Constraints on the emitting region of the gamma-rays observed in GW170817

The gravitational waves from the neutron star merger event GW170817 were accompanied by an unusually weak short gamma-ray burst (sGRB) 170817A, by an optical/IR macronova/kilonova and by a long-lasting radio to X-rays counterpart. While association of sGRBs with mergers was predicted a long time ago, the luminosity of this prompt γ -ray emission was weaker by a few orders of magnitude than all known previous sGRBs and it was softer than typical sGRBs. This raises the question whether the γ -rays that we have seen were a regular sGRB viewed off-axis. We revisit this question following recent refined analyses of the γ -ray signal and the VLBI observations that revealed the angular structure of the relativistic outflow: observing angle of ∼20^◦, a narrow jet with a core ≾5^◦ and E_iso > 10⁵² erg. We show here that if the γ -ray spectrum is described by a Comptonized power law with E_p > 185 keV, then regardless of the detailed angular structure of the jet - top-hat or structured: (i) the region emitting the observed γ -rays must have been moving with a Lorentz factor Γ ≳ 5; (ii) the observed γ -rays were not 'off-axis' emission (viewing angle >1/Γ) emerging from the core of the jet (where a regular sGRB was most likely produced); (iii) the γ -ray emission region was either 'on-axis' (at an angle < 1/Γ) or if it was 'off-axis' then the observing angle must have been small (<5^◦) and the on-axis emission from this region was too faint and too hard to resemble a regular sGRB.