be more frequent than bright bursts and may
have been missed by smaller telescopes such
as Parkes. the bursts to a precision of 4 milliarcseconds
(Marcote et al., 2017), allowing us to further
constrain the environment of the FRB.
The Repeater became the focus of our effort
for the interferometric localization of FRBs.
Radio interferometry is a technique to com-
bine the signals from different radio tele-
scopes to effectively achieve the resolving
power of a radio telescope that is as large
as the separation between the telescopes.
Using the Very Large Array (VLA) and col-
lecting interferometric data with a sampling
time of 5 milliseconds (instead of a few tens
of seconds), our team was able to search for
repeated bursts. This was a watershed moment in the emerg-
ing field of FRBs. For the first time, the local-
ization was sufficient to search for optical
and infrared counterparts and perhaps iden-
tify where FRBs originate.
Compared to the 10-arcminute localization
of the single 300-m dish at Arecibo Observa-
tory, the ~30 kilometer baseline of the VLA
was able to localize nine bright bursts from
the Repeater to a 100 milliarcsecond preci-
sion (Figure 2; Chatterjee et al., 2017). Using
the European Very Long Baseline Interfer-
ometry Network of radio telescopes spread
across Europe, our team further localized
April 2017
Optical Counterpart
Archival R-band images from the Keck Observa-
tory showed a very faint (R = 24.5 magnitude)
object detected at about 5-σ but it was not
clear whether it was an extended source or a
point-like object. We were granted nine hours
of Gemini Director’s Discretionary Time for fur-
ther imaging and spectroscopy with the Gem-
ini Multi-Object Spectrograph to characterize
the counterpart and investigate whether the
FRB was Galactic or extragalactic.
Gemini Observatory’s flexible queue sched-
uled observations were critical to the success
of this project. The faint target required dark
GeminiFocus
Figure 2.
The interferometric
localization of the
bursts from FRB 121102
using the VLA. Panel
(a): radio image of the
burst; the black circles
are Arecibo beam sizes.
Panel (b): zoom-in of
the radio image; the
position precision for
each burst is less than
one arcsecond. Panel
(c): the de-dispersed
time profile, spectrum,
and dynamic spectrum
of the burst. Figure
adapted from
Chatterjee et al., 2017.
5