J. Lyman (Univ. of Warwick), D. Homan (Univ. of Edinburgh), K. Maguire
(QUB), M. T. Botticella (INAF-Capodimonte), M. Fraser (UCD), C.
Inserra (Southampton), E. Kankare (QUB), S. J. Smartt (QUB), K. W.
Smith (QUB), M. Sullivan (Southampton), S. Valenti (UC Davis), O.
Yaron (Weizmann), I. Manulis (Weizmann), D. Young (QUB), T.-W. Chen
(MPE), S. Campana (INAF-Brera), S. Benetti, L. Tomasella
(INAF-Padova), G. Leloudas (DARK, Copenhagen), Z. Cano (IAA, Granada).
Under the extended Public extened ESO Spectroscopic Survey for
Transient Objects (ePESSTO; see Smartt et al. 2015, A&A, 579, 40
http://www.pessto.org), we report spectra of the optical/NIR candiate
counterpart of the gravitational wave source G298048 in NGC4993.
This object was discovered by Coulter et al. (GCN 21529, named SSS17a)
and further detected by Allam et al. (GCN 21530), Yang et al. (GCN
21531), Melandri et al. (GCN 21532), Nicholl et al. (GCN21541), Tanvir
& Levan (GCN21544), Chambers et al. (GCN 21553), Yoshida et al. (GCN
The observations were performed on the ESO New Technology Telescope at
La Silla with the EFOSC2 instrument in spectroscopic mode starting
from 2018-08-18 at 23:20 UT, using the Gr11 and Gr16 which provide
coverage 3345-9995 at a resolution between 260-396 (v = 1150 - 756
km/s) for a 1.5" slit, which was used in roughly 1.5-2.0" seeing.
Quick look reductions (not including fringe and flat correction)
indicate the spectrum is highly unusual for a nearby transient.
We have dereddening the spectrum with A_v = 0.34, for Milky
The continuum is featureless, similar to the report of the first
Magellan spectrum in Drout et al. (GCN 21547). But it peaks at 5500
and falls off rapidly in the blue. This is consistent with the Swift
colours reported and the fading U-band and UV flux. The red part
(beyond 5000Angs) can be fit with a black-body temperature of around
6000K or a power law. But neither are consistent with the turn over
in the blue. There are no strong features in the blue that indicate
line absorption or could be identified with known transitions. There
are no Balmer lines, no Ca II H&K, or Si II absorption seen in
We can certainly conclude this is not a young supernova of any type in
this host, nor is it consistent with being a background supernova of
any standard type. The UV dimming and maintaining of the red and NIR
flux appears broadly consistent with many flavours of kilonova models.
Further observations with ePESSTO will continue over the next 4
nights, combining EFOSC2 and SOFI observations. During the 5 years of
PESSTO spectroscopic surveys, this transient stands out as exceptional
in its colours, evolution and spectra (supporting the ideas in Foley