Detection of VHE gamma-ray transients with monitoring facilities
G. La Mura, G. Chiaro, R. Conceiçao, A. De Angelis, M. Pimenta, B. Tomé
MMNRAS , 000–000 (2020) Preprint 20 July 2020 Compiled using MNRAS L A TEX style file v3.0
Detection of VHE gamma-ray transients with monitoringfacilities
G. La Mura (cid:63) , G. Chiaro, R. Concei¸c˜ao, , A. De Angelis, , , M. Pimenta, , B. Tom´e , Laborat´orio de Instrumenta¸c˜ao e F´ısica Experimental de Part´ıculas (LIP), Av. Prof. Gama Pinto 2, 1649-003 Lisboa, Portugal Istituto di Astrofisica Spaziale e Fisica cosmica - INAF , Via A. Corti 12, 20133 Milano, Italia Instituto Superior T´ecnico (IST), Av. Rovisco Pais 1, 1049-001 Lisboa, Porugal Dipartimento di Fisica e Astronomia - Universit`a di Padova, Via Marzolo 8, 35131 Padova, Italia Dipartimento di scienze matematiche, informatiche e fisiche - Universit`a degli Studi di Udine, Via Palladio 8, 33100 Udine, Italia Istituto Nazionale di Fisica Nucleare sez. Padova (INFN), Via Marzolo 8, 35131 Padova, Italia
Received 2020 June 18. Accepted 2020 July 17; in original form 2020 January 13
ABSTRACT
The observation of Very High Energy γ rays (VHE, E >
100 GeV) led us to the iden-tification of extremely energetic processes and particle acceleration sites both withinour Galaxy and beyond. We expect that VHE facilities, like CTA, will explore thesesources with an unprecedented level of detail. However, the transient and unpredictablenature of many important processes requires the development of proper monitoringstrategies, to observe them. With this study, we estimate the properties of VHE tran-sients that can be effectively detected by monitoring facilities. We use data collectedby the
Fermi -LAT instrument, during its monitoring campaign, to select events thatare likely associated with VHE emission. We use this sample to estimate the frequency,the luminosity and the time-scales of different transients, focusing on blazar flares andGamma Ray Bursts (GRBs). We discuss how the balance between Field of View, sen-sitivity and duty cycle of an observatory affects the likelihood to detect transients thatoccur at the inferred rates and we conclude describing the contribution that currentand near-future monitoring facilities can bring to the identification and study of VHEtransient emission.
Key words: instrumentation: detectors – gamma rays: general – galaxies: active –gamma ray burst: general
The recent detection of a Gamma-Ray Burst (GRB) asso-ciated with a Gravitational Wave event (GW170817, Ab-bott et al. 2017a,b) and a flaring blazar consistent with thedirection of an ultra-relativistic neutrino (TXS 0506+056,IceCube Collaboration 2018) demonstrated the importanceof γ -ray monitoring in the identification of multi-messengerevents. In both cases, the existence of an electromagneticcounterpart to the signals was first identified by γ -ray in-struments, namely the Gamma-ray Burst Monitor (GBM,Meegan et al. 2009), for the neutron star merger originatingGW170817, and the Large Area Telescope (LAT, Atwoodet al. 2009), for the blazar flare associated with IceCube-170922A, the two instruments carried by the Fermi Gamma-ray Space Telescope . Follow-up observations executed in dif-ferent wavelengths led to the clarification of other important (cid:63)
E-mail : [email protected] characteristics, such as the identification of the GRB coun-terpart as a kilonova (e.g. Cowperthwaite et al. 2017), aswell as the redshift and, hence, the distance and the lumi-nosity of the processes (Blanchard et al. 2017; Paiano et al.2018). In particular, the blazar TXS 0506+056 has been de-tected at Very High Energy (VHE, E ≥
100 GeV) for thefirst time by MAGIC, shortly after the Fermi outburst (Mir-zoyan 2017).Detecting VHE photons from sources like blazars andGRBs has several important implications. On one hand, theemission of VHE radiation in close connection with the pro-duction of ultra-relativistic particles is a strong hint towardsthe role of these sources as cosmic particle accelerators. Onthe other, the interaction of γ rays with the lower energyphoton field, forming the Extragalactic Background Light(EBL), is an extremely powerful tool to constrain the effectsof star formation and active galactic nuclei (AGN) in theprocess of cosmological evolution. During its long monitor-ing campaign, the Fermi -LAT telescope has firmly identified c (cid:13) a r X i v : . [ a s t r o - ph . I M ] J u l G. La Mura et al.
AGNs - blazars in particular - and GRBs as the most pow-erful extragalactic sources of photons above 10 GeV (Ajelloet al. 2017, 2019). Some AGNs are known to have energyspectra that extend up to several TeV, while the combi-nation of observed GRB spectra with the identification ofreliable counterparts at measurable redshifts suggested thatGRBs could be intrinsically able to produce photons wellabove E = 100 GeV. This expectation would be eventuallyconfirmed by the MAGIC and H.E.S.S. observations of someGRBs (Mirzoyan 2019; Abdalla et al. 2019; MAGIC Collab-oration 2019).In recent times, our ability to observe VHE sourcesgreatly improved, thanks to the construction of largeground-based observatories using either the Imaging Atmo-spheric Cherenkov Telescope approach (IACT, like H.E.S.S.,VERITAS, and MAGIC, Aharonian et al. 2006; Holder et al.2008; Aleksi´c et al. 2016) or the Extensive Air Shower de-tector array (EAS, such as HAWC and LHAASO, DeYoung2012; Di Sciascio & LHAASO Collaboration 2016). Due tothe strong implications of VHE observations on the physicsof relativistic jets and light propagation through the Uni-verse, great efforts are currently ongoing to improve thecharacteristics of these observatories. We can expect thatthe upcoming Cherenkov Telescope Array (CTA, CTA Con-sortium 2019) will achieve outstanding performances in thisfield. Although CTA will be able to observe the VHE sky atan unprecedented level of detail, its ability to perform reg-ular monitoring of sources, to map vast regions of the skyand to promptly respond to fast transients will be strictlylimited by its narrow field of view and its duty cycle. Whilededicated monitoring programs, such as the one carried outby FACT (Anderhub et al. 2013), can cover a list of se-lected targets and, possibly, provide transient follow-up ca-pabilities, the necessary observational requirements implyunavoidable gaps in the data flow. On the other hand, thepresence of survey instruments that continuously scan wideareas of the sky, like Fermi, is essential to study transientphenomena. Unfortunately, the maintenance of long termsurvey missions in space is hard to attain and it comes atthe cost of reduced efficiency in the VHE domain. It hasbeen proposed that the observational coverage and continu-ity problems can be overtaken using the wide Field of View(FoV) of EAS facilities. In this work, we use γ -ray monitor-ing data to study the distribution of transients that are morelikely to be associated with VHE emission. We analyze theproperties of these events and we compare them with theperformance of instruments used to monitor or follow-uptransients. The paper is structured as follows: in §
2, we use
Fermi -LAT data to study the distribution of events that areknown or predicted to produce VHE emission at a detectablelevel for CTA, describing our selection, assumptions and im-plications; in §
3, we discuss the possibilities that EAS arrayshave to contribute to VHE transient monitoring; in § § VHE transients are a fundamental probe of multi-messengerastrophysics and particle physics, but, at present, very little is known about the rate and the properties of these events.Since the performance of VHE instruments will have a crit-ical impact on our ability to explore these events, we carryout a systematic study to assess how monitoring facilitiescan contribute to their investigation, by taking into accountthe source positions in the sky, their redshift, and the time-scales with which they can be detected. We develop thisanalysis in a sequence of steps that, starting from the avail-able data, leads us to infer what instrument characteristicsare required and which performances can be achieved. Asa starting point, we use Fermi-LAT monitoring data to se-lect a sample of transients that led to HE emission and aremost likely connected with VHE activity. We use the factthat some of these transients are associated with sourcesthat have been already detected at VHE, suggesting thattheir spectra are not subject to severe cut-offs. Finally, wecompare the selected data and our assumptions with theperformance of different observing facilities, demonstratingthat instruments with optimized monitoring capabilities canboth act as effective alert systems and explore a spectralwindow that would be otherwise very difficult to monitor.
We want to determine the distribution and the properties ofthe most important extragalactic VHE sources. To achievethis purpose, we need to use a model that can be comparedwith observational data. In general, we may express the VHEphoton spectrum observed from an astrophysical source atredshift z in the form of:d N ( E )d E = N (cid:18) EE (cid:19) − [ α + β log ( E/E )] e − [ τ E ( z )+ E/E c . o . ] [GeV − cm − s − ] , (1)where α is the photon index ( α ≥ . β is the curvature parameter ( β = 0 for a purepower-law spectrum), E is a scaling energy, E c . o . is thecut-off energy, and τ E ( z ) is the Universe opacity at energy E as a function of redshift.Eq. (1) implies that the number of photons availableat energies E (cid:29) E can quickly become lower than 1 pho-ton every few square metres. This means that instrumentswith small collecting areas, like those carried by satellites,require long observing times to characterize the high energyspectra of sources and they may not be able to effectively de-tect the highest energy emission of a short time-scale event.Thanks to the long duration of the Fermi -LAT monitoringcampaign, however, we are able to select a sample of sources,among the ones which were detected at high energy (HE,
E >
10 GeV) and to study their flaring activity.
Since most EBL models predict that the Universe opticaldepth for γ -ray photons with E (cid:39) τ ≥ z ≈ . E <
MNRAS000
MNRAS000 , 000–000 (2020) onitoring VHE transients To study the distribution of VHE flaring sources, we un-dertook an analysis aimed at characterizing the frequency,the duration, and the possible spectral features of VHE AGNflares. We used the 2 nd Fermi -LAT All-Sky Variability Anal-ysis (2FAV, Abdollahi et al. 2017) as a starting point toselect a uniform sample of γ -ray flares. Since the second ver-sion of the catalog includes 7 . γ -ray light curves of specific skyareas and the possibility to extract preliminary spectral fitsto the soft (100-800 MeV) and the hard (0 . γ -ray flares and were detected with an energy flux largerthan 10 − erg cm − s − above 10 GeV in the Third Catalogof Hard Fermi -LAT sources (3FHL, Ajello et al. 2017). Thisselection led to the identification of 160 sources, which havebeen associated with 2367 γ -ray flares, detected in the 2FAVhard band.Although using the Fermi -LAT variability analysis tostudy the properties of flares has some limitations, we canidentify examples of flares that triggered specific follow-upanalysis. Comparing 2FAV selected flares with almost simul-taneous analyses of data suggests that the power-law fits tothe 2FAV hard band give reasonable representations of thespectral energy distribution (SED). In general, flares are as-sociated with harder spectral indexes than those that char-acterize the average SEDs of the 8 yr data set of the Fourth
Fermi -LAT Gamma-ray Catalogue (4FGL, The Fermi-LATcollaboration 2019). While this procedure can effectivelyidentify the events with strongest HE activity, the extrap-olation of spectra towards the VHE domain is subject tosome degree of uncertainty, because sources may exhibit in-trinsic spectral cut-offs. These, however, are not expectedto be found below an energy of approximately 300 GeV. In-deed, some objects associated with our flare sample are de-tected by Fermi-LAT above 100 GeV, although several yearsof monitoring were necessary to collect the required statis-tics. In addition, many of the sources that exhibited thebrightest flares have been detected at VHE by IACT facili-ties (Wakely & Horan 2008), suggesting that intrinsic spec-tral cut-offs are unlikely to occur below 500 GeV. Therefore,in general, we do not expect severe cut-offs in flaring states,where, on the contrary, many works point to the existenceof possible additional emission components (H. E. S. S. Col-laboration 2017; Prince et al. 2017; Zacharias et al. 2019),while it is well established that the blazar TXS 0506+056was first detected in the VHE domain after being identifiedas a LAT flaring blazar consistent with an IceCube neutrinoevent (Abeysekara et al. 2018).The selected data describe observations under the ef-fect of the EBL opacity. Therefore, we can use the spectralparameters inferred from the variability analysis to estimatethe flux produced by flaring sources in the VHE domain. Todo this, we apply the EBL absorption model of Dom´ınguezet al. (2011) to the extrapolation of Eq. 1 to
E >
100 GeVand we use the resulting fluxes to select flaring events, whichare estimated to be brighter than 4 . · − erg cm − s − between 100 GeV and 200 GeV, roughly corresponding to https://fermi.gsfc.nasa.gov/ssc/data/access/lat/FAVA/ Figure 1.
The estimated frequency of VHE transients associatedwith AGN flares of different duration and with GRBs. The eventstaken into account are limited to the ones which are expectedto be bright enough to be detected by CTA South if they werelocated within 20 o from zenith, and they could be observed for1 hr. the limiting flux for a 5 σ detection with CTA South (CTA-S) in 1 hr of observation. This further limitation leads to theselection of 1374 flares that were associated with AGNs in7 . ±
12 per year) and 237 in one thathas been detectable for a few hours (resulting in averagely32 ± At present, the data that we possess on the VHE proper-ties of GRBs are still very scarce, and any attempt to modelthem are subject to large uncertainties, due to the many im-portant free parameters that affect the predicted luminosi-ties and spectra (Galli & Piro 2008; Bernardini et al. 2019).If we are interested in an estimate of the rate of VHE events,we can start from the observation that GRBs detected with
MNRAS , 000–000 (2020)
G. La Mura et al.
Figure 2.
Differential sensitivity to a point-like source for HAWC(cyan continuous line), LHAASO (blue line), and SWGO (yellowline) as compared with the
Fermi -LAT Pass 8 sensitivity (darkgrey line) computed on 1 year of observations. For comparison,the plot also shows different fractions of the Crab Nebula fluxspectrum (short dashed light grey curves), as well as the sensi-tivity achieved by CTA-S in 50 hours of observation (long dashedred line). a maximum photon energy above 100 GeV are brighter to-wards the lower energy limit. As a consequence, we expectthat GRBs with a significant VHE emission should in princi-ple be detected by
Fermi -LAT in the HE domain, if properlypointed.According to the data presented in the second cata-log of LAT detected GRBs (Ajello et al. 2019), in 10 yearsof operation, the LAT has been able to detect 169 GRBswith photons above 100 MeV, while only 15 events were as-sociated with photons detected above 10 GeV. By normal-izing the detection rates with respect to the LAT effectivearea inferred from the P8R3 TRANSIENT V2 instrumentresponse function (IRF) and assuming an isotropic distri-bution of GRBs, we can relate the different detection rateswith the probability that a specific GRB spectrum extendsabove a critical energy. If we denote as N GRB ( E ) the cu-mulative number of events per year that emit photons upto energy E and we assume an isotropic GRB distribution,we can reproduce the observed GRB rate from a power-lawdistribution in the form of: N ( E ) = 291 (cid:18) E
100 MeV (cid:19) − . . (2)Clearly, a simple extrapolation of the observed LAT trendis still a very limited approximation of the real GRB prop-erties, because the reduced sensitivity of the LAT above100 GeV and the extreme rarity of the most energetic pho-tons can lead to a poor sampling of the actual high en-ergy properties. However, the emission of VHE photons re-quires favourable energetic conditions in the source and,since the EBL opacity limits their propagation within a rel-atively small horizon, we can expect that a rate of at least 3 ( ±
2) GRBs per year may be associated with detectableVHE emission (see Fig. 1).
With our study of
Fermi -LAT γ -ray transients, we estimatedthat several hundreds transient events can represent highpriority targets for sub-TeV VHE investigations every year.Their unpredictable nature and the possibility that a signif-icant fraction of triggers issued by lower energy monitoringinstruments may have little or no VHE emission, however,poses the question to explore what instrument performanceis required to identify the most relevant ones. While CTAis obviously expected to provide the best sensitivity in thisenergy range, the performance of monitoring instrumentson short time-scales is still very limited. Fermi -LAT haslow sensitivity to VHE photons, while the largest groundbased facilities of HAWC and LHAASO are located in theNorthern hemisphere and, therefore, do not provide full-skycoverage.VHE particles - either cosmic rays or photons - thatinteract with the atmosphere generate a shower of rela-tivistic secondary products, including charged particles andCherenkov radiation, which develops downward, approxi-mately in the direction of the incoming primary. These show-ers can be tracked by collecting the Cherenkov light andby detecting the secondary particles that reach the ground.Arrays of particle detectors, placed on the ground, but athigh altitude, can infer the direction and the energy of at-mospheric showers, provided that they are able to deter-mine with sufficient accuracy the location, the arrival timeand the energy carried by the secondary charged particlesthat reach them. For this work, we took the yellow curvein Fig. 2, as the reference curve for SWGO (Schoorlemmer2019). This curve was for the most part obtained through astraw-man model which assumed a facility based on waterCherenkov detectors (WCD) covering an area of 80 000 m with an 80% filling factor (Albert et al. 2019). For the lowerpart, below 300 GeV, the most important energy region forthis work, the assumed sensitivity is based on an end-to-endsimulation (performed with Geant4, Agostinelli et al. 2003)of a detector concept which combines a WCD with whitereflective walls and a Resistive Plate Chamber (RPC), toeffectively lower the energy threshold (Assis et al. 2018).The success of this concept relies on the ability to triggeron low-energy secondary photons with the WCD while be-ing able to measure time with a resolution better than 2 ns,crucial for a good geometry reconstruction. The instrumentperformances illustrated in Fig. 2 are computed for a steadypoint source located at a zenith distance of 20 o and underthe assumption that it can be observed 6 hr per day. Forwide FoV instruments, we need to correct this sensitivityaccording to the visibility of different sky areas. Although a wide FoV instrument can track simultaneouslytargets located in different sky regions, the corresponding000
Fermi -LAT γ -ray transients, we estimatedthat several hundreds transient events can represent highpriority targets for sub-TeV VHE investigations every year.Their unpredictable nature and the possibility that a signif-icant fraction of triggers issued by lower energy monitoringinstruments may have little or no VHE emission, however,poses the question to explore what instrument performanceis required to identify the most relevant ones. While CTAis obviously expected to provide the best sensitivity in thisenergy range, the performance of monitoring instrumentson short time-scales is still very limited. Fermi -LAT haslow sensitivity to VHE photons, while the largest groundbased facilities of HAWC and LHAASO are located in theNorthern hemisphere and, therefore, do not provide full-skycoverage.VHE particles - either cosmic rays or photons - thatinteract with the atmosphere generate a shower of rela-tivistic secondary products, including charged particles andCherenkov radiation, which develops downward, approxi-mately in the direction of the incoming primary. These show-ers can be tracked by collecting the Cherenkov light andby detecting the secondary particles that reach the ground.Arrays of particle detectors, placed on the ground, but athigh altitude, can infer the direction and the energy of at-mospheric showers, provided that they are able to deter-mine with sufficient accuracy the location, the arrival timeand the energy carried by the secondary charged particlesthat reach them. For this work, we took the yellow curvein Fig. 2, as the reference curve for SWGO (Schoorlemmer2019). This curve was for the most part obtained through astraw-man model which assumed a facility based on waterCherenkov detectors (WCD) covering an area of 80 000 m with an 80% filling factor (Albert et al. 2019). For the lowerpart, below 300 GeV, the most important energy region forthis work, the assumed sensitivity is based on an end-to-endsimulation (performed with Geant4, Agostinelli et al. 2003)of a detector concept which combines a WCD with whitereflective walls and a Resistive Plate Chamber (RPC), toeffectively lower the energy threshold (Assis et al. 2018).The success of this concept relies on the ability to triggeron low-energy secondary photons with the WCD while be-ing able to measure time with a resolution better than 2 ns,crucial for a good geometry reconstruction. The instrumentperformances illustrated in Fig. 2 are computed for a steadypoint source located at a zenith distance of 20 o and underthe assumption that it can be observed 6 hr per day. Forwide FoV instruments, we need to correct this sensitivityaccording to the visibility of different sky areas. Although a wide FoV instrument can track simultaneouslytargets located in different sky regions, the corresponding000 , 000–000 (2020) onitoring VHE transients Figure 3.
Distribution of VHE flaring blazars, compared with the sky coverage of LHAASO (blue shaded region) and SWGO (orangeshaded region) within a FoV of 50 o from zenith, assuming an SWGO site with latitude 23 o S. The colour shading represents the sensitivitydegradation with respect to a source that culminates at the zenith. The sensitivity comparison in the overlapping region is computed at300 GeV and only the instrument with the best estimated performance is plotted. The red dots represent objects for which
Fermi -LATdetected flares that would be bright enough to be observed by CTA in 1 hr, but are too faint for the computed EAS sensitivities. Thegreen dots are sources whose flares are comparable to the sensitivity of the relevant monitoring instrument, while the stars representobjects whose flares are either already detected by HAWC, or their
Fermi -LAT spectral fit is significantly brighter than the predictedEAS sensitivity limit scaled down to the duration of the flare. Every source can produce more than one flare. sensitivity depends on the zenith distance of every source.Due to computational limitations, the dependence of thesensitivity with the shower inclination was obtained usingthe amount of shower electromagnetic energy at the groundas a proxy. This allowed us to use the shower simulationsgenerated with CORSIKA (Heck et al. 1998), while skip-ping the full detector simulation and the application of theshower reconstruction analyses, as described in detail in As-sis et al. (2018). While this is a crude estimate, we believeit is a conservative approach. On the one hand, the triggerprobability decreases with the increase of the zenith angle,as fewer particles (energy) reach the ground due to the atmo-sphere increasing thickness. On the other hand, the showeris more spread over the array, which eases the geometricreconstruction. This has an impact not only on the astro-physical source position determination but also effectivelyreduces the hadronic background.Since the Earth’s daily rotation induces a transit thatchanges the fraction of time that a source spends at a specificzenith distance θ , depending on the latitude of the observ-ing site and the source’s declination, we derived the fractionof time ∆ t ( θ ) that every point in the sky spends at a givenzenith distance θ , according to its declination. If an instru-ment operates for an observing time ∆ t , under conditionsthat change with time, like θ for a transiting source, thesensitivity of the observation scales as: S (∆ t ) = (cid:34) T (cid:88) i ∆ t i S i ( T ) (cid:35) − / (3)where ∆ t i are the amounts of time during which we can con- sider the instrument to have a regular performance, while T and S i ( T ) represent, respectively, the standard time inter-val, on which the sensitivity is computed, and the sensitivityunder constant observing conditions over such interval.To derive the effective sensitivities that would result bythe combined operation of LHAASO and SWGO observa-tories, we used the corresponding sensitivity curves, com-puted for one year of observations, assuming that SWGOis located at a latitude of 23 o S. We subdivided the sky in2 o wide strips of constant declination, and we computed thefraction of time that a point belonging to each one of thesestrips spends at a zenith distance θ ≤ o . We used the sensi-tivities computed as a function of θ for an energy of 300 GeV(Assis et al. 2018) to obtain the sensitivity degradation re-sulting when the zenith distance increases from θ = 0 o to θ = 50 o in steps of 5 o . Finally, we applied Eq. (3) to extractthe sensitivity that an observation can effectively achieve inthe whole FoV, thus deriving an estimate of the limiting fluxto detect a target within the monitored area. The result ofthis calculation, carried out for the SWGO and LHAASOexperiments, is shown in Fig. 3, together with the distribu-tion of VHE flaring AGNs. This is the latitude of the ALMA site, one of the possible can-didate sites for such experiment.MNRAS , 000–000 (2020)
G. La Mura et al.
Figure 4.
Histograms illustrating the predicted detection rates of VHE transients after 1 yr of observations with SWGO and CTA South.The color bars represent the expected rates at which VHE transients can be detected by SWGO (yellow), CTA-S without a trigger (red),and CTA-S assuming a perfect monitoring program that issues triggers for all VHE transients (pink). The analysis takes into accountthe instrument sensitivities, duty cycles and FoVs and divides the transients in long events, for which the transit over the observing siteis granted (left panel) and short events, where the additional possibility that the transient is in an unobservable portion of the sky istaken into account (right panel). To estimate the CTA serendipitous detection rate, we assume that the area scanned by CTA increaseswith the duration of the flare as result of the combination of different pointings.
Taking as reference the performance of CTA-S, our anal-ysis of the Fermi -LAT monitoring data led to the identifi-cation of 160 blazars and of a fraction of GRBs distributedacross the whole sky, whose light-curves show transient VHEemission that is strong enough for detection in 1 hr of obser-vation. Since blazars are sources of potential VHE activityfor which we know the position in the sky, we can comparetheir emission with the predicted sensitivity of monitoringinstruments that regularly scan the sky, such as LHAASOand SWGO, as it is illustrated in Fig. 3. The shaded areasrepresent the FoV covered within 50 o from zenith, with acolour scale that illustrates the sensitivity degradation to-wards different sky regions, with respect to a target thatculminates at the zenith. Only the instrument with the bestestimated sensitivity is plotted in the overlapping region.If we apply an extrapolation of Eq. (1) to the spectraof the HE flares identified in our sample, taking into ac-count the effects of EBL (Dom´ınguez et al. 2011; Kudoda& Faltenbacher 2017), we can compare the predicted VHEfluxes with the chances that a monitoring instrument hasto detect a specific transient. With the assumed instrumentperformance, it turns out that 21 blazars out of 160, listedin Table 1, showed a flaring activity that is comparable tothe detection threshold of monitoring facilities (green sym-bols). Eight of these sources, marked as stars in Fig. 3 andlisted in bold-face in Table 1, have a predicted spectrumthat is significantly brighter than the limiting flux expectedfor the instrument covering their location in the sky, or, inthe case of MRK 421 and MRK 501, are already detectedby HAWC, while six of them appear in the TeVCat list ofobjects detected in VHE domain. The remaining sources, instead, show an estimated flaring activity that is too faintfor detection by a monitoring instrument. All of these ob-jects, however, remain of potential interest, because mostof them have been associated with multiple flares and theuse of their averaged spectral properties can still smear outthe possible existence of sharp peaks in their light-curve atshorter time-scales.Indeed, the importance to have efficient monitoring in-struments is connected with the chances to detect and trackflaring activity. CTA, with an observing budget of 1500 hrper year per site, corresponding to a duty cycle smaller than20 per cent (Actis et al. 2011), and in the assumption that itcan collect observational data on average for 5 hr per night,covering a FoV with an angular radius of 2 o , has practicallynull chances to be in place, when a flare occurs. Assumingthat CTA follows a scheduled observing plan (i.e., not tak-ing into account Target of Opportunity observations), theprobability to detect a transient in a random position of theobservable sky is proportional to the rate of the transientand to the ratio between the area covered by CTA obser-vations and the total visible sky (approximately 2 sr, below1 TeV). Recalling that our estimate of transient rates arebased on the events which are bright enough to be detectedin 1 hr of observation and assuming that CTA exposure haveno significant overlap above 1 hr, the scanned area increaseswith the duration of the flare, resulting in a higher detectionchance for longer transients. Clearly, the detection probabil-ity can be even higher, if CTA is alerted about the occur-rence of a transient, with accurate positional information. Atpresent, however, the available instruments are not able tofilter which low energy transients will be actually associatedwith VHE emission and following all the possible triggerswould probably result in a high rate of false alarms.Conversely, if we focus our attention to the sky regionthat will be covered by CTA-S, because of its better perfor-mance, we can illustrate the role of a monitoring facility, by MNRAS , 000–000 (2020) onitoring VHE transients Table 1.
List of AGNs that have been detected by
Fermi -LAT with an energy flux larger than 10 − erg cm − s − above 10 GeV andassociated with flaring activity above the sensitivity limit of monitoring facilities. The columns report the name of the AGN, its 3FHLassociation, the sky coordinates (Right Ascension and Declination, J2000), the redshift, the number of associated flares in 7 . TeVCat , and the energy flux obtained from a power-law fit to the 3FHL data. We mark withbold-faced names those whose flares are significantly stronger than the corresponding detection limit (marked with stars in Fig. 3).
Name 3FHL source R.A. Dec. z N flares
TeVCat 3FHL en. flux hh:mm:ss dd:mm:ss 10 − erg cm − s − TXS 0214+083 J0217.1+0836 02 : 17 : 17 .
12 +08 : 37 : 03 .
89 0 .
085 2 N 2 . ± . PKS 0301–243
J0303.4–2407 03 : 03 : 26 . −
24 : 07 : 11 .
42 0 .
260 3 Y 36 . ± . .
37 +18 : 00 : 41 .
58 0 .
416 9 N 5 . ± . . −
36 : 27 : 30 .
85 0 .
055 11 N 5 . ± . . −
44 : 05 : 08 .
94 0 .
892 91 N 37 . ± . . −
26 : 05 : 44 .
64 0 .
414 3 N 11 . ± . PKS 0736+0174
J0739.3+0137 07 : 39 : 18 .
03 +01 : 37 : 04 .
62 0 .
191 33 Y 2 . ± . MRK 421
J1104.4+3812 11 : 04 : 27 .
31 +38 : 12 : 31 .
80 0 .
031 9 Y 437 . ± . .
70 +02 : 03 : 08 .
60 0 .
158 66 N 1 . ± . .
09 +25 : 18 : 07 .
14 0 .
135 18 Y 8 . ± . . −
05 : 47 : 21 .
53 0 .
536 79 Y 18 . ± . PKS 1510–089
J1512.8–0906 15 : 12 : 50 . −
09 : 05 : 59 .
83 0 .
360 187 Y 35 . ± . . −
24 : 22 : 19 .
48 0 .
049 1 Y 20 . ± . MRK 501
J1653.8+3945 16 : 53 : 52 .
22 +39 : 45 : 36 .
61 0 .
033 2 Y 156 . ± . PKS 2155–304
J2158.8–3013 21 : 58 : 52 . −
30 : 13 : 32 .
12 0 .
116 3 Y 132 . ± . PKS 2233–148
J2236.5–1433 22 : 36 : 34 . −
14 : 33 : 22 .
19 0 .
325 17 N 9 . ± . . −
28 : 06 : 39 .
32 0 .
525 6 N 4 . ± . .
75 +16 : 08 : 53 .
56 0 .
859 301 N 32 . ± . J2324.7–4040 23 : 24 : 44 . −
40 : 40 : 49 .
44 0 .
174 1 N 9 . ± . . −
49 : 55 : 40 .
64 0 .
518 76 N 7 . ± . . −
15 : 55 : 07 .
83 0 .
621 34 N 2 . ± . looking at the predicted detection rates illustrated in Fig. 4.If we use the results of our analysis to infer the average rateof transients with detectable VHE emission, applying visi-bility and duty cycle constraints, we can expect that CTAwill be able to detect dozens of transients per year withjust 1 hr of observation, if every VHE event were associatedwith an accurate positional trigger. The number of poten-tially detectable sources is naturally expected to decrease forevents of shorter time-scale and to drop practically at zero,in absence of a trigger. At present, the largest set of mon-itoring data in the HE domain is the one provided by the Fermi -LAT, but even this information can only be accessedafter a time lag of approximately 6 hr, due to data down-linkand processing requirements. As a result, the CTA abilityto detect transients, without a monitoring facility, can beseriously affected. The existence of a monitoring instrumentwith the quoted SWGO performance would recover a sig-nificant fraction of the longest duration triggers and, pos-sibly, achieve even better results on the shortest ones. Thisstrategy, therefore, would represent a viable system to trig-ger CTA observations on the most relevant VHE transientevents and also to collect independent data, extending ourability to monitor VHE activity towards sources that lie at z ≈ . The study of VHE transient phenomena is growing in impor-tance, now that flaring activity in blazars has been shown totake part in the acceleration of ultra-energetic particles andthat VHE photons have been firmly detected from GRBs. In this work we analyzed the results of the γ -ray monitor-ing campaign, carried out by the Fermi -LAT instrument, toidentify the sources of flaring activity that are most likelyassociated with transient VHE emission. We estimated theexpected fluxes of these VHE transients and we comparedthem with the predicted performance of CTA-S and of mon-itoring facilities like LHAASO and SWGO. We pointed outthat a fraction of the selected sources is expected to havestrong enough flaring activity to be tracked by EAS arraysthat continuously scan wide portions of the sky. We showthat instruments with optimized sub-TeV capabilities areable to provide relevant information on a substantial fractionof events that CTA could only detect if properly triggered.At present, most of the triggers are provided by satellites atrelatively low energy, while the high energy information onlycomes with some delay (for instance, it takes approximately6 hr before LAT data can be processed). EAS arrays, onthe other hand, can directly trigger on the VHE band, withshorter response times. Given that EAS arrays can coververy large instrumented surfaces at a relatively low cost andthat they have the further advantages to be maintainableand upgradable, with respect to a space mission, we pointout that obtaining good sensitivities to the sub-TeV rangeof these detectors will bring to major improvements in therole of VHE transients as cosmic messengers.
DATA AVAILABILITY
There are no new data associated with this article.
MNRAS , 000–000 (2020)
G. La Mura et al.
ACKNOWLEDGEMENTS
The authors would like to thank F. Longo, N. Omodei andthe referee for useful discussion and suggestions.This work was partly performed under projectPTDC/FIS-PAR/29158/2017, Funda¸c˜ao para a Ciˆencia eTecnologia. RC is grateful for the financial support by OE -Portugal, FCT, I. P., under DL57/2016/cP1330/cT0002.The
Fermi
REFERENCES
Abbott B. P., Abbott R., Abbott T. D., et al. 2017a, ApJ, 848,L12Abbott B. P., Abbott R., Abbott T. D., et al. 2017b, ApJ, 848,L13Abdalla H., Adam R., Aharonian F., et al. 2019, Nature, 575, 464Abdollahi S., Ackermann M., Ajello M., et al. 2017, ApJ, 846, 34Abeysekara A. U., Archer A., Benbow W., et al. 2018, ApJ, 861,L20Actis M., Agnetta G., Aharonian F., et al. 2011, ExperimentalAstronomy, 32, 193Agostinelli S., et al., 2003, Nucl. Instrum. Meth., A506, 250Aharonian F., Akhperjanian A. G., Bazer-Bachi A. R., et al. 2006,A&A, 457, 899Ajello M., Atwood W. B., Baldini L., et al. 2017, ApJS, 232, 18Ajello M., Arimoto M., Axelsson M., et al. 2019, ApJ, 878, 52Albert A., Alfaro R., Ashkar H., et al. 2019, arXiv e-prints, p.arXiv:1902.08429Aleksi´c J., Ansoldi S., Antonelli L. A., et al. et al. et al. 2016,Astroparticle Physics, 72, 61Anderhub H., Backes M., Biland A., et al. 2013, Journal of In-strumentation, 8, P06008Assis P., Barres de Almeida U., Blanco A., et al. et al. et al. 2018,Astroparticle Physics, 99, 34Atwood W. B., Abdo A. A., Ackermann M., et al. 2009, ApJ,697, 1071Bernardini M. G., Bissaldi E., Bosnjak Z., Carosi A., D’AvanzoP., Di Girolamo T., Inoue S., Gasparetto T., 2019, in 36thInternational Cosmic Ray Conference (ICRC2019). p. 598 Blanchard P. K., et al., 2017, ApJ, 848, L22CTA Consortium 2019, Science with the Cherenkov TelescopeArray, doi:10.1142/10986.Cowperthwaite P. S., Berger E., Villar V. A., et al. 2017, ApJ,848, L17DeYoung T., 2012, Nuclear Instruments and Methods in PhysicsResearch A, 692, 72Desai A., Helgason K., Ajello M., Paliya V., Dom´ınguez A., FinkeJ., Hartmann D., 2019, ApJ, 874, L7Di Sciascio G., LHAASO Collaboration 2016, Nuclear and Parti-cle Physics Proceedings, 279, 166Dom´ınguez A., et al., 2011, MNRAS, 410, 2556Galli A., Piro L., 2008, A&A, 489, 1073H. E. S. S. Collaboration 2017, A&A, 600, A89Heck D., Schatz G., Thouw T., Knapp J., Capdevielle J. N., 1998,FZKA-6019Holder J., Acciari V. A., Aliu E., et al. 2008, in AharonianF. A., Hofmann W., Rieger F., eds, American Institute ofPhysics Conference Series Vol. 1085, American Institute ofPhysics Conference Series. pp 657–660 ( arXiv:0810.0474 ),doi:10.1063/1.3076760IceCube Collaboration 2018, Science, 361, eaat1378Kudoda A. M., Faltenbacher A., 2017, MNRAS, 467, 2896MAGIC Collaboration 2019, Nature, 575, 455Meegan C., et al., 2009, ApJ, 702, 791Mirzoyan R., 2017, The Astronomer’s Telegram, 10817Mirzoyan R., 2019, The Astronomer’s Telegram, 12390Paiano S., Falomo R., Treves A., Scarpa R., 2018, ApJ, 854, L32Prince R., Majumdar P., Gupta N., 2017, ApJ, 844, 62Schoorlemmer H., 2019, in 36th International Cosmic Ray Con-ference (ICRC2019). p. 785 ( arXiv:1908.08858 )The Fermi-LAT collaboration 2019, Fermi Large Area TelescopeFourth Source Catalog ( arXiv:1902.10045 )Wakely S. P., Horan D., 2008, International Cosmic Ray Confer-ence, 3, 1341Zacharias M., et al., 2019, Galaxies, 7, 41MNRAS000