γ
Fe ν
p Going further with high-energy multi-messenger astronomy
Kumiko Kotera - Institut d’Astrophysique de Paris
EuCAPT Symposium - 07/05/2021 Multi-messengers!
Gravitational waves
Cosmic rays
p Fe ν Gamma rays ν
Neutrinos
2 Multi-messengers! cosmic rays + others —> temporal coincidence impossible (deflections) but studies of di!use fluxes
Gravitational waves GW + electromagnetic (+gamma) GW170817
neutrinos + gammas? TXS0506+056 (blazar flare) Cosmic rays
p Fe ν Gamma rays ν
Neutrinos Neutrinos
3 RESEARCH | RESEARCH ARTICLE
as a fitted parameter. The model parameters are the RA coordinates within each data-taking period. contributing events at boundary times). For the correlated and are expressed as a pair, (F100, g), The resultant P value is defined as the fraction of box-shaped time window, the uncertainties are where F100 is the flux normalization at 100 TeV. randomized trials yielding a value of TS greater than discontinuous and not well defined, but the un- The time-dependent analysis uses the same for- or equal to the one obtained for the actual data. certainties for the Gaussian window show that it mulation of the likelihood but searches for Because the detector configuration and event is consistent with the box-shaped time window Neutrino flares andclustering TXS in time as well0506+56 as space by introducing selections changed as shown in Table 1, the time- fit. Despite the different window shapes, which an additional time profile. It is performed sep- dependent analysis is performed by operating on lead to different weightings of the events as a IC-170922A – a 290arately TeV for two different Neutrino generic profile shapes: a each data-taking period separately. (A flare that function of time, both windows identify the same Gaussian-shaped time window and a box-shaped spans a boundaryRESEARCH between two periods could be time interval as significant. For the box-shaped time window. Each analysis varies the central partially detected in either period, but with re- time window, the best-fitting parameters are sim- ◥ trinos, IceCube provides real-time triggers for time of the window, T0,andthedurationTW duced significance.) An additional look-elsewhere ilar to those of the Gaussian window, with fluence RESEARCH ARTICLE SUMMARY observatories2 around the world measuring (from seconds to years) of the potential signal to correction then needs to be applied for a result in at 100 TeV and spectral index giveng-rays, by x-rays,E J100 optical,= radio, and gravitational T T 1:0 4 –2 waves, allowing for the potential identification find the four parameters (F100, g, 0, W)that an individual dataNEUTRINO segment, ASTROPHYSICS given by the ratio of 2:2þ0:8 10À TeV cm and g =2.2±0.2.This maximize the likelihood ratio, which is defined the total 9.5-year observation time to the obser- fluenceÀ  corresponds to an averageof even flux rapidly over fading sources. 0:7 15 –1 –2 –1 as the test statistic TS. (For the Gaussian time vation time of that data segment (30). 158 days of F100 =1:6þ0:6 10À TeVRESU LTS:cm A high-energys . neutrino-induced Multimessenger observations ofÀ a  muon track was detected on 22 September 2017, window, TW represents twice the standard de- When we estimate the significance of the time- Neutrinos from the direction of automatically generating an alert that was viation.) The test statistic includes a factor that flaring blazar coincidentdependent with result by performing the analysis at distributed worldwide ◥ corrects for the look-elsewhere effect arising TXS 0506+056 the coordinates of TXS 0506+056 onON randomized OUR WEBSITE within 1 min of detection and prompted follow-up from all of the possible time windows that could The results of thehigh-energy time-dependent analysis neutrino per- datasets, IceCube-170922A we allow in each trial a newRead the fit full for article all at http://dx.doi. searches by telescopes over RESEARCH ARTICLE RESEARCH | be chosen (30). formed at the coordinatesThe IceCube of Collaboration, TXS 0506+056Fermi-LAT, are MAGIC,theAGILE parameters:, ASAS-SN, HAWC,F100, g H.E.S.S.,, T0, TW. Weorg/10.1126/ find that the a broad range of wave- INTEGRAL, K anata, K iso, K apteyn, Liverpool Telescope, Subaru, Swift/NuSTAR, science.aat1378 lengths. On 28 September For each analysis method (time-integrated and shown in Fig. 1 for each of the six data periods. fraction of randomized trials that result...... in a more 2017, the Fermi Large Area VERITAS, and VLA/17B-403 teams*† –5 Downloaded from lower limit of 183 TeV, dependingtime-dependent), only weakly on a robusttrophysical significance spectrum, together estimate with is simulatedOne ofcannot the data be excluded. periods, Elect IC86bromagnetic from observations 2012 to 2015, significant excess than the real dataTelescope is 7 × 10 Collaborationfor reported that the di- the assumed astrophysicalobtained energy spectrum by performing (25). data, was the used identical to calculate analysis the probability on thatcontains a are avaluable significant to assess excess, the possible which association is identified of the box-shaped time window and 3 rection× 10–5 offor the neutrino the was coincident with a The vast majority of neutrinos detected by neutrino at the observed track energy and zenith a single neutrinoINTRO to an DU astrophysical CTIO N: Neutrinos source. are tracers of mic rays. The discovery of an extraterrestrial cataloged g-ray source, 0.1° from the neutrino IceCube arise from cosmic-raytrials interactions with randomized within angle datasets. in IceCube These is of are astrophysical produced origin.by This both time-windowFollowing thecosmic-ray alert, shapes. IceCube acceleration: The performed excess electrically consists a neutral Gaussiandiffuse flux of high-energy time window. neutrinos, Thisannounced fraction,direction. once The cor- source, a blazar known as TXS Downloaded from Earth’s atmosphere. Althoughby randomizingatmospheric neu- theprobability, event times the so-called and recalculating signalness of the eventof 13 ±complete 5 events analysis aboveand traveling of the relevant expectation at nearly data the prior speed from ofto light, the they rectedby IceCube for in the 2013, ratio has characteristic of the total prop- observation0506+056 at time a measured redshift of 0.34, was trinos are dominant at energies below 100 TeV, (14), was reported to be 56.5% (17). Although 31 October 2017.can Although escape the no densest additional environments excess and may erties that hint at contributions from extra- in a flaring state at the time with enhanced atmospheric background.be tracedThe back to significance their source of depends origin. High- togalactic the sources, IC86b although observation the individual time sources (9.5 years/3g-ray activity years), in the GeV range. Follow-up ob- their spectrum falls steeply with energy, allowing IceCube can robustly identify astrophysical neu- of neutrinos was found from the direction of TXS –4 –4 astrophysical neutrinos to be moreTable easily 1. identi-IceCubetrinos neutrino at PeV data energies, samples. for individual neutrinoson the0506+056 energies near ofenergy the the time events, neutrinos of the theirare alert, expected proximitythere to are be produced to resultsremain as inyet unidentified.P values Continuously of 2 × 10 mon-andservations 10 ,respec- by imaging atmospheric Cherenkov in blazars: intense extragalactic radio, optical, itoring the entire sky for astrophysical neu- telescopes, notably theMajorAtmospheric
fied at higher energies. The muon-neutrino as- at several hundred TeV, an atmospheric originthe coordinatesindications at the of 3s TXSlevel of 0506+056, high-energy neutrino and their tively, corresponding to 3.5s and 3.7s.Because http://science.sciencemag.org/ Six data-taking periods make up the full x-ray, and, in some cases, g-ray sources Gamma Imaging Cherenkov (MAGIC) Candidate9.5-year data sample. Sample Neutrino numbers Source:clustering in time.characterized This TXS is illustrated by relativistic 0506+056 jets in of Fig. 2, there is no a priori reason to prefer onetelescopes, of the revealed periods where http://science.sciencemag.org/ plasma pointing close to our line of the detected g-ray flux from the blazar Fig. 1. Event display for which shows the time-independent weight of generic time windows over the other, we take the correspond to the number of detector sight. Blazars are among the most reached energies up to 400 GeV. Mea- neutrino event IceCube- strings that were operational. During the individual events inpowerful the likelihood objects in the analysis Universe and during more significant one and include a trial factorsurements of of the source have also 170922A. The time at which a the IC86b data period.are widely speculated to be sources 2forthefinalsignificance,whichisthen3.5beens completed. at x-ray, optical, and DOM observed a signal is first three periods, the detector was still of high-energy cosmic rays. These cos- radio wavelengths. We have inves- reflected in the color of the hit,under construction. The last three periods The Gaussian timemic rays window generate high-energy is centered neutri- at 13 Outside the 2012–2015 time period, thetigated next models associating neutrino with dark blues for earliest hitscorrespond to different data-taking December 2014 [modifiednos and g-rays, Julian which day are (MJD) produced 57004] most significant excess is found using the Gauss-and g-ray production and find that and yellow for latest. Times with an uncertaintywhen of the ±21 cosmic days rays andaccelerated a duration in ian window in 2017 and includes the IceCube-correlation of the neutrino with the conditions and/or event selections with the the jet interact with nearby gas or shown are relative to the first 35 flare of TXS 0506+056 is statistically
T =110 days. The best-fitting parameters for 170922A event. This time window is centered DOM hit according to the trackfull 86-string detector. W þ24 photons. On 22 September 2017, the IC170922A:significant at the level of 3 standard on July 12, 2018 À cubic-kilometer IceCube Neutrino deviations (sigma). On the basis of the reconstruction, and earlier and the fluence J100 = ∫F100(t)dt and the spectral at 22 September 2017 with duration TW =19days, later times are shown with the Observatory2 detected0:9 a ~290-TeV4 –Downloaded from 2 2 0:4 redshift4 of TXS 0506+056, we derive index are given by EneutrinoJ100 =2 from:1þ a dir0:7ection10 consistentÀ TeV cm g =1.7±0.6,andfluenceE J100 =0:2þ0:2 constraints10À for the muon-neutrino same colors as the first and À Â –2 À Â
with the flaring g-ray blazar TXS luminosity for this source and find
290 TeV neutrino on July 12, 2018 last times, respectively. The Sample Start End at 100 TeV and g = 2.1 ± 0.2, respectively. The TeV cm at 100 TeV. No other event besides the 0506+056. We report the details of them to be similar to the luminosity total time the event took to joint uncertainty on these parameters is shown IceCube-170922A event contributes significantly IC40 5 April 2008 20 May 2009 this observation and the results of a observed in g-rays. cross the detector is ~3000 ns...... in Fig. 3 along withmultiwavelength a skymap follow-up showing campaign. the result to the best fit. As a consequence,IceCube the Coll. uncertainty The size of a colored sphere isIC59 20 May 2009 31 May 2010 CO NCLU SIO N: The energies of the ...... proportional to the logarithm of the time-dependentRATIO NALE: analysisMultimessenger performed astron- at the on the best-fitting window location andg width-rays and the neutrino indicate that IC79 31 May 2010 13 May 2011 http://science.sciencemag.org/ Science (2018) of the amount of light ...... location of TXS 0506+056omy aims for globally and in coordinated its vicinity spans the entire IC86c period, because anyblazar win- jets may accelerate cosmic rays observed at the DOM, with IC86a 13 May 2011 16 May 2012 observations of cosmic rays, neutri- to at least several PeV. The observed ...... during the IC86bnos, data gravitational period. waves, and electro- dow containing IceCube-170922A yields a similarassociation of a high-energy neutrino Signalness: 56.5% larger spheres corresponding IC86b 16 May 2012 18 May 2015 to larger signals. The total ...... The box-shapedmagnetic time radiation window across is a broad centered value of the test statistic. Following the trialwith cor- a blazar during a period of en- range of wavelengths. The combi- hanced g-ray emission suggests that charge recorded is ~5800 photoelectrons.IC86c Inset 18 May is an 2015 overhead perspective 31 October view 2017 of the event. The13 daysbest-fitting later track with direction duration is shownT as=158days(from an arrow, rection procedure for different observation periods ...... nation is expectedW to yield crucial Multimessenger observations of blazar TXS 0506+056. The blazars may indeed be one of the long- 0:50 consistent with a zenith angle 5:7þ0:30 degrees below the horizon. MJD 56937.81 toinformation MJD 57096.21, on the mechanisms inclusive50% of and 90%as described containment above, regions for the the significance neutrino IceCube- of thissought excess sources of very-high-energy IceCube, Fermi-LAT, MAGIC, AGILE, ASAS-SN,À HAWC, H.E.S.S, INTEGRAL, Kapteyn, EM13±5 flare above the backgroundenergizing of the[Science mostatmospheric powerfulPage astro- 361 8 170922A (2018) (dashed neutrinos, red no.6398, and solid gray contours, eaat1378] respectively),3.5σ cosmic rays, and hence responsible for physical sources. That the produc- overlain on a V-band optical image of the sky. Gamma-ray sources a sizable fraction of the cosmic neu- Kanata, Kiso, Liverpool, Subaru, Swift, VERITAS, VLA, Science 2018 tion of neutrinos is accompanied by in this region previously detected with the Fermi spacecraft are trino flux observed by IceCube. electromagnetic radiation from the shown as blue circles, with sizes representing their 95% positional ▪ The list of author affiliations is available in the full source favors the chances of a multi- uncertainty and labeled with the source names. The IceCube on July 12, 2018 wavelength identification. In par- neutrino is coincident with the blazar TXS 0506+056, whose article online. *The full lists of participating members for each ticular, a measured association of optical position is shown by the pink square. The yellow circle team and their affiliations are provided in the high-energy neutrinos with a flaring shows the 95% positional uncertainty of very-high-energy g-rays supplementary materials. source of g-rays would elucidate the detected by the MAGIC telescopes during the follow-up campaign. †Email: [email protected] Cite this article as IceCube Collaboration et al., mechanisms and conditions for ac- The inset shows a magnified view of the region around TXS 0506+056 Science 361, eaat1378 (2018). DOI: 10.1126/
celeration of the highest-energy cos- on an R-band optical image of the sky. science.aat1378 IMAGES: PHOASAS-SN FOR THE V-BAND OPTICAL; KANATA FOR THE R-BAND IN MAGNIFIED VIEW
The IceCube Collaboration et al., Science 361, 146 (2018) 13 July 2018 1 of 1
4
Fig. 1. Time-dependent analysis results. The orange curve corresponds fitting time windows (durations TW) over all times T0, with the height to the analysisneutrino using the Gaussian-shaped flare time profile. The central time T indicating[Science the significance 361 (2018) of that window.no.6398, In each 147-151] period, the most 0
and width TW are plotted for the most significant excess found in each significant time window forms a plateau, shaded in blue. The large blue period, with the P value of that result indicated by the height of the peak. band centered near 2015 represents the best-fitting 158-day time window Fig. 2. Fermi-LATand MAGICThe observations blue curveMarkus of IceCube-170922A corresponds Ahlers (NBI)’ tos the analysispixel, using using detected theNeutrinos box-shaped photons with and energyγ time-rays of 1from to 300found Extragalactic GeV in using a 2° by the 2° Sources box-shaped timeAugust profile. The28, 2018 vertical dotted slide line 19 in IC86c location. Sky position of IceCube-170922A in J2000 equatorial coordinates region around TXS0506+056. The map has a pixel size of 0.02° and was overlaying the g-ray countsprofile. from Fermi The-LAT curve above 1traces GeV (A) the and theouter signal edgesmoothed of the with superposition a 0.02°-wide Gaussian of the kernel. best- MAGICindicates data are shown the as time of the IceCube-170922A event. significance as observed by MAGIC (B) in this region. The tan square signal significance for g-rays above 90 GeV. Also shown are the locations of indicates the position reported in the initial alert, and the green square a g-ray source observed by Fermi-LAT as given in the Fermi-LAT Third indicates the final best-fittingIceCube position Collaboration, from follow-up reconstructionsScience 361, (18 147). –151Source (2018) Catalog 13 July (3FGL) 2018 (23) and the Third Catalog of Hard Fermi-LAT 2 of 5 Gray and red curves show the 50% and 90% neutrino containment regions, Sources (3FHL) (24) source catalogs, including the identified positionally respectively, including statistical and systematic errors. Fermi-LAT data are coincident 3FGL object TXS 0506+056. For Fermi-LAT catalog objects, shown as a photon counts map in 9.5 years of data in units of counts per marker sizes indicate the 95% CL positional uncertainty of the source.
The IceCube Collaboration et al., Science 361, eaat1378 (2018) 13 July 2018 2 of 8 RESEARCH | RESEARCH ARTICLE
as a fitted parameter. The model parameters are the RA coordinates within each data-taking period. contributing events at boundary times). For the correlated and are expressed as a pair, (F100, g), The resultant P value is defined as the fraction of box-shaped time window, the uncertainties are where F100 is the flux normalization at 100 TeV. randomized trials yielding a value of TS greater than discontinuous and not well defined, but the un- NeutrinoThe time-dependent flares analysis and uses the TXS same for- 0506+56or equal to the one obtainedPost-trials for the actual p-value data. certainties for association: for the Gaussian window 3.0σ show that it mulation of the likelihood but searches for Because the detector configuration and event is consistent with the box-shaped time window Theclustering Multi-Messenger in time as well as space by introducing selections Light changed as shownCurve in Table 1, the time- fit. Despite the different window shapes, which an additional time profile. It is performed sep- dependent analysis is performed by operating on lead to different weightings of the events as a IceCubearately Collaboration, for two different Fermi-LAT, generic profile MAGIC, shapes:a AGILE,each ASAS-SN, data-taking periodHAWC, separately. H.E.S.S., (A flareINTEGRAL, that function et al., ofScience time, both (2018) windows identify the same Gaussian-shaped time window and a box-shaped spans a boundaryRESEARCH between two periods could be time interval as significant. For the box-shaped time window. Each analysis varies the central partially detected in either period, but with re- time window, the best-fitting parameters are sim- ◥ trinos, IceCube provides real-time triggers for time of the window, T0,andthedurationTW duced significance.) An additional look-elsewhere ilar to those of the Gaussian window, with fluence RESEARCH ARTICLE SUMMARY observatories2 around the world measuring VHE(from gamma-rays seconds to years) of the potential signal to correction then needs to be applied for a result in at 100 TeV and spectral index giveng-rays, by x-rays,E J100 optical,= radio, and gravitational T T 1:0 4 –2 waves, allowing for the potential identification find the four parameters (F100, g, 0, W)that an individual dataNEUTRINO segment, ASTROPHYSICS given by the ratio of 2:2þ0:8 10À TeV cm and g =2.2±0.2.This maximize the likelihood ratio, which is defined the total 9.5-year observation time to the obser- fluenceÀ  corresponds to an averageof even flux rapidly over fading sources. 0:7 15 –1 –2 –1 as the test statistic TS. (For the Gaussian time vation time of that data segment (30). 158 days of F100 =1:6þ0:6 10À TeVRESU LTS:cm A high-energys . neutrino-induced Multimessenger observations ofÀ a  muon track was detected on 22 September 2017, GeVwindow, gamma-raysTW represents twice the standard de- When we estimate the significance of the time- Neutrinos from the direction of automatically generating an alert that was viation.) The test statistic includes a factor that flaring blazar coincidentdependent with result by performing the analysis at distributed worldwide ◥ corrects for the look-elsewhere effect arising TXS 0506+056 the coordinates of TXS 0506+056 onON randomized OUR WEBSITE within 1 min of detection and prompted follow-up from all of the possible time windows that could The results of thehigh-energy time-dependent analysis neutrino per- datasets, IceCube-170922A we allow in each trial a newRead the fit full for article all at http://dx.doi. searches by telescopes over RESEARCH ARTICLE RESEARCH | X-raysbe chosen (30). formed at the coordinatesThe IceCube of Collaboration, TXS 0506+056Fermi-LAT, are MAGIC,theAGILE parameters:, ASAS-SN, HAWC,F100, g H.E.S.S.,, T0, TW. Weorg/10.1126/ find that the a broad range of wave- INTEGRAL, K anata, K iso, K apteyn, Liverpool Telescope, Subaru, Swift/NuSTAR, science.aat1378 lengths. On 28 September For each analysis method (time-integrated and shown in Fig. 1 for each of the six data periods. fraction of randomized trials that result...... in a more 2017, the Fermi Large Area VERITAS, and VLA/17B-403 teams*† –5 Downloaded from lower limit of 183 TeV, dependingtime-dependent), only weakly on a robusttrophysical significance spectrum, together estimate with is simulatedOne ofcannot the data be excluded. periods, Elect IC86bromagnetic from observations 2012 to 2015, significant excess than the real dataTelescope is 7 × 10 Collaborationfor reported that the di- the assumed astrophysicalobtained energy spectrum by performing (25). data, was the used identical to calculate analysis the probability on thatcontains a are avaluable significant to assess excess, the possible which association is identified of the box-shaped time window and 3 rection× 10–5 offor the neutrino the was coincident with a The vast majority of neutrinos detected by neutrino at the observed track energy and zenith a single neutrinoINTRO to an DU astrophysical CTIO N: Neutrinos source. are tracers of mic rays. The discovery of an extraterrestrial cataloged g-ray source, 0.1° from the neutrino IceCube arise from cosmic-raytrials interactions with randomized within angle datasets. in IceCube These is of are astrophysical produced origin.by This both time-windowFollowing thecosmic-ray alert, shapes. IceCube acceleration: The performed excess electrically consists a neutral Gaussiandiffuse flux of high-energy time window. neutrinos, Thisannounced fraction,direction. once The cor- source, a blazar known as TXS X-rays spectral index Downloaded from Earth’s atmosphere. Althoughby randomizingatmospheric neu- theprobability, event times the so-called and recalculating signalness of the eventof 13 ±complete 5 events analysis aboveand traveling of the relevant expectation at nearly data the prior speed from ofto light, the they rectedby IceCube for in the 2013, ratio has characteristic of the total prop- observation0506+056 at time a measured redshift of 0.34, was trinos are dominant at energies below 100 TeV, (14), was reported to be 56.5% (17). Although 31 October 2017.can Although escape the no densest additional environments excess and may erties that hint at contributions from extra- in a flaring state at the time with enhanced atmospheric background.be tracedThe back to significance their source of depends origin. High- togalactic the sources, IC86b although observation the individual time sources (9.5 years/3g-ray activity years), in the GeV range. Follow-up ob- their spectrum falls steeply with energy, allowing IceCube can robustly identify astrophysical neu- of neutrinos was found from the direction of TXS –4 –4 astrophysical neutrinos to be moreTable easily 1. identi-IceCubetrinos neutrino at PeV data energies, samples. for individual neutrinoson the0506+056 energies near ofenergy the the time events, neutrinos of the theirare alert, expected proximitythere to are be produced to resultsremain as inyet unidentified.P values Continuously of 2 × 10 mon-andservations 10 ,respec- by imaging atmospheric Cherenkov in blazars: intense extragalactic radio, optical, itoring the entire sky for astrophysical neu- telescopes, notably theMajorAtmospheric
fied at higher energies. The muon-neutrino as- at several hundred TeV, an atmospheric originthe coordinatesindications at the of 3s TXSlevel of 0506+056, high-energy neutrino and their tively, corresponding to 3.5s and 3.7s.Because http://science.sciencemag.org/ Six data-taking periods make up the full x-ray, and, in some cases, g-ray sources Gamma Imaging Cherenkov (MAGIC)
optical http://science.sciencemag.org/ Candidate9.5-year data sample. Sample Neutrino numbers Source:clustering in time.characterized This TXS is illustrated by relativistic 0506+056 jets in of Fig. 2, there is no a priori reason to prefer onetelescopes, of the revealed periods where plasma pointing close to our line of the detected g-ray flux from the blazar Fig. 1. Event display for which shows the time-independent weight of generic time windows over the other, we take the correspond to the number of detector sight. Blazars are among the most reached energies up to 400 GeV. Mea- neutrino event IceCube- strings that were operational. During the individual events inpowerful the likelihood objects in the analysis Universe and during more significant one and include a trial factorsurements of of the source have also 170922A. The time at which a the IC86b data period.are widely speculated to be sources 2forthefinalsignificance,whichisthen3.5beens completed. at x-ray, optical, and DOM observed a signal is first three periods, the detector was still The Gaussian timeof high-energy window cosmic is rays. centered These cos- at 13 Outside the 2012–2015 time period, theradio next wavelengths. We have inves- reflected in the colorradio of the hit, under construction. The last three periods mic rays generate high-energy neutri- tigated models associating neutrino with dark blues for earliest hitscorrespond to different data-taking December 2014 [modifiednos and g-rays, Julian which day are (MJD) produced 57004] most significant excess is found using the Gauss-and g-ray production and find that and yellow for latest. Times with an uncertaintywhen of the ±21 cosmic days rays andaccelerated a duration in ian window in 2017 and includes the IceCube-correlation of the neutrino with the conditions and/or event selections with the the jet interact with nearby gas or shown are relative to the first 35 flare of TXS 0506+056 is statistically
T =110 days. The best-fitting parameters for 170922A event. This time window is centered DOM hit according to the trackfull 86-string detector. W þ24 photons. On 22 September 2017, the significant at the level of 3 standard on July 12, 2018 À cubic-kilometer IceCube Neutrino deviations (sigma). On the basis of the reconstruction, and earlier and the fluence J100 = ∫F100(t)dt and the spectral at 22 September 2017 with duration TW =19days, later times are shown with the Observatory2 detected0:9 a ~290-TeV4 –Downloaded from 2 2 0:4 redshift4 of TXS 0506+056, we derive index are given by EneutrinoJ100 =2 from:1þ a dir0:7ection10 consistentÀ TeV cm g =1.7±0.6,andfluenceE J100 =0:2þ0:2 constraints10À for the muon-neutrino same colors as the first and À Â –2 À Â
with the flaring g-ray blazar TXS luminosity for this source and find last times, respectively. The Sample Start End at 100 TeV and g = 2.1 ± 0.2, respectively. The TeV cm at 100 TeV. No other event besides the on July 12, 2018 0506+056. We report the details of them to be similar to the luminosity total time the event took to joint uncertainty onthis observation these parameters and the results is of shown a IceCube-170922A event contributes significantlyobserved in g-rays. cross the detector is ~3000IceCube ns.IC40, Fermi-LAT, 5 April 2008 MAGIC, 20 May 2009 AGILE, ASAS-SN, HAWC, H.E.S.S, INTEGRAL, Kapteyn, ...... in Fig. 3 along withmultiwavelength a skymap follow-up showing campaign. the result to the best fit. As a consequence,Page 16 the uncertainty The size of a colored sphere isIC59 20 May 2009 31 May 2010 CO NCLU SIO N: The energies of the Kanata,...... Kiso, Liverpool, Subaru, Swift, VERITAS, VLA, Science 2018 proportional to the logarithm of the time-dependentRATIO NALE: analysisMultimessenger performed astron- at the on the best-fitting window location andg width-rays and the neutrino indicate that IC79 31 May 2010 13 May 2011 http://science.sciencemag.org/ of the amount of light ...... location of TXS 0506+056omy aims for globally and in coordinated its vicinity spans the entire IC86c period, because anyblazar win- jets may accelerate cosmic rays observed at the DOM, with IC86a 13 May 2011 16 May 2012 observations of cosmic rays, neutri- to at least several PeV. The observed ...... during the IC86bnos, data gravitational period. waves, and electro- dow containing IceCube-170922A yields a similarassociation of a high-energy neutrino larger spheres corresponding IC86b 16 May 2012 18 May 2015 to larger signals. The total ...... The box-shapedmagnetic time radiation window across is a broad centered value of the test statistic. Following the trialwith cor- a blazar during a period of en- range of wavelengths. The combi- hanced g-ray emission suggests that charge recorded is ~5800 photoelectrons.IC86c Inset 18 May is an 2015 overhead perspective 31 October view 2017 of the event. The13 daysbest-fitting later track with direction duration is shownT as=158days(from an arrow, rection procedure for different observation periods z~0.337 Paiano...... et al. 2018 nation is expectedW to yield crucial Multimessenger observations of blazar TXS 0506+056. The blazars5 may indeed be one of the long- 0:50 consistent with a zenith angle 5:7þ0:30 degrees below the horizon. MJD 56937.81 toinformation MJD 57096.21, on the mechanisms inclusive50% of and 90%as described containment above, regions for the the significance neutrino IceCube- of thissought excess sources of very-high-energy À EM flare energizing the[Science most powerful astro- 361170922A (2018) (dashed red no.6398, and solid gray contours, eaat1378] respectively), cosmic rays, and hence responsible for physical sources. That the produc- overlain on a V-band optical image of the sky. Gamma-ray sources a sizable fraction of the cosmic neu- tion of neutrinos is accompanied by in this region previously detected with the Fermi spacecraft are trino flux observed by IceCube. electromagnetic radiation from the shown as blue circles, with sizes representing their 95% positional ▪ The list of author affiliations is available in the full source favors the chances of a multi- uncertainty and labeled with the source names. The IceCube on July 12, 2018 wavelength identification. In par- neutrino is coincident with the blazar TXS 0506+056, whose article online. *The full lists of participating members for each ticular, a measured association of optical position is shown by the pink square. The yellow circle team and their affiliations are provided in the high-energy neutrinos with a flaring shows the 95% positional uncertainty of very-high-energy g-rays supplementary materials. source of g-rays would elucidate the detected by the MAGIC telescopes during the follow-up campaign. †Email: [email protected] Cite this article as IceCube Collaboration et al., mechanisms and conditions for ac- The inset shows a magnified view of the region around TXS 0506+056 Science 361, eaat1378 (2018). DOI: 10.1126/
celeration of the highest-energy cos- on an R-band optical image of the sky. science.aat1378 IMAGES: PHOASAS-SN FOR THE V-BAND OPTICAL; KANATA FOR THE R-BAND IN MAGNIFIED VIEW
The IceCube Collaboration et al., Science 361, 146 (2018) 13 July 2018 1 of 1
Fig. 1. Time-dependent analysis results. The orange curve corresponds fitting time windows (durations TW) over all times T0, with the height to the analysisneutrino using the Gaussian-shaped flare time profile. The central time T indicating[Science the significance 361 (2018) of that window.no.6398, In each 147-151] period, the most 0
and width TW are plotted for the most significant excess found in each significant time window forms a plateau, shaded in blue. The large blue period, with the P value of that result indicated by the height of the peak. band centered near 2015 represents the best-fitting 158-day time window Fig. 2. Fermi-LATand MAGICThe observations blue curveMarkus of IceCube-170922A corresponds Ahlers (NBI)’ tos the analysispixel, using using detected theNeutrinos box-shaped photons with and energyγ time-rays of 1from to 300found Extragalactic GeV in using a 2° by the 2° Sources box-shaped timeAugust profile. The28, 2018 vertical dotted slide line 19 in IC86c location. Sky position of IceCube-170922A in J2000 equatorial coordinates region around TXS0506+056. The map has a pixel size of 0.02° and was overlaying the g-ray countsprofile. from Fermi The-LAT curve above 1 traces GeV (A) the and theouter signal edgesmoothed of the with superposition a 0.02°-wide Gaussian of the kernel. best- MAGICindicates data are shown the as time of the IceCube-170922A event. significance as observed by MAGIC (B) in this region. The tan square signal significance for g-rays above 90 GeV. Also shown are the locations of indicates the position reported in the initial alert, and the green square a g-ray source observed by Fermi-LAT as given in the Fermi-LAT Third indicates the final best-fittingIceCube position Collaboration, from follow-up reconstructionsScience 361, (18 147). –151Source (2018) Catalog 13 July (3FGL) 2018 (23) and the Third Catalog of Hard Fermi-LAT 2 of 5 Gray and red curves show the 50% and 90% neutrino containment regions, Sources (3FHL) (24) source catalogs, including the identified positionally respectively, including statistical and systematic errors. Fermi-LAT data are coincident 3FGL object TXS 0506+056. For Fermi-LAT catalog objects, shown as a photon counts map in 9.5 years of data in units of counts per marker sizes indicate the 95% CL positional uncertainty of the source.
The IceCube Collaboration et al., Science 361, eaat1378 (2018) 13 July 2018 2 of 8 16
expectation of an associated HE neutrino detection by Table 8. Parameter values for hadronic models (HMs) for IceCube. TXS 0506+056 discussed in the text and presented in Fig. 6.
3.3. Hadronic Models (HMs) HM1 HM2 HM3 In hadronic scenarios, while the low-energy peak in the B! [G] 85 16 blazar’s SED is explained by synchrotron radiation from R! [in 10 cm] 2 3 4.5 14 relativistic primary electrons, the HE peak is explained δ 5.2 10 15 by EM cascades induced by pions and muons as de- 43 1 Table 7. Model-specific parameter values for leptonic models (LMs)L! for[in TXS 10 0506+056erg s− ]9.3 discussed 0.6 in 0.06 the text cay products of the photomeson production (Mannheim e 1993; M¨ucke et al. 2003), or synchrotron radiation from se,1 1.8 LMBB1a LMBB1b LMBB1c LMBB2a LMBB2b LMBB2c LMPL1a LMPL1b LMPL2a LMPL2b relativistic protons in the ultrahigh-energy range (Aha- s 4.2 3.6 3.6 (max) 44 1 e,2 Lp! [10 erg s− ]0.540.270.341 5.4100.540.541010 ronian 2000; M¨ucke et al. 2003). We coin this scenario 2 ! sp 22.53 2 2γ 2e,min [in 2 10 ]6.311 2 2 2 “HM”, which stands for Hadronic Model6 ,inreference6 2 γ! 1310 3 10 1111111 p,min × × γe,! br [in 10 ]7.96.35 to the hadronic origin8 of the γ-rays. The synchrotron γp,! max [10 ]3030301.60.160.01630300.0160.0164 γ! 10 and IC emission of3 secondary pairs may have an im- e,max uext! [erg cm− ]0.0330.0330.0670.040.08 46 1 5 portant contributionT ! [K] to the bolometric radiation 3 of10 the Lp! [in 10 erg s− ]2.70.10.01n/a source. In contrast to the leptonic scenario (Sec.× 3.2), α n/asp 3 2 32.1 2
the parametersεmin! [keV] describing the proton distribution n/a can be 0.05 γp,! min 1 directly constrainedεmax! [keV] from the NuSTAR and Fermi n/aLAT 5 9 γ! 2 10 data. For the TXS 0506+056 flare, in the hadronic sce- p,max × Notenario,—See the Table SED5 for can parameter be fully definitions, explained and without Table 6 invokingfor parameter values common to all LMs. In LMBB models, the external photon fieldexternal is blackbody-like radiation with fields. comoving temperature T !,whileinLMPLmodels,itisapower-lawbetweencomovingenergiesNote—Parameter definitions are providedε inmin! Tableand 5ε.max! , withThere photon are index diffαerent.Inallcases, combinationsuext! is the of comovingparameters energy that density of the external photon field. Note that the isotropic-equivalent 4 cosmic-ray proton luminosity is Lp = δ Lp! . can successfullyCandidate explain Neutrino the SED in Source: the HM sce- TXS 0506+056 Keivani et al. 2018 nario (B¨ottcher et al. 2013; Cerruti et al. 2015). As a starting point,leptonic we search model for combinations of δ and hadronic model 1047 −11 1047 10 δ=δ ∼ 5 B! that lead to(max) rough energy equipartition between eq Lp δ=10 LMBB1a the10− magnetic10 field× (max) and protons, since the primary elec- 10−10 LMBB1b 2 Lp LM δ=15 HM tron energy density is negligible in this scenario. With LMBB1c LMBB2a analytical calculations we derive rough estimates10 of46 the cascade 1046 LMBB2b ] 1 ] parameter values for equipartition: δeq 5, Beq! 80 G, LMBB2c ] 1 1 − −12 − −11 −11 − ] 16 9 ∼ ∼ ] 10 1 10 1 10 s LMPL1a − R 10 cm, and ε 10 GeV (Petropoulou & − eq! p,! max ! 2 − s s 2 ∼ ∼ 2 LMPL1b − Dermer 2016). − ! LMPL2a 1045 1045 The parameter values obtained by numerically mod- LMPL2b [erg cm [erg cm HM3 ε elingε −12 the SED (see Fig. 6) are summarized in Table 8 −12 [erg cm
10 ν 10 F F at z=0.33 [erg s at z=0.33 [erg s ε ε ε ε ε
and are similar to the estimates provided above. The L L
F −13 ε ε ν
10 44 ε 44 jet power computed for this parameter set (HM1)10 is 10 close to the minimum value expected in the hadronic −13 X-rays −13 X-rays scenarios.10 10−12 More specifically, the absolute power of a 10 10−12 two-sided jet inferred for these parameters is Lj 103 104 105 103 104 105 2 2 47 101 43 ≈ 1043 2πcR! (δ/2) (up! + ue! + uB! ) 4 10 erg s− ,with −14 0 5 3 ∼10 × 15 10 0 5 10 15 up! 2uB! 10500 erg cm10 − ,where10 up! , u10e! , uB! are comov- 102 10 4 10 6 10 8 ≈ ∼ ε [eV] 10 10 ε [eV] 10 10 ing energy densities of relativistic protons, electrons, and ε leptonic with radiatively subdominant hadronic component cascade implies high X-rayν [TeV] level to match magnetic fields, respectively. As demonstrated in Fig. 6, detection proba. with IC in real time during 6-month flare = 1-2% observed neutrino flux Figurethe emission 4. fromLeptonic the EM Model cascade (LMBB2b) forms a “bridge” for the be- Figure 6. Hadronic Model (HM3) for the SED of TXS 0506+056 flare (Ep. 1). Two SED cases (gray TXS 0506+056 flare (Ep. 1), as computed for different values tween the low-energy and high-energy peaks of the SEDlucky?Figure 5. Upper limits on the all-flavorGao neutrinoet al. 2018 (ν +¯ν) lines) are plotted against the observations (colored points, of the Doppler factor (gray curves), together with resulting for δ = δ (graybut dottedthe 2014-2015 line). Despite neutrino minimizing flares require the higherfluxes rates predicted for our modeling of theCerruti SED in et the al. leptonic2018 showing allowedeq ranges at 90% confidence), one with all-flavor neutrino fluxes (red curves) and electromagnetic (L~ 1047 erg/s over 158 days ~ 4 x average gamma-ray(LMx) andluminosity) hadronic (HMx) models. Zhang, Fang & Li 2018 hadronicpower of component the jet, the set adopted to the set maximum of parameters allowed for proton HM1 observations (colored points, showing allowed ranges at 90% (max) 50 1 Gokus et al.11 2018 luminositycannot explainLp the2 SED10 dueerg to s the− (solid associated gray), significant and the confidence). Photon attenuation at εγ > 3 10 eV due to ≈ × Sahakyan× 2018 otherEM set cascade to twice component. this maximal value (dashed gray line). interactionsProton maximum with the energy extragalactic — Motivated background∼ by the light hypoth- is not Multi-zone or more complicated models? CorrespondingThe EM cascade all-flavor emission neutrino can fluxes be forsuppressed the maximal if the esisincluded that blazars here. are UHECR accelerators, i.e., at ener- - Additional photomeson production by external radiation fields Murase, 18 (solidsource red) becomes and “twice less opaque maximal” to (dashed the intra-source line) casesγγ areab- gies above 3 10 eV (Murase et al. 2012), we ex- - hadronuclear production (e.g., jet-cloud interaction)11 Oikonomou,× also shown. Photon attenuation at εγ > 3 10 eV due to plore the effect of the proton maximum energy on the More parameters introduced, the setup is× ad-hoc Petropoulou 2018 interactions with the extragalactic background∼ light is not neutrino flux upper limits. We thus explore cases with included here. 8 9 9 6 γ! =1.6 10 , 1.6 10 , and 3 10 – see Ta- p,max × × × ble 7. Our results on the neutrino fluxes are presented In what follows, we show that our neutrino flux limits in Fig. 5. are fairly insensitive to the exact parameter values that Neutrino spectra in the LMBB1x models are more may affect the photomeson production optical depth. extended in energy compared to the default case (LMBB2b). They peak around 10 PeV (100 PeV) for Multi-messengers! cosmic rays + others —> temporal coincidence impossible (deflections) but studies of di!use fluxes
Gravitational waves GW + electromagnetic (+gamma) GW170817
neutrinos + gammas? TXS0506+056 (blazar flare) Cosmic rays
p Fe ν Gamma rays
ν GW + neutrinos ???
Neutrinos
7 15