AGILE Observations of the Gravitational Wave Event GW150914
M. Tavani, C. Pittori, F. Verrecchia, A. Bulgarelli, A. Giuliani, I. Donnarumma, A. Argan, A. Trois, F. Lucarelli, M. Marisaldi, E. Del Monte, Y. Evangelista, V. Fioretti, A. Zoli, G. Piano, P. Munar-Adrover, L.A. Antonelli, G. Barbiellini, P. Caraveo, P.W. Cattaneo, E. Costa, M. Feroci, A. Ferrari, F. Longo, S. Mereghetti, G. Minervini, A. Morselli, L. Pacciani, A. Pellizzoni, P. Picozza, M. Pilia, A. Rappoldi, S. Sabatini, S. Vercellone, V. Vittorini, P. Giommi, S. Colafrancesco, M. Cardillo
SSubmitted to the
Astrophysical Journal Letters , April 1, 2016.
AGILE Observations of the Gravitational Wave EventGW150914
M. Tavani , , , C. Pittori , , F. Verrecchia , , A. Bulgarelli , A. Giuliani ,I. Donnarumma , A. Argan , A. Trois , F. Lucarelli , , M. Marisaldi ,E. Del Monte , Y. Evangelista , V. Fioretti , A. Zoli , G. Piano ,P. Munar-Adrover , L.A. Antonelli , , G. Barbiellini , P. Caraveo ,P.W. Cattaneo , E. Costa , M. Feroci , A. Ferrari , F. Longo ,S. Mereghetti , G. Minervini , A. Morselli , L. Pacciani , A. Pellizzoni ,P. Picozza , M. Pilia , A. Rappoldi , S. Sabatini , S. Vercellone ,V. Vittorini , P. Giommi , S. Colafrancesco , M. Cardillo . INAF-IAPS, via del Fosso del Cavaliere 100, I-00133 Roma, Italy Dip. di Fisica, Univ.di Roma “Tor Vergata”, via della Ricerca Scientifica 1, I-00133 Roma, Italy Gran Sasso Science Institute, viale Francesco Crispi 7, I-67100 L’Aquila, Italy ASI Science Data Center (ASDC), Via del Politecnico, I-00133 Roma, Italy INAF-OAR, via Frascati 33, I-00078 Monte Porzio Catone (Roma), Italy INAF-IASF-Bologna, Via Gobetti 101, I-40129 Bologna, Italy INAF-IASF Milano, via E.Bassini 15, I-20133 Milano, Italy INAF, Osservatorio Astronomico di Cagliari, Poggio dei Pini, strada 54, I-09012 Capoterra, Italy Dip. di Fisica, Universita’ di Trieste and INFN, Via Valerio 2, I-34127 Trieste, Italy INFN-Pavia, Via Bassi 6, I-27100 Pavia, Italy CIFS, c/o Physics Department, University of Turin, via P. Giuria 1, I-10125, Torino, Italy Dip. di Matematica, Univ.di Roma “Tor Vergata”, via della Ricerca Scientifica 1, I-00133 Roma, Italy INFN Roma Tor Vergata, via della Ricerca Scientifica 1, I-00133 Roma, Italy INAF/IASF-Palermo, Via U.La Malfa 153, I-90146 Palermo, Italy University of Witwatersrand, Johannesburg, South Africa INAF Osservatorio Astronomico di Arcetri, Largo Enrico Fermi, 5, I-50125 Firenze, Italy a r X i v : . [ a s t r o - ph . H E ] A p r ABSTRACT
We report the results of an extensive search in the AGILE data for a gamma-ray counterpart of the LIGO gravitational wave event GW150914. Currentlyin spinning mode, AGILE has the potential of covering with its gamma-ray in-strument 80% of the sky more than 100 times a day. It turns out that AGILEcame within a minute from the event time of observing the accessible GW150914localization region. Interestingly, the gamma-ray detector exposed ∼
65 % ofthis region during the 100 s time intervals centered at -100 s and +300 s from theevent time. We determine a 2 σ flux upper limit in the band 50 MeV - 10 GeV, U L = 1 . × − erg cm − s − obtained ∼
300 s after the event. The timing of thismeasurement is the fastest ever obtained for GW150914, and significantly con-strains the electromagnetic emission of a possible high-energy counterpart. Wealso carried out a search for a gamma-ray precursor and delayed emission overtimescales ranging from minutes to days: in particular, we obtained an optimalexposure during the interval − / −
30 s. In all these observations, we do notdetect a significant signal associated with GW150914. We do not reveal the weaktransient source reported by
Fermi -GBM 0.4 s after the event time. However,even though a gamma-ray counterpart of the GW150914 event was not detected,the prospects for future AGILE observations of gravitational wave sources aredecidedly promising.
Subject headings: gravitational waves, gamma rays: general.
1. Introduction
The recent discovery of gravitational waves (hereafter, GW) by the LIGO experimentimpulsively emitted from the source GW150914 started a new era in astronomy (Abbott etal., 2016a,b,c,d; hereafter A16a,b,c,d). The detection occurred at the beginning of an acqui-sition run of the LIGO experiment in an enhanced configuration (A16a). The LIGO-VIRGOdetectors are expected to operate soon at even improved sensitivity, and the potential for alarge number of detections of gravitational phenomena will shape future ground-based andspace research. The characteristics of gravitational phenomena emitting detectable gravita-tional waves are those of the most extreme and energetic events of our universe. The energyradiated under the form of gravitational waves as inferred from GW150914 is about 3 M (cid:12) ,a huge value (A16a). This energy was emitted during a few hundreds of milliseconds. It isclear that this type of gravitational phenomena opens the way to study with unprecedented 3 –synergy the interplay between gravitation, the astrophysical context, and the quantum prop-erties of fields and matter. In general, GW-emitting final stages of coalescences involvingcompact stars (neutron stars (NSs) and black holes (BHs)) are the most likely candidates forLIGO-VIRGO events. X-ray and gamma-ray counterparts are expected from different typesof coalescing compact star systems. The quest for electromagnetic counterparts of extremegravitational events is now open.The characteristics of the binary system associated with GW150914 are somewhat sur-prising given the current observations and understanding of the evolutionary processes lead-ing to the formation of black holes. The event results from the coalescence of two blackholes of relatively large masses (near 30 M (cid:12) ) in an unknown stellar environment (A16d).Where and how such events can be produced is an open and interesting question (A16d):the physical conditions of the coalescing compact objects are far from being understood.Even though purely gravity-systems such as BH-BH binaries are not anticipated to emitdetectable electromagnetic (e.m.) radiation, nevertheless such a radiation can be emittedbefore, during and after coalescence depending on the physical conditions of the system. Itis then of great interest to explore this possibility and search for e.m. counterparts of GWevents.The AGILE satellite, today at its ninth year of operations in orbit, is observing thegamma-ray sky with excellent monitoring capabilities in the gamma-ray range 30 MeV -30 GeV. The satellite is currently in spinning mode covering a large fraction of the skywith a gamma-ray sensitivity to transient emission that can reach flux levels near F =(1 − × − erg cm − s − for ∼
100 s integrations. This timescale is typical of passes of theimaging gamma-ray instrument (Field of View (FoV) of 2.5 sr) over exposed sky regions inspinning mode. Each accessible sky region is exposed more than 100 times a day with 100 sintegrations each. In addition, non-imaging detectors such as the AGILE Mini-Calorimeterand anticoincidence system (routinely employed for the detection of GRBs and TGFs) canbe used. AGILE has therefore unique capabilities for the search of gamma-ray counterpartsof GW sources.In this Letter we present the results of an extensive search in AGILE data for a high-energy counterpart of GW150914 that we performed retroactively once the event was madepublic . The paper is organized as follows. In Sect. 2 we state the case of AGILE as an ideal Depending on Earth occultations and SAA passages. The AGILE Team was not part of the multifrequency follow-up collaboration with the LIGO-VIRGOteam at the time of the GW150914 detection and subsequent investigations; we learned of the GW150914event on February 11, 2016.
2. AGILE’s Capability for the Search of Gamma-Ray Counterparts of GWSources
The AGILE satellite, launched on April 23rd, 2007, is orbiting the Earth in a near equa-torial orbit of current altitude ∼
500 km. The instrument consists of an imaging gamma-raySilicon Tracker (sensitive in the energy range 30 MeV - 30 GeV), Super-AGILE (currentlyworking in ratemeter mode in the energy range 20-60 keV due to current temporary teleme-try limitations), a Mini-Calorimeter (MCAL, working in the range 0.4 - 100 MeV) and ananticoincidence (AC) system (for a summary of the AGILE mission features, see Tavani etal. 2009). The combination of Tracker, MCAL and AC working as a gamma-ray imagerconstitutes the AGILE-GRID. The instrument is capable of detecting gamma-ray transientsand GRB-like phenomena for timescales ranging from sub-milliseconds to tens-hundreds ofseconds. In addition to the hundreds of GRBs detected by MCAL and Super-AGILE, severalprominent GRBs were detected by the gamma-ray imager since the beginning of operations(GRB080514B, Giuliani et al 2008; GRB090401B, Moretti et al. 2009; GRB090510, Giu-liani et al. 2010; GRB100724B, Del Monte et al. 2011; GRB130327B, Longo et al. 2013;GRB130427A, Verrecchia et al. 2013; GRB131108A, Giuliani et al. 2013 and Giuliani etal. 2015). Furthermore, AGILE so far detected about 1,000 Terrestrial Gamma-Ray Flashes(TGFs) with durations ranging from hundreds to thousands of microseconds (Marisaldi etal. 2014, Tavani et al. 2011). A special sub-millisecond search for transient events detectedby MCAL is operational on board (Tavani et al., 2009).The characteristics that make AGILE in spinning mode an important instrument forfollow-up observations of large GW source localization regions are: (1) a very large FoV ofthe GRID (2.5 sr); (2) an accessible region of 80% of the whole sky that can be exposed every7 minutes (see Fig. 1); (3) 100-150 useful passes every day for any region in the accessiblesky ; (4) a gamma-ray exposure of ∼ ∼ − erg cm − s − above 30 MeV for typical single-pass of unocculted sky regions; (5) sub-millisecond trigger for very fast events detectable byMCAL in the range 0.4-100 MeV; (6) hard X-ray (20 −
60 keV) triggers of GRB-like events The total number of 7 minute rotations is ∼ ∼
3. AGILE Observations of GW150914
The GW150914 event occurred at time T = 09:50:45 UTC on Sept. 14, 2015 (A16a).At that time AGILE was scanning the sky in spinning mode with the Earth only partiallyocculting the GRID FoV. Fig. 2 shows the gamma-ray exposure above 50 MeV for thewhole satellite 7 minute revolution that includes the GW150914 event time. As anticipated,the Earth only marginally covers the GW150914 localization region of the most accurateGW150914 localization map (LALInference, Abbott et al. 2016e, Veitch et al. 2015): mostof the localization region is not occulted, and therefore available for AGILE exposure. Weperformed a search for: (1) the prompt event involving the GRID, MCAL, AC and Super-AGILE ratemeters (both inside and outside the GRID FoV); (2) delayed emission on multipletimescales involving the GRID; (3) precursor emission involving the GRID. Figure 3 shows the gamma-ray exposure of a typical point inside the GW150914 local-ization region for the two satellite rotations of interest closer to the prompt event. AGILEhad GRID exposure of a substantial fraction (65%) of the GW150914 localization region afew tens of seconds before T , but not at the event time (see also Fig. 4). Had the instrumentobtained an exposure of the field equal to that of a few tens of seconds earlier, AGILE couldhave obtained a gamma-ray sensitivity near 10 − erg cm − s − for a few second integration inthe range 50 MeV - 10 GeV. A large fraction of the GW localization region was not occultedby the Earth and a strong X/gamma-ray signal, if any, could have been detected by theAGILE non-imaging detectors.The MCAL did not trigger events within an interval covering -100 / +100 seconds from T . Within -200 / +200 seconds we found four triggers, all of them on the sub-millisecondor ∼ E >
50 MeV gamma-ray exposure maps (in cm s sr, using a 0 . ◦ pixel size) as the satellite rotates in spinning mode scanning 80% of the sky in about 7minutes. The AGILE-GRID FoV radius is assumed to be 70 ◦ . In this case, the Earth isocculting the FoV in the Northern Galactic hemisphere. The exclusion region for the albedophotons is 80 ◦ from the Earth center. This sequence applies to the satellite rotation thatincludes the prompt event time of GW150914 (localization region marked by the purplecontour, LALinference 90% contour level, A16e). 7 –Fig. 2.— Hammer-Aitoff projection in Galactic coordinates. Top panel:
AGILE cumulativegamma-ray exposure above 50 MeV (in cm s sr, using a 0 . ◦ pixel size) during the satellite 7minute rotation that includes the GW150914 event time. Bottom panel:
Gamma-ray photonsabove 50 MeV detected during the satellite 7 minute rotation. The GW150914 localizationregion is marked by the area inside the purple line (LALinference 90% contour level, A16e). 8 –not detect the transient reported by the
Fermi
GBM team (Connaughton et al., 2016) at T + 0 . s . For the reported spectral shape (best-fit single power-law with index − . +0 . − . )we estimate a minimum detectable MCAL fluence at the standard trigger threshold (fivestandard deviations above background) of ∼ . × − erg cm − in the 400 − ◦ off axis. This is only 13% larger than the fluencereported for the Fermi -GBM event (2 . +1 . − . × − erg cm − in the 10 − σ fluence upper limit of 2 . × − erg cm − . We remind here that the interestingshort GRB090510 (Giuliani et al. 2010, see discussion below) is detected by the SA rate-memeters with a significance level of ∼ . σ although it occurred outside the 1 sr SA fieldof view. Also a search of a signal in AC data does not produce a significant detected flux.AGILE was optimally positioned in the GW150914 localization region at interesting timeintervals preceding and following the prompt event. The most interesting time intervals areduring the time intervals ∆ T − = − ±
50 s, and ∆ T +1 = +333 ±
50 s, taking T = 09:50:45UTC on Sept. 14, 2015 as a time-zero reference. In the following, we focus on the analysisof the gamma-ray data searching for precursor and delayed gamma-ray emission. Table 1shows the AGILE-GRID passes over the GW150914 localization region and the results ofour analysis in search of transient gamma-ray emission in that region. The AGILE-GRID exposed a good fraction (75%) of the GW150914 localization regionwithin 250 s from the prompt event. As shown in Figs. 3 and 5, important information canbe obtained during the first useful pass, ∆ T +1 . Considering the local photon backgroundand exposure, a search in the localization region for a transient gamma-ray source producesthe 2 σ UL for emission in the range 50 MeV - 10 GeV: U L = 1 . × − erg cm − s − . Thisupper limit is significant in the context of a possible gamma-ray signal from a cosmic eventassociated with compact object coalescence. Figure 6 shows the upper limit obtained byAGILE in the context of the gamma-ray lightcurve expected from the short GRB090510 re-positioned at redshift z = 0 .
09 of GW150914. This short GRB, which shows several featuresexpected from compact object coalescences possibly emitting GWs (e.g., Berger 2014), wasdetected by AGILE and
Fermi -LAT with a significant delayed emission above 30 MeV lastingup to 100 s with a hard spectral component (Giuliani et al. 2010, Abdo et al. 2009, and For a hard spectrum similar to the short GRB delayed emission of GRB090510 discussed below. T of GW150914. The GW150914 event time is marked by a dotted blueline. The black curve is obtained for a 7 ◦ × ◦ field centered at Galactic coordinates( l, b ) = (282 . , − . ◦ radius circle centered at Galacticcoordinates ( l, b ) = (283 . , − . − s − ) during the 4 s interval that includes the GW150914event time. The AGILE-GRID FoV is taken to be 70 ◦ . AGILE just missed coveringthe GW150914 localization region during the prompt event with useful sensitivity. TheGW150914 localization region is marked by the purple contour (LALinference 90% contourlevel, A16e). 11 – Table 1: Analysis of individual passes over the GW150914 localization region
Interval Central Duration 2 σ UL (*) Commentsnumber time bin (**) (s) (10 − erg cm − s − )-13 -5203 100 2.7 88% of error box not-occulted-12 -4779 100 – affected by SAA-11 -4355 100 – affected by SAA-10 -3931 100 – affected by SAA-9 -3507 100 – affected by SAA-8 -3083 100 2.3 93% of error box not-occulted-7 -2663 100 4.5 78% of error box not-occulted-6 -2235 100 1.5 68% of error box not-occulted-5 -1807 100 1.5 65% of error box not-occulted-4 -1379 100 1.5 20% of error box not-occulted-3 -951 100 1.0 48% of error box not-occulted-2 -523 100 1.0 56% of error box not-occulted-1 -95 100 1.5 65% of error box not-occulted+1 +333 100 1.9 75% of error box not-occulted(*) Flux upper limit obtained for emission in the range 50 MeV - 10 GeV and for a spectrumsimilar to the delayed gamma-ray emission of the short GRB090510.(**) Time calculated from T = 09 : 50 : 45 UTC on Sept. 14, 2015, the event time ofGW150914. 12 – Table 2: Long-integration time analysis of the GW150914 localization region
Interval Duration 2 σ UL (*) Commentsname (10 − erg cm − s − )-3d 3 days 0.3-2d 2 days 0.5-1d 1 day 0.7-12h 12 hours 0.8-6h 6 hours 2.5-3h 3 hours 3.5+3h 3 hours – telemetry interruption (**)+6h 6 hours 3.5 with telemetry interruption (**)+12h 12 hours 1.8 with telemetry interruption (**)+1d 1 day 1.1 with telemetry interruption (**)+2d 2 days 0.9 with telemetry interruption (**)+3d 3 days 0.7 with telemetry interruption (**)+5d 5 days 0.4 with telemetry interruption (**)(*) Gamma-ray upper limits (ULs) for photons in the range 100 MeV - 10 GeV and for a E − spectrum obtained in the best exposed regions of the GW150914 localization region. Weestablished a range of ULs within a factor of 2 inside the exposed localization region. TheULs were obtained taking into account the effective exposure that has an overall efficiencynear 70% due to Earth occultation and SAA passages.(**) Telemetry interruption from UT 10:00 to UT 13:00 of Sept. 14, 2015.Ackermann etal. 2010). Thus AGILE data uniquely determine a gamma-ray upper limitwithin 250-350 s from the GW150914 event.Due to a temporary telemetry interruption caused by ground station operations inMalindi (Kenya), satellite data for two consecutive orbits following the interval ∆ T +1 arenot available. The next useful pass over the GW150914 localization region is about 3 hoursafter ∆ T +1 , at T = 10 ,
800 s after T .We carried out a long-timescale search for transient gamma-ray emission during thehours immediately following the prompt event. No significant gamma-ray emission in theGW150914 localization region was detected during individual passes 3-4 hours after theGW150914 event.We also performed a search on longer timescales up to several days after the event.Table 2 summarizes our results and upper limits for these long integrations. 13 – An interesting possibility arises from the situation of two approaching massive BHs ina shrinking orbit. Even though current theoretical scenarios for BH-BH coalescences donot envision gamma-ray emission preceding or following the final event (e.g., Baumgarte &Shapiro 2011), a residual gaseous and/or plasma environment can in principle induce e.m.radiation during the approaching phase by non-thermal processes. It is in any case relevantdetermining if gamma rays were emitted before the coalescence itself.As reported in Tables 1 and 2, we carried out a search for a gamma-ray precursor over alarge dynamic time range. The most significant observations are the available AGILE-GRIDpasses preceding the prompt event. Table 1 and Fig. 5 show the sequence of passes. Oneof the most interesting passes is ∆ T − covering the time interval − ±
50 s from T . Thegamma-ray 2 σ UL in this case is comparable with what obtained for the ∆ T +1 interval, thatis U L = 1 . × − erg cm − s − . As Fig. 5 shows, for the successive passes (retrograde intime) -2 and -3, the Earth progressively enters in the exposed GW150914 localization region.Passes -4 and -5 have a marginal exposure with a relatively high background because of theEarth albedo gamma-ray photons. The relevant region is then better exposed for intervals-6, -7, and -8. Interval -8 is one of the best cases, with more than 90% of the GW150914localization region region well sampled. Unfortunately, the successive intervals -9, -10, -11,-12 are affected by a particularly deep passage in the SAA and no useful data are availablefor these intervals. Pass no. -13 is very similar to the pass +1, and occurs after one orbit( ∼
95 minutes).We also extended retrogradely in time our search up to 3 days before T . No significantgamma-ray emission was detected during any time interval reported in Tables 1 and 2.
4. Discussion
AGILE observed the field containing GW150914 with very good coverage and signifi-cant gamma-ray exposure within tens-hundreds of minutes before and after the event. TheAGILE-GRID missed the coverage of the prompt event in its field of view, but could de-termine important limits immediately before and after the event. In light of the broadcampaign of follow-up observations of the GW150914 localization region ranging from radioto gamma-rays reported in Abbott et al. 2016e, the AGILE observations are significant in Usually, passages into the SAA affect 1-2 rotations. In the period near T of GW150914 , the SAApassage was “deep” and affected 4 satellite rotations.
14 –Fig. 5.— Sequence of (
E >
50 MeV) maps in Galactic coordinates showing the AGILE-GRIDpasses with the best sensitivity over the GW150914 localization region obtained duringthe period (- 5,303 s, + 433 s) with respect to T . The color maps show the gamma-ray flux 2 σ upper limits in the range 50 MeV - 10 GeV with the most stringent valuesbeing U L = (1 − × − erg cm − s − . The sequence shows 14 maps for all the 1-orbitpasses of Table 1, corresponding to the 100 s interval numbers (from top left to bottomright): − , − , − , − , − , − , − , − , − , − , − , − , − , +1. Note that the passes − , − , − , − − erg cm − s − in the range 50 MeV - 10 GeV obtained in the interval 250-350 s after T . Itis interesting to compare this UL with other observations obtained by imaging X-ray/gamma-ray space instruments with FoVs larger that 1 sr. The BAT instrument on board of the Swift satellite could not observe the GW150914 field because it occurred outside its field of view(Evans et al. 2016). The same applies to
Fermi -LAT that could not cover the GW150914localization region at the moment of the prompt event (
Fermi -LAT Collaboration, 2016).Compared with the
Fermi -LAT gamma-ray UL that was obtained more than 70 minutesafter the event (
Fermi -LAT Collab., 2016), the AGILE-GRID observation at ∆ T +1 providesa more stringent constraint to any delayed emission above 50 MeV shortly after the event.We note that “delayed” gamma-ray emission is sometimes detected from both long andshort GRBs (e.g., Giuliani et al. 2010, Ackermann et al. 2010, De Pasquale et al. 2010). Inparticular, the case of the gamma-ray bright short GRB 090510 is relevant to our purpose be-cause of its characteristics and possible association with a compact star coalescence involvingGW emission. The event, localized through its afterglow at z = 0 .
9, showed a first quasi-thermal interval lasting about 300 ms with a spectrum peaking at a few MeV, and a second“gamma-ray afterglow” phase lasting hundreds of seconds detected above 100 MeV with ahighly non-thermal spectrum (Giuliani et al. 2010, Ackermann et al. 2010). The possibilitythen of detected delayed gamma-rays from GRBs is of great relevance for detectors such asthe AGILE-GRID. As shown in Fig. 6 the AGILE UL obtained 300 seconds after the GWevent is just below the gamma-ray emission expected from a short GRB090510-like eventre-located at the distance of z = 0 .
1. Our data are close to excluding a delayed gamma-rayafterglow of the type of GRB090510. We notice that a short GRB of this kind positioned at400 Mpc would have been detectable by the AGILE non-imaging detectors (AC, MCAL andSuper-AGILE ratemeters) as well as by
Fermi -GBM and INTEGRAL/SPI-ACS if the eventhad the same hard X-ray spectrum as GRB090510. We checked that the current sensitivityof the AGILE AC system is about a factor of 5-10 better than the flux detected in the caseof the short GRB 090510, implying that a signal 10 times weaker than that associated withGRB090510 would have been detected.Also of interest is the limit obtained 50-100 s before the coalescence. For a total massnear 60 M (cid:12) , we are sampling the radiative environment when the orbital distance a is ∼ (0 . − R (cid:12) . We are excluding precursor gamma-ray activity at the distance comparable withthe solar size during the compact object approach. If dynamically formed in a dense stellarenvironment, the BH-BH binary associated to GW150914 might carry a gaseous remnantwith it as a product of the 3-body encounter that formed the binary. This gaseous component 16 –subject to the gravitational influence of the approaching massive BHs can settle into a short-lived disk that might produce e.m. radiation by thermal and non-thermal processes. Theobserved flux UL translates into an upper limit to the (isotropically) radiated gamma-rayluminosity, L γ < × erg s − , assuming the GW150914 luminosity distance of ∼ E γ ∼ . × − M (cid:12) .AGILE does not detect the very weak event reported by Fermi -GBM about 0.4 s after T (Connaughton etal., 2016). The MCAL did not register any event above the triggerthreshold in the energy range 0.4-100 MeV, and the AC and Super-AGILE detectors didnot register enhancements of their countrates during the GW150914 prompt event (in theranges 80-200 keV and 20-60 keV, respectively). Considering the faintness and spectrum ofthe Fermi -GBM event this lack of detection is not surprising.
5. Future Observations
In the near future, we expect a relatively large number of aLIGO and aLIGO-VIRGOdetections of GW sources. As the GW sensitivity and event positioning is improving, thecapability to alert the community within timescales progressively short is to be envisioned.With aLIGO-aVIRGO being capable of detecting NS-NS and NS-BH coalescing systems, thelikelihood of detecting electromagnetic emission from these events will increase substantially.We showed in this paper that AGILE can effectively observe relatively large sky regionsassociated with GW events with high efficiency. As demonstrated in the case of GW150914,AGILE might obtain significant results if the event occurs in the accessible sky region whichdepends on the sky coverage per 7 minute rotation ( ∼ .
8) and Earth occultation ( ∼ . ∼ . × . ’ .
5. In this case, AGILE can obtain imaging gamma-raydata within a timescale ranging from seconds to 300-400 s. This is exactly the case thatoccurred for GW150914. The GRID FoV just missed the prompt event, and had exposurefor most of the localization region only ∼
300 s after T . As shown in Sect. 2, this is theworst case that could have happened given the conditions. Nevertheless, the upper limitobtained for this “late” exposure provides the fastest data in constraining the GW150914e.m. emission. This latter event was also not confirmed by the INTEGRAL SPI-ACS observation covering theGW150914 prompt event (Savchenko et al. 2016). The SPI-ACS upper limit is important since it is obtainedby another satellite not occulted by the Earth at the time of T . The analysis of the SPI-ACS instrument isreported in the interval from −
30 s to +30 s from T .
17 –Fig. 6.— The AGILE (blue circles) and
Fermi -LAT (black squares) gamma-ray lightcurvesof the short GRB090510 (originally at z = 0 .
9) scaled in flux and time corrected as if itoriginated at the GW150914 luminosity distance of 400 Mpc ( z = 0 . Fermi -LAT spectral dataare from Fermi-LAT Collaboration (2016). The AGILE upper limit to gamma-ray emissionabove 100 MeV from the 65% of the GW150914 localization region during the time interval∆ T +1 is marked in red. The Fermi -LAT upper limit for GW150914 , obtained in the interval4,442-4,867 s after the event, is marked in black color (
Fermi -LAT Collaboration, 2016). Thegreen dotted curve shows the estimated AGILE gamma-ray UL derived by extrapolating theUL near 300 s back to 1 s. 18 –The probability of catching the “prompt” event in the GRID FoV is about 0 . ∼ . × . .
1. This is a relatively large probability compared to other satellites or groundinstruments.Further improvement of the efficiency and speed of the AGILE data processing is atask for the near future. The AGILE-GRID data management system allows to obtainresults within ∼ REFERENCES
Abbott B.P. et al., 2016a, Phys. Rev. Letters, 116, 061102.Abbott B.P. et al., 2016b, Observing gravitational-wave transient GW150914 with minimalassumptions [LIGO-P1500229].Abbott B.P. et al., 2016c, GW150914: First results from the search for binary black holecoalescence with Advanced LIGO [LIGO-P1500269-v12].Abbott B.P. et al., 2016d, Properties of the binary black hole merger GW150914 [LIGO-P1500218].Abbott B.P. et al., 2016e, preprint, Localization and broadband follow-up of thegravitational-wave transient GW150914, ApJL, submitted [LIGO-P1500227-v9].Abdo et al., 2009, Nature, 462, 331.Ackermann M., et al., 2010, ApJ, 716, 1178.Baumgarte T.W. & Shapiro S., 2011, Physics Today, October 2011, 32.Berger E., 2014, ARA&A, 52, 43.Bulgarelli A., et al., 2014, ApJ, 781, 19.Connaughton V., et al, 2016, ArXiv e-prints, arXiv:1602.03920.Del Monte E., et al. 2011, A&A, 535, A120.De Pasquale M., et al., 2010, ApJ, 709, L146.Evans P. A., Kennea J. A., Barthelmy S. D., et al. 2016, arXiv:1602.03868.
Fermi -LAT Collab., 2016, Fermi-LAT Observations of the LIGO Event GW150914, ApJ,submitted [arXiv:1602.04488v2].Giuliani A., et al 2008, A&A, 491, L25-L28.Giuliani A., et al., 2010, ApJ, 708, L84.Giuliani A., et al. 2013, GCN 15479.Giuliani A., et al. 2015, submitted to A&A.Longo F., et al. 2013, GCN 14344. 20 –Marisaldi M., et al., 2014, Journal of Geophys. Res., 119, 1337.Moretti E., et al. 2009, GCN 9069.Pittori C., et al., 2013, NuPhS, 239, 104.Savchenko, V. et al., 2016, arXiv:1602.04180v1.Tavani M. et al., 2009, A&A, 502, 995.Tavani M., et al., Phys. Rev. Letters, 2011, 106, 8501.Verrecchia F., et al. 2013, GCN 14515.Veitch J., Raymond V., Farr, B. et al. 2015, Phys. Rev. D, 91, 042003.