Simone Dichiara

Short-duration gamma-ray bursts (sGRBs) are brief (duration <2 sec) flashes of gamma-rays detected by the low orbit satellites with a rate of about 1 or 2 per week. They are produced in binary mergers of compact objects (e.g. binary neutron stars or black-hole-neutron star mergers). Most of the energy is released along the axis of the jet lunched from the newly formed central object along the polar regions. The observed flux decrease of several orders of magnitudes passing from an on-axis observer (jet aligned along the line of sight) to an off axis one.
The discovery of the first gravitational wave originated from a binary neutron star merger (GW170817) directly related to a sGRB (GRB 170817A) revealed a population of local bursts observed off-axis. Therefore, it becomes important to investigate the contribution of this specific population over the total sample of observed GRBs.
In order to do this we selected all the GRBs observed by the Neil Gehrels Swift satellite from January 1, 2005 to January 1, 2019 with short duration (T90<1 s) and no afterglow detection (e.g. signal fainter than the detector sensitivity) ending up with 32 possible off-axis bursts. Comparing the X-ray upper limits with the typical on-axis afterglows we conclude that these events need to be either observed off-axis or occurred in a very tenuous environment (atypical low densities; n<4e-3 cm^-3 for epsilon_B>1e-4).
Then we cross-matched the BAT GRB refined positions (90% error region) with the GLADE v2.3 galaxy catalogue in order to search for possible matches with local galaxies finding 4 possible candidates at distance <200 Mpc (12.5% of the sample). This fraction is higher than the rate expected from a chance alignment (about 3%; derived from simulations) and it possibly indicates a physical connection between some of these bursts and the nearby galaxies.
We derived optical and UV upper limits using archival follow-up observation for these four GRBs and we used them to constrain the physical properties of the associated kilonova.
Assuming that each event occurred inside the matched local galaxy, we constrain the ejecta mass and velocity for signals viewed toward their polar regions (lanthanide-poor ejecta) deriving M_ej ≤ 1e-3 M_sun for v_ej ≥ 0.2c. Our constraints are less tight in the internal region of the galaxies and we can not exclude emission observed at high latitude (e.g. lanthanide-rich).

Moreover, we derive a sGRBs all sky rate of R_sGRBs = 160 (-100;+200) Gpc^-3 yr^-1 assuming a local origin for these events. This implies a detection rate of sGRBs below 200 Mpc of 0.16 (-0.1;+0.2) yr^-1 and 0.8 (-0.5;+1) yr^-1 considering Swift and Fermi satellite, respectively.
On the other side, if the bursts occurred in distant (high redshift) galaxies, the rate should be lower than 180 Gpc^-3 yr^-1.
According to the most recent results from the Advanced LIGO and Virgo O3 the neutron star merger rate is 110<R_BNS<3840 Gpc^-3 yr^-1.
Comparing it with our rates, we derive a beaming factor for the boosted jet emission of fb^-1<80 and fb^-1>10 considering the local and cosmological origin, respectively.

In conclusion we do not find evidence for a large population of sGRBs in the local universe (4 possible matches with local galaxies).
Assuming that they are all local off-axis sGRBs, we derive a rate of R_sGRBs ~ 160 (beaming factor: fb^-1<80) and we can exclude the presence of a lanthanide-poor kilonova.
If they are cosmological on-axis sGRBs the rate is R_sGRBs>180 and the lack of X-ray afterglow could be explained by tenuous environment.
An optimization of the follow-up strategies and a systematic search for untriggered bursts could be crucial to increase the detection rate of local events.