Hunting for the gamma-ray counterparts to binary black hole mergers

January 2022

Previous | Index | Next

Gravitational Waves (GWs) are produced by the coalescence of compact objects like neutron stars and black holes. These events are in no doubt in the list of the most cataclysmic events in the Universe. GW events are also promising candidates for producing very high energy (VHE) emission through particle acceleration processes. This is why H.E.S.S. dedicates a large amount of observation time to the follow-up of GWs.

Fig. 1: Artists impression of a binary black hole system producing gravitational waves. (Image credit: LIGO/Caltech/MIT/Sonoma State, Aurore Simonnet)

H.E.S.S. was the first ground-based instrument to observe the event GW170817 [1-3] on 2017 August 17, the neutron star merger that was accompanied by a gamma-ray burst. These observations resulted in the first stringent upper limits on the VHE emission from a neutron star merger. Since 2017, H.E.S.S has also observed four merger events of two black holes. Black hole mergers are not typically expected to produce gamma-ray emission because, due to long merger times, most of the surrounding matter has been accreted before the merger happens. However, a weak gamma-ray transient detected by the gamma-ray telescope Fermi-GBM [4] less than half a second after a merger of two black holes triggered much interest in the astrophysical community and encouraged the search for VHE emission from such events. Moreover, GW signals from black hole mergers are detected much more often than from neutron star mergers and are therefore a good way to test H.E.S.S.'s response to GW events.

The four observations reported here have allowed us to commission the H.E.S.S. GW follow-up program and to improve it for future observations. The localization regions of GW sources can span several tens to hundred of square degrees in the sky, while the H.E.S.S. field of view is around 20 square degrees. Therefore, follow-up observations of GW alerts are challenging and require special strategies to be implemented. These strategies are detailed in [5]. In a nutshell, the goal of these observation strategies is to identify the regions that are most likely to have hosted the event of the GW event, taking into consideration the visibility and observational constraints of the H.E.S.S. telescopes.

Fig. 2: H.E.S.S. coverage of the GW190728_064510 event. The project of the GW event localization on Earth is shown on the left at the time of start of the observation. The yellow and brown patches indicate the regions where light from the Sun and Moon, respectively, was too bright to allow for H.E.S.S. observations. The red square indicates the location of the H.E.S.S. telescopes and the red line indicates the region of the sky that was visible to H.E.S.S. The sky map with the probability of the position of the event together with the fields-of-view of individual H.E.S.S. observations (black circles) is shown on the right.The green dotted circle is a neutrino candidate detection by IceCube [10].

Only alerts of GW events whose localization regions do not exceed a few hundred degrees in the sky are picked for observation by our algorithms; otherwise, the region would be too large to cover in a manageable time. H.E.S.S. was able to schedule observations on several GW events. However, nature is not always on our side, and clouds and rain disturbed two of these observations. In total, and in addition to GW170817, four successful observations were performed on four different black hole mergers: GW170814 [6-7] in 2017 during the second GW observing period, GW190512_180714 and GW190728_064510 in 2019 during the first part of the third observation period [8] and GW200224_222234 in 2020 during the second part of the third observing period [9]. Electromagnetic counterparts of these events were never found by the astronomical community and thus they remain poorly localized. However, H.E.S.S. was able to cover large portions of their localization regions which increases the chance of having covered the true positions of the events. The start of the H.E.S.S. observations ranges from three hours after the detection of the event GW200224_222234 to two days for GW170814.

No significant VHE signal was found but the good quality data obtained permit us to compute upper limits on the VHE gamma-ray emission in the sky regions covered by H.E.S.S. [11-12]. These upper limits constrain for the first time VHE emission from black hole mergers.

Fig. 3: The H.E.S.S. upper limits (orange) on the observed energy flux (left) and luminosity (right) vs the observation time as measured from the merger, T - T_0 from four black hole merger events. The limits are compared to the emission of VHE GRBs (grey dots) and extrapolations of Fermi-LAT detected GRBs (grey lines). The upper limit for GW 170817 as measured by H.E.S.S. is also included (blue). (Image credit: [11])

The limits obtained in this study can be used by scientists over the world to constrain their emissions models. In Fig. 3 we compare our upper limits to other VHE transients. These transient are VHE-detected GRBs [13-14] or extrapolations of high-energy GRBs detected by the gamma-ray telescope Fermi-LAT into the VHE domain [15]. As seen in the left plot of Fig. 3, the H.E.S.S. energy flux upper limits are higher than the extrapolated emission of most of the transient events. But this plot does not tell the whole story, since GRBs are usually detected at much greater distances than GW events. For the luminosity upper limits, as shown in the right plot of Fig. 3, one can see that the H.E.S.S. upper limits are more constraining if the distance of the event is taken into consideration. If there had been any GRB-like VHE emission from the black hole mergers at the time of the observations, H.E.S.S. would have had a good chance of detecting it. Of course, these limits can be improved by having earlier and longer observations. Assuming that the H.E.S.S. observation strategy remains the same, and due to the expected high rate of GW detections with additional and/or more sensitive detectors in the future, H.E.S.S. will have on average the chance to observe several GW events with minimal delay. Moreover, the localization of GW events is expected to improve as more GW detectors join the network, which means that some of these events will be sufficiently well localized that they will be able to be observed with a single pointing. This will allow us to spend more time on one individual position and obtain longer, more sensitive observations, instead of spending time on covering large areas in the sky. In conclusion, this will result in stronger constraints on the VHE emission from black hole mergers.


[1] Abbott, B. P., Abbott, R., Abbott, T. D., et al. 2017, ApJL, 848, L13

[2] Abdalla, H., Abramowski, A., Aharonian, F., et al. 2017, ApJL, 850, L22

[3] Abdalla, H., Adam, R., Aharonian, F., et al. 2020, ApJL, 894, L16

[4] Connaughton, V., Burns, E., Goldstein, A., et al. 2016, ApJL, 826, L6

[5] Ashkar, H., Brun, F., F¨ ußling, M., et al. 2021, JCAP, 2021 (03), 045

[6] Abbott, B. P., Abbott, R., Abbott, T. D., et al. 2017, Phys. Rev. Lett., 119, 141101

[7] Ashkar, H., Schüssler, F., & Seglar-Arroyo, M. 2019, Mem. S.A.It., 90, 49

[8] Abbott, R., Abbott, T. D., Abraham, S., et al. 2020, arXiv e-prints, arXiv:2010.14527

[9] Abbott, R., Abbott, T. D., Abraham, S., et al. 2021, arXiv e-prints, arXiv: 2111.03606

[10] IceCube Collaboration. 2019, GCN circ. 25192

[11] H.E.S.S. collaboration. 2021, ApJ, 923, 109


[13] Abdalla, H., Adam, R., Aharonian, F., et al. 2019, Nature, 575, 464

[14] H.E.S.S. Collaboration et al. 2021, Science, 372, 1081

[15] Ajello, M., Arimoto, M., Axelsson, M., et al. 2019, ApJ, 878, 52