ural 37.06 −33.25 badweather
MASTER-Kislovodsk 43.50 −28.31 goodconditions,observationsbegan
MASTER-SAAO −8.36 0.93 belowthehorizonatnighttime
MASTER-IAC 78.22 20.25 snowstorm
MASTER-OAFA −1.1 69.06 belowthehorizonatnighttime
dates on which this field was observed during the 30 days be- unfiltered).TocollectasmanyphotonsaspossibletheMASTER
fore the trigger, and also the typical 5σ V band detection limit telescopes are usually operated without a filter when searching
reached,inthe3×90-secditheredexposures.Theresultinglim- fortransients.Inadditioneachobservatoryhostsverywide-field
itsareshowninSect.5.2. cameras which cover 400 square degrees and are sensitive to
Followingtheneutrinotrigger,wescheduled20×90-secex- sourcesbrighterthan15thmagnitude.
posures of the field containing the trigger position, which were
AnimportantcomponentofMASTERisitsin-housedetec-
taken between UTC 2016-02-19.229 and 2016-02-19.253. The
tionsoftwarewhichprovidesphotometricandastrometricinfor-
ASAS-SN field contains about 90% of the final 50% error cir-
mationaboutallopticalsourcesintheimagewithin1-2minutes
cle of 1◦. Because of the bright Moon, the combined depth of
aftertheframereadout.Theprocessingtimeincludesprimaryre-
V (cid:46)18.0 is relatively shallow while the 5σ depth of the indi-
duction(bias,dark,flatfield),sourceextractionwithhelpofthe
vidual 90-secexposures is V (cid:46)16.5. Notransient sources were
SExtractor algorithm4 (Bertin & Arnouts 1996), the identifica-
detected.
tion of cataloged objects and the selection of unknown objects.
New sources detected in two images at the same position are
4.1.2. LCO classified as optical transients (Lipunov et al. 2016). The unfil-
tered magnitudes are calibrated using stars from the USNO-B1
TheLasCumbresObservatory(LCO3;Brownetal.2013)con-
catalogwherethecatalogmagnitudesareconvertedtounfiltered
sists of seven 0.4-m, nine 1-m and two 2-m robotic telescopes magnitudesvia0.2×B+0.8×R.Foreachimagealimitingmag-
situated in six sites around the world (two additional 1-m tele-
nitudeiscalculated.
scopeswillbedeployedinthenearfuturetoaseventhsite).The
network specializes in time domain astronomy, and has the ca- TheMASTERnetworkreceivedtheneutrinotripletcoordi-
pabilityofperformingimmediatetarget-of-opportunityobserva- nates by email at 2016-02-18 17:15:58 UTC. The altitudes and
tionsofalmostanypointintheskywithinminutes. visibility constraints of the position at the different observato-
The error circle of the neutrino triplet was tiled with nine ries are listed in Table 2 for the time when the neutrino detec-
pointings that were observed with the LCO 1-m Telescope at tion was communicated. Observations started at the MASTER-
McDonaldObservatoryinTexas.Theobservationscoverthein- Kislovodsktelescopeswithinlessthanonehourandtheposition
ner ∼60% of the 50% error circle of the final triplet location. wasmonitoredbyMASTER-Kislovodsk,MASTER-Tunkaand
ObservationswereobtainedusingvariouscombinationsofUB- MASTER-IACforthefollowingmonth(compareTableB.1).
Vgri filters on different nights (Table B.2 and Sect. 5.2). The
ThemajorityoftheobservationslistedinTableB.1arecen-
limitingmagnitudeswerecalculatedfollowingcalibrationtothe
teredonthetripletpositionandincludethecomplete50%error
APASScatalog(seeAppendixBofValentietal.2016formore
circleofthefinalposition.Moreover,exceptforsmallgaps,the
details). Due to the proximity of the field to the sun, additional
complete90%errorcirclewascoveredbothbeforeandafterthe
epochs could not be obtained in the weeks following the alert
neutrinodetection.Notransientswerefoundabovethe5σlim-
to determine whether any transient sources were present in the
itingmagnitudesgiveninTableB.1andshowninSect.5.2.The
images.
verywidefieldcamerasdidnotdetectanytransientbrighterthan
15th magnitude within the 400 square degrees surrounding the
tripletlocation.
4.1.3. MASTER
The Mobile Astronomical System of the Telescope-Robots
(MASTER; Lipunov et al. 2010; Kornilov et al. 2012; Gor- 4.2. X-rayobservations
bovskoyetal.2013)GlobalRoboticNetconsistsofsevenobser-
vatoriesinbothhemispheres(seeTable2).AllMASTERobser- We triggered the X-ray Telescope (XRT) on board the Swift
vatoriesincludeidenticaltwin40-cmwide-fieldtelescopeswith satellite(Gehrelsetal.2004)tosearchforGRBafterglows,AGN
two4squaredegreefieldsofviewwhichmonitortheskydown flares or other X-ray transients (see Sect. 4.2.2). By chance the
to 21st magnitude. In divergent mode, the twin telescopes can Swift Burst Alert Telescope (BAT; Barthelmy et al. 2005) ob-
cover 8 square degrees per exposure and the telescope mounts servedthetripletpositionwithinaminuteaftertheneutrinode-
allow rapid pointing to follow up short transient sources. Each tectionasdescribedinSect.4.2.1.
MASTER node is equipped with BVRI Johnson/Bessel filters,
two orthogonal polarization filters and two white filters (called
3 http://lco.global 4 http://www.astromatic.net/software/sextractor
Articlenumber,page6of23
IceCubeetal.:Follow-upofaneutrinomultiplet
4.2.1. SwiftBATobservations
40.5 720
Swift BAT detects hard X-rays in the energy range from 15– X2
150keV.Thefieldofviewcoversabout10%oftheskyandthe 640
detectorisilluminatedthroughapartiallycoded-aperturemask.
40.0 560
Just30safterthelastneutrinowasdetected,theSwiftsatel- s)
litecompletedapreplannedslewtoRA=23.38◦,Dec=+41.12◦ 00) X6 480me (
wathaicpharptliaaclecdodthinegtrfirpalcettiopnosoiftio6n0%w.itWhienrtehteriBevAeTd tfiheeldBoAfTvdieawta, c (J2039.5 X3 X5 400ure ti
for this pointing from the Swift Quick Look website (ObsID De X4 X1 320pos
00085146016). No rate- or image-triggered transients were de- x
tected above the significance threshold of S>6.5σ during the 39.0 240e
pointing, so only survey mode data are available. Survey data 160
for the pointing consist of three exposures of 59s, 10s, and
15s,withinterveninggapsformaintenanceoperations.TheBAT 38.5 80
analysis was conducted using the heasoft5 (v. 6.18) software 0
27.5 27.0 26.5 26.0 25.5 25.0 24.5
toolsandcalibration,closelyfollowingtheanalysesfromMark-
RA (J2000)
wardtetal.(2005);Tuelleretal.(2008,2010)andBaumgartner
etal.(2013).
Fig. 3: Exposure map of the 37 Swift XRT pointings averaging
We used the heasoft tool batcelldetect on the summed ex-
320spertiling.Theredcircleshowsthe50%confidencebound
posure as well as on the first exposure over the full bandpass
tothetripletposition.XRTsources(compareTable3)areshown
(15−150keV), with a detection threshold of S = 3.5σ (the
asblackpoints.
lowest allowed setting). The most significant detection within
the triplet 90% confidence region was in the first exposure at
RA=28.6083◦,Dec=37.34583◦ (henceforthreferredtoasthe
mapshowninFig.3.Theachievedexposureperpointingis0.3–
BATBlip)withsingle-trialsignificanceS=4.6σ.
0.4ks. Data were analyzed as described in Evans et al. (2015),
ToestimatethesignificanceoftheBATBlipgiventhesearch leadingto asingle unifiedX-ray image,exposure map,and list
area,wefindthenumberofsimilarormoresignificantfluctua- of X-ray sources. The Swift XRT observations cover nearly the
tions in a rectangular region of the BAT image plane centered complete50%containmentregion.
around the position of BAT Blip in 2655 BAT pointings with
Six X-ray sources were identified (Table 3) with the detec-
similar exposure times. We find an average of 0.13 such candi- tionflaggoodwhichmeansthattheirprobabilityofbeingspuri-
date sources per pointing. Since the triplet 90%-confidence re- ousis<0.3%(Evansetal.2015).Asrevealedfromsearchesof
gioncorrespondsto41%oftherectangularregion,thisyieldsa theNASAExtragalacticDatabase7 andexaminationofarchival
p-value of p=9.9% for the BAT Blip. A trial factor penalty of
opticalimages,X1isspaciallycoincidentwithaknownSeyfert1
twowasincludedsinceboththesummedandthefirstexposure
galaxy;X2,X3,X4,andX5correspondtostarsandX6remains
wereanalyzed.TheBATBlipishenceconsistentwitharandom
unidentified. We note that X-rays associated with a bright star
fluctuationofthebackground. were detected when Swift followed up a neutrino candidate de-
Flux upper limits were derived from the summed expo- tectedbytheANTARESneutrinotelescope(Dornicetal.2015;
sure noise map, including the BAT Blip, over the triplet 90%- Smartt et al. 2015). The large number of stars detected in our
confidence region, and we find a 4σ upper limit to the fluence observationsshowsthatsuchchancecoincidencesarefrequent.
of3.3×10−7ergcm−2 fortheenergyrangeof15–150keV.This Wedonotconsiderthestarsaspotentialsourcesofhigh-energy
corresponds to a limit of 3.9×10−9ergcm−2s−1 on the aver- neutrinos.
age flux between 100s to 256s after the detection of the first TheX-raysourceX1isclassifiedasaFlatSpectrumRadio
neutrino. BAT count limits are converted to fluences using the Quasar by Healey et al. (2007) and is located at a redshift of
PIMMS6 online tool, assuming a power law with a spectral in- z=0.08 (Wills & Browne 1986). It has been detected several
dexofΓ=−2.ThisspectralindexcorrespondstoatypicalGRB timesbyROSAT,XMM-NewtonandtheSwift XRT.Compared
spectrum in this energy range. It is moreover very close to the tothepreviousdetections,X1wasnotflaringduringtheseXRT
mean AGN spectral index which was measured to be −1.95 by observations.
Burlonetal.(2011).InSect.5.3wecomparethelimittotypical
Among the identified sources, X6 is unique in not having
promptfluxesofGRBsdetectedbytheBAT.
anobviouscounterpartwithinits90%errorcircle.Torefinethe
localization and study the X-ray variability, X6 was followed
up with 1ks and 8.6ks Swift observations on 2016-03-18 and
4.2.2. SwiftXRTobservations
2016-07-23 (Target ID 34429). The source was re-detected in
TheSwiftXRTisanX-rayimagingspectrometersensitivetothe thedeepestXRTobservation;itfadedbyafactorofninewithin
energyrangefrom0.3−10keV.Thetelescope’sfieldofviewhas fivemonths.TheXRTlightcurve,showninFig.A.1,isconsis-
adiameterof0.4◦.TosearchforpossibleX-raycounterpartsto tentwithat−0.5 decayoverfivemonthswhichistooshallowfor
theneutrinotripletoverthelargestfeasibleregion,werequested aGRBafterglow(seeSect.5.3)oratypicaltidaldisruptionevent
a 37-pointing mosaic of Swift observations. These observations whichfadeswitht−5/3intheX-rayregime(Komossa2015).The
began at 2016-02-18 17:57:42 (22.6h after the neutrino detec- latterdetectionrulesoutthepossibilitythatX6isaGRB.
tion; Target IDs 34342 to 34379), with the resulting exposure In archival PTF images we find two bright stars, hereafter
referred to as S1 and S2, located close to the 90% error circle
5 heasoftwebsite:http://heasarc.nasa.gov/lheasoft/
6 available at https://heasarc.gsfc.nasa.gov/docs/ 7 NASA Extragalactic Database: https://ned.ipac.caltech.
software/tools/pimms.html edu/
Articlenumber,page7of23
SubmittedtoAstronomy&Astrophysics
Table3:XRTsources
Name R.A. Dec. ExposureTime Rate Alt.Name ObjectType
(s) (counts/s)
X1 25.4909 +39.3921 308 0.097±0.020 B20138+39B Seyfert1Galaxy
X2 25.6546 +40.3788 285 0.047±0.015 HD10438 Star
X3 25.5324 +39.4129 324 0.035±0.012 V*OQAnd VariableStar
X4 26.7475 +39.2575 284 0.024±0.011 1RXSJ-14658.4+391526 Star
X5 25.0723 +39.5886 221 0.029±0.014 HD10169 Star
X6 25.0107 +39.6033 506 0.017±0.007 - unknown
Notes.CoordinatesareprovidedinJ2000.
of X6. To look for fainter optical sources we obtained a Keck 4.3.1. TheFermiLAT
imageinwhichathirdobject,O3,isdetected(seeFig.A.2).The
propertiesofthethreepotentialopticalcounterpartsarespecified
inTableA.1.
The Fermi Gamma-ray Space Telescope consists of two
To search for short lived optical emission we analyze si- primary instruments, the Large Area Telescope (LAT) and
multaneous UVOT observations. During the first XRT obser- the Gamma-Ray Burst monitor (GBM). The LAT is a pair-
vation the UVOT observed in the U band (Target ID 34357). conversion telescope comprising a 4×4 array of silicon strip
Weusetheheasofttooluvotdetecttomeasuretheapertureflux trackers and cesium iodide (CsI) calorimeters. The LAT covers
within a circle with a radius of 3(cid:48)(cid:48)centered around the best fit the energy range from 20MeV to more than 300GeV with a
location of X6. This small radius was choosen to avoid con- FoVof∼2.4steradian,observingtheentireskyeverytwoorbits
tamination from the star S2. No source is detected and the 3σ (∼3hours) while in normal survey mode (Atwood et al. 2009).
limitis17.39magAB whichcorrespondstoafluxupperlimitof TheGBMiscomprisedof12sodiumiodide(NaI)andtwobis-
10−15erg s−1cm−2Å−1atawavelengthof3501Å. muth germanate (BGO) scintillation detectors that have an in-
stantaneousviewof70%ofthesky.TheNaIandBGOdetectors
Considering all available observations we identify two pos-
aresensitivetoemissionbetween8keVand1MeV,and150keV
siblescenarios:X6couldeitherbeanextremestellarflareorit
and40MeV,respectively(Meeganetal.2009).
couldbeanobscuredanddistantAGN.Wediscussthenatureof
The triplet location was occulted by the Earth at the detec-
X6inmoredetailinAppendixA,wherewecometotheconclu-
tion time of the first neutrino event (T0). As a result, the GBM
sionthatitlikelyisnotassociatedwiththeneutrinotriplet.
and LAT can place no constraints on the existence of a prompt
Except for X-ray source X6, the Swift follow-up observa- gamma-ray transient coincident with the detection of the neu-
tions identified no unknown X-ray sources within the 50%- trinoevents.Withinthe24hoursbeforeandafterT0,therewere
containment region of the neutrino triplet. Our upper limits on a total of four reported GBM detections8. They were all sepa-
anysourceoverthisregionarederivedfromthe0.3–1.0keV,1– ratedbymorethan50◦ fromthetripletlocationandanassocia-
2keV, 2–10keV, and 0.3–10keV (full band) background maps. tioncanbeexcluded.
Backgroundcountratesforeachbandpasswereestimatedfrom The region of interest entered the LAT field-of-view after
three regions, sampling the on-axis, off-axis, and field-overlap roughly 1600s and in the following we analyze the LAT data
portionsofthetotalexposurepattern;theseprovidea3σcount- recorded within the days before and after the detection of the
rate upper limit following the Bayesian method of Kraft et al. neutrino alert. We focussed on limiting the intermediate (hours
(1991).Theupperlimitswerethenmultipliedbyafactorof1.08 todays)tolong(weeks)timescaleemissionfromanewtransient
to correct for the finite size of the aperture (a 20-pixel radius). sourceorflaringactivityfromaknowngamma-rayemitterinthe
The rate upper limits are converted to fluxes for each of two LATenergyrange.
spectral models: a typical AGN spectrum in the X-ray band (a We employed two different techniques to search for such
powerlawwithphotonindexΓ=−1.7,NH=3×1020cm−2)and emission in the LAT data; the Fermi All-sky Variability analy-
aGRBspectrum(apowerlawwithΓ=−2,NH=3×1021cm−2). sis (FAVA; Ackermann et al. 2013a) and a standard unbinned
TherangeofresultingupperlimitsislistedinTableB.3.InSect. likelihood analysis. FAVA is an all-sky photometric analysis in
5.3wecomparethelimitstodetectedGRBafterglows. whicharegionoftheskyissearchedfordeviationsfromtheex-
pected flux based on the mission-averaged data. The unbinned
likelihoodanalysisisthestandardmethodofdetectingandchar-
4.3. Gamma-rayobservations acterizingsourcesintheLATdataandisdescribedinmoredetail
inAbdoetal.(2009).Weadditionallyemployedaprofilelikeli-
ThepositionofthetripletwasobservedbytheFermiLATabout hoodmethoddescribedinAckermannetal.(2012)tocalculate
30minutes after the neutrino detection (see Sect. 4.3.1). Bad upperlimitsinsituationswhennosignificantexcessemissionis
weather conditions in La Palma did not allow immediate ob- detected.
servations with either MAGIC (Aleksic´ et al. 2016) or FACT The FAVA search was performed on 24h timescales brack-
(Anderhub et al. 2013) and the position is not observable for eting T0, covering the periods of [T0−24h to T0], [T0−12h
H.E.S.S.. VERITAS observed the direction with a delay of one to T0+12h], and [T0 to T0+24h] (see Table 4). A single
week(seeSect.4.3.2)andthepositionwaswithinHAWC’sfield
ofviewatthearrivaltimeofthetriplet(seeSect.4.3.3). 8 http://gcn.gsfc.nasa.gov/fermi_grbs.html
Articlenumber,page8of23
IceCubeetal.:Follow-upofaneutrinomultiplet
+44° 25.0
22.5
3FGL J0133.3+4324
20.0
+42°
3FGL J0202.5+4206
17.5
0)+40° 15.0
0
0
ec (J2 3FGL J0156.3+3913 3FGL J0136.5+3905 12.5TS
D 10.0
+38°
7.5
3FGL J0152.2+3707
5.0
+36°
2.5
0.0
32° 30° 28° 26° 24° 22°
RA (J2000)
(a)TheFermiLATlikelihoodratioteststatistic(TS)withinthere- (b) Fermi LAT 95% upper limits on the flux in the 100MeV to
gionofinterest.Thesignificanceofflucuationsabovetheexpected 100GeVenergyrange.
√
backgroundscalesroughlywith TS.
Fig.4:FermiLATresultsfromtheunbinnedlikelihoodanalysiswithintheregionofinterestusingalldatawithin14daysafterthe
neutrinodetection.Thedashedcirclesshowthe50%and90%errorcirclesoftheneutrinotriplet.
Table4:FermiLATfluxupperlimits
Interval Duration StartDate EndDate MedianU.L.(95) MedianU.L.(95)
(UTC) (UTC) (phcm−2s−1) (GeVcm−2s−1)
T 24hrs 2016-02-1719:21:32 2016-02-1819:21:32 – –
FAVA1
T 24hrs 2016-02-1619:21:32 2016-02-1719:21:32 – –
FAVA2
T 24hrs 2016-02-1707:21:32 2016-02-1807:21:32 – –
FAVA3
T 7days 2016-02-1515:43:35 2016-02-2215:43:35 – –
FAVA4
T 6hrs 2016-02-1719:21:32 2016-02-1801:21:32 3.32×10−7 1.82×10−7
Like1
T 12hrs 2016-02-1719:21:32 2016-02-1807:21:32 1.86×10−7 1.01×10−7
Like2
T 24hrs 2016-02-1719:21:32 2016-02-1819:21:32 1.27×10−7 6.96×10−8
Like3
T 24hrs 2016-02-1619:21:32 2016-02-1719:21:32 1.15×10−7 6.30×10−8
Like4
T 24hrs 2016-02-1707:21:32 2016-02-1807:21:32 1.11×10−7 6.08×10−8
Like5
T 14days 2016-02-1719:21:32 2016-03-0219:21:32 1.73×10−8 9.48×10−9
Like6
Notes.AsummaryoftheFAVAandlikelihoodanalysistimescales.FAVAdoesnotprovidefluxupperlimitestimated.Theupperlimitestimates
quotedforthelikelihoodanalysisarethemedian95%C.L.consideringallupperlimitswithinthe90%errorcircle.Theyhavebeenobtainedfor
theenergyrangefrom100MeVto100GeVandaspectralindexofΓ=−2.1hasbeenassumed.
week-long timescale was also searched, covering the period The unbinned likelihood analysis was performed using the
of [T0−2.15days to T0+4.85days]. The FAVA analysis selects standardLATanalysistools(ScienceToolsversionv10r01p0)10,
flaresthathaveasignificanceof6σabovethemissionaverage bymodelingallphotonswithinaregionofinterest(ROI)witha
emission at the location. Within the analyzed time windows no radiusof12◦,coveringanenergyrangeof100MeVto100GeV,
suchflarewasdetectedatthetripletlocation. andutilizingtheP8R2_TRANSIENTR020_V6eventclassandthe
AnexaminationofthesecondFAVAcatalog(2FAV,paperin corresponding instrument response functions. For the purposes
preparation),whichlistsallflaringsourcesdetectedintheLAT of this analysis, all modeled sources were fixed to their cata-
data on weekly timescales over the course of the entire mis- log values, while the normalization of the Galactic and diffuse
sion, shows only one period of flaring activity within the 90% isotropiccomponentsofthefitwereallowedtovary.Becauseof
error circle of the triplet location9. This period of activity was theuncertaintyinthetripletlocation,thisanalysiswasrepeated
between 2009-08-31 and 2009-09-07 and was associated with overa10◦×10◦gridofcoordinateswith0.15◦binning.
3FGLJ0156.3+3913 which is a blazar candidate of uncertain Thissearchwasperformedoveravarietyoftimescales,rang-
type (Acero et al. 2015). No further activity from this source ing from 6h to 14days (Table 4). The resulting significance
hasbeendetectedbyFAVA. maps show no emission in excess of the expected background
onanyofthetimescalesconsidered.Foreachbininthecoordi-
9 http://fermi.gsfc.nasa.gov/ssc/data/access/lat/FAVA/
LightCurve.php?ra=26.1&dec=39.5 10 http://fermi.gsfc.nasa.gov/ssc/
Articlenumber,page9of23
SubmittedtoAstronomy&Astrophysics
nategrid,wecalculatedthe95%confidencelevels(C.L.)upper
limitonthephotonfluxofacandidatepointsourcewithafixed
spectral index of Γ=−2.1. This value is appropriate for both
AGN(compareAckermannetal.2015)andGRBs(Gruberetal.
2014;Ackermannetal.2013b)andisusedasthestandardvalue
when searching for GRBs. An example of the significance and
energy upper limit maps for the T0+14day timescale is shown
in Fig. 4. The median photon flux and energy flux upper limits
calculatedforeachtimescalearelistedinTable4.
4.3.2. VERITAS
VERITAS is a ground-based instrument for very-high-energy
(VHE) gamma-ray astronomy with maximum sensitivity in the
80GeV to 30TeV range. It is located at the Fred Lawrence
WhippleObservatory(FLWO)insouthernArizona(31◦ 40’N,
110◦ 57’ W) at an altitude of 1.3km above sea level. The ar-
rayconsistsoffour12-m-diameterimagingairCherenkovtele-
scopeseachequippedwithacameracontaining499photomulti-
pliertubes(PMTs)coveringa3.5◦ fieldofview.Fulldetailsof
theVERITASinstrumentperformanceandsensitivityaregiven
Fig. 5: Significance sky map for the VERITAS observations of
inPark(2015).
the neutrino triplet region. The dashed white (gray) line indi-
AtthetimethetripletdetectionwascommunicatedtoVER-
cates the 50% (90%) error circle for the triplet. No gamma-ray
ITAS, the Moon was approaching its full phase and the night
excesswasdetectedinthefieldofview.TheknownVHEsource
skywastoobrighttosafelyoperatethesensitivePMTcameras. RGBJ0136+391 is located approximately 1.6◦ away from the
It is, however, not uncommon for some variable VHE sources
tripletcentralposition.
such as active galactic nuclei to exhibit extended periods of in-
tense flaring activity that can be detected days after the source
has reached its peak flux (Dermer & Giebels 2016). Observa-
200,000 litres of purified water and instrumented with four
tions were started eight days after the detection of the neutrino
photo-multipliertubes.Light-tightbladdersprovideopticaliso-
events on 2016-02-25, when VERITAS observed the triplet lo-
lation. The observatory is optimized to detect Cherenkov light
cation between 02:32 and 03:20 UTC. Additional observations
from extensive air showers produced by gamma-ray primaries
weretakenon2016-02-26between02:36and03:43UTC.The
atenergiesbetween100GeVand100TeV.HAWCislocatedin
combinedexposuretimeduringthesetwonightswas62.8min,
the state of Puebla, Mexico at an altitude of 4,100m (97.3◦W,
afterqualitycutswereapplied.Theseobservationswerecarried
19.0◦N).HAWCoperatescontinuouslyandhasanaveragedown
outinthenormal‘wobble’mode,wherethepointingdirectionof
thetelescopesisoffsetfromthesourcepositiontoallowforsi- timeduetomaintenanceofonly∼5%.Awidefieldofview,ap-
proximatelydefinedbyaconewithanopeningangleof45◦from
multaneousmeasurementofthebackground(Bergeetal.2007).
Awobbleoffsetof0.7◦ wasselectedtocoveralargerregionof zenith,spansthedeclinationrangeof−26◦ to+64◦ androtates
with the Earth through the full range of right ascension every
skygiventheuncertaintyintheaveragedtripletposition.
day.Foradetaileddescriptionofthearrayandanalysismethods
An analysis of the VERITAS data showed no significant
seeAbeysekaraetal.(2017).
gamma-ray excess in the triplet region of interest (see Fig. 5).
Consequently, differential flux upper limits were calculated at Atthedetectiontimeoftheneutrinotriplet,itspositionhad
the 95% confidence level in four energy bins for a gamma-ray just entered HAWC’s field of view. HAWC was operating nor-
pointsourcelocatedattheaveragedtripletpositionandaregiven mally and observed the full transit (∼6hours at zenith angles
inTableB.4.Furthermore,nonewgamma-raysourceswerede- <45◦) of the triplet location between 19:15 UTC on 2016-02-
tectedanywherewithinthetriplet50%errorregionorwithinthe 17 and 01:30 UTC on 2016-02-18. HAWC data are being con-
VERITASfieldofview. tinuously reconstructed on computers at the array site with an
The only known VHE source in the vicinity of the triplet average time lag of approximately 4s and were immediately
is the high-synchrotron-peaked blazar RGB J0136+39111 (also available for a follow-up analysis when the IceCube alert was
3FGLJ0136.5+3905; see Fig. 4a). It has an approximate angu- received.
lardistanceof1.6◦ fromthetripletcentralpositionandwasnot Ascanoftheregionaroundthetripletcoordinateswasper-
detectedduringtheVERITASobservations(seeSect.5.4forfur- formed with the standard HAWC maximum-likelihood tech-
ther discussion of this source). Therefore, the data show no in- nique, using nine energy-proxy analysis bins that sort data ac-
dicationofapersistentVHEgamma-raysource,orahighstate cordingtotheairshowersize(Abeysekaraetal.2016).Theanal-
ofRGBJ0136+391,whichcouldbeassociatedwiththeneutrino ysis bins account for the varying angular resolution and back-
events. ground suppression efficiency. For each bin, the event count in
eachpixelofaHEALPix(Górskietal.2005)mapiscomparedto
apredictioncomposedoftheaverage,smoothedbackgroundof
4.3.3. HAWC
cosmic rays measured from data and the simulated expectation
The High Altitude Water Cherenkov (HAWC) Observatory of gamma-ray events from a point-like source. The signal ex-
is an array of 300 detectors, each filled with approximately pectationincludesthemodelingoftheangularresolution,which
improveswithenergyfrom∼1◦ to<0.2◦ intherangefrom1to
11 http://tevcat.uchicago.edu/?mode=1;id=244 100TeV. The differential flux in each analysis bin is described
Articlenumber,page10of23