J-GEM optical and near-infrared follow-up of gravitational wave events during LIGO’s and Virgo’s third observing run

In 2015, gravitational waves (GWs) were detected for the first time by the Laser Interferometer Gravitational-wave Observatory (LIGO) (2015CQGra..32k5012A, ), which marked the dawn of GW astronomy. The most plausible targets for the current GW detectors are signals emitted from compact binary coalescences (CBCs). CBCs include three types: Mergers of binaries composed of two black holes (BHs), two neutron stars (NSs), and one NS and one BH, called BBH, BNS, and NSBH, respectively. The first GW event, GW150914, detected by Advanced LIGO was a coalescing BBH (2016PhRvL.116x1102A, ). In 2017, the first GW event from a BNS, GW170817, was also detected by both the Advanced LIGO and Advanced Virgo (2015CQGra..32b4001A, ) detectors (2017PhRvL.119p1101A, ).

The direction of a GW signal is predominantly determined by the differences of the arrival times of the GW to more than two detectors. The uncertainty of the localization depends on the number of detectors that observe the GW signal, the source direction, and the signal-to-noise ratio. The median 90% credible region is 120–180 square degrees 2018LRR….21….3A , and is regularly exceed $\sim$1000 square degrees.

Depending on the nature of the coalescing source, a GW source may accompany electromagnetic (EM) emission 1989Natur.340..126E ; 1998ApJ…507L..59L ; 2010MNRAS.406.2650M . The identification and observation of the EM counterpart to a GW signal are important not only for understanding the physics of the merging event but also for maximally utilizing information in the GW signal. However, identifying an EM counterpart to a GW source is challenging, particularly in optical bands because of limited field-of-views (FoVs) of instruments. A wide-field blind survey using a wide FoV instrument can search any EM counterpart associated with the GW source, even without any association with host-like galaxy. Another approach is to survey cataloged galaxies within the three-dimensional (3D) localization map under the assumption that the EM counterpart is located close to or on a galaxy (2016ApJ…820..136G, ). This approach is called a galaxy-targeted survey.

In fact, EM follow-up surveys to GW150914 were performed across the entire EM spectrum, from radio to gamma-ray bands (2016ApJ…826L..13A, ). However, no evident EM counterpart was discovered (2016MNRAS.462.4094S, ; 2016ApJ…824L..24K, ; 2016ApJ…823L..33S, ; 2016PASJ…68L…9M, ). The non-detection in GW150914 may be a natural consequence of BBH merger. Note that a possible gamma-ray counterpart to GW150914 was tentatively claimed by 2016ApJ…826L…6C , which triggered the discussion on the possible presence of relativistic jets in BBH mergers as in gamma-ray bursts (e.g. (2016PTEP.2016e1E01Y, ), (2017NewA…51….7J, ), though see the counterargument presented by 2016ApJ…827L..38G ).

Identification of EM counterpart to a GW event was achieved for the first time for GW170817, where a gamma-ray signal was detected 1.74$\pm$0.05 sec after the GW detection (GRB 170817A, 2017ApJ…848L..13A ; 2017ApJ…848L..15S ). Subsequently, the optical counterpart (AT 2017gfo) was identified by galaxy-targeted surveys, at 10.86 hours after the GW detection (2017ApJ…848L..12A, ; 2017Sci…358.1556C, ). The optical emission of AT 2017gfo rapidly declined, while the near-infrared (NIR) emission lasted relatively longer (e.g. 2017ApJ…848L..16S ; 2017ApJ…848L..17C ; 2017ApJ…848L..24V ; 2017ApJ…850L…1L ; 2017PASJ…69..101U .) The optical spectra of AT 2017gfo rapidly evolved from blue to red (2017Sci…358.1574S, ; 2017Natur.551…67P, ). These properties are clearly different from those of supernovae. Also, EM emission was detected in X-ray (e.g. (2017ApJ…848L..25H, ), (2017Sci…358.1565E, )) and radio wavelengths (e.g. (2017ApJ…848L..21A, )) at 9 and 16 days after the GW detection, respectively. The X-ray and radio emission were detected for hundreds of days (2019ApJ…886L..17H, ). At the same time, the Very Long Baseline Interferometric observation detected a knot with a superluminal motion (2018Natur.561..355M, ).

When a BNS coalesces, a portion of matter of the merger material is ejected into interstellar space (1986ApJ…308L..43P, ). In this material, $r$-process elements including lanthanides are likely to be synthesized. This matter gives rise to thermal emission mainly in the optical and IR bands, which is powered by the radioacitve decay of the $r$-process nuclei (1989Natur.340..126E, ), called a ”kilonova”. The observed temporal evolution of the optical and NIR emission of AT 2017gfo was consistent with simple kilonova models (2017PhRvD..96l3012S, ; 2017PASJ…69..102T, ; 2017ApJ…851L..21V, ). However, the blue component seen in the early phase may not be fully reproduced by the simple kilonova model. To explain the blue emission, two models were proposed; (1) a kilonova model with a higher electron fraction (2017PASJ…69..102T, ), or (2) a cocoon generated by interactions between the NS ejecta and the jet that powered the emission from radio to gamma-ray wavelengths (2017ApJ…848L..20M, ; 2018ApJ…855..103P, ; 2018MNRAS.478L..18T, ; 2018MNRAS.479..588G, ; 2018PhRvL.120x1103L, ). Both models can explain the light curve observed after 0.5 days. But these models predict different brightness at $\leq$ 0.5 days (2018ApJ…855L..23A, ). Thus, early-phase observations are required to determine the origin of the optical blue emission.

The third LIGO/Virgo observing run (O3) started on 1st April, 2019. During O3, 56 GW alerts were announced ¹¹1https://gracedb.ligo.org/superevents/public/O3/. A GW signal that is consistent with the CBCs can be estimated plausible masses of its components. Based on the masses, the detected GW signals in O3 were classified into 37 BBHs, six BNSs, five NSBHs, and four MassGap objects of which masses consist of between 3$\;M_{\odot}$ and 5$\;M_{\odot}$. Details of four GW events were already published by the LIGO Scientific Collaboration and the Virgo Collaboration (LVC): BBH event GW190412 was likely originated from the coalescence of BBH with asymmetric masses of $\sim 8\;M_{\odot}$ and $\sim 30\;M_{\odot}$ (2020PhRvD.102d3015A, ), and GW190521 was originated from the BBH merger with a total mass of 150 $M_{\odot}$ 2020PhRvL.125j1102A . In this event, a candidate of an optical counterpart was claimed 2020PhRvL.124y1102G . A BNS event GW190425, which was the first event with a NS in O3, was likely originated from a coalescence of BNS with a total mass of $3.4\;M_{\odot}$ (2020ApJ…892L…3A, ). Finally, a NSBH event GW190814 was likely originated from the coalescence of a 23 $M_{\odot}$ BH with a 2.6 $M_{\odot}$ compact object 2020ApJ…896L..44A .

In the O3, EM follow-up observations have been performed by various telescopes and satellites from a gamma-ray to radio wavelengths. In the optical and NIR wavelengths, a lot of follow-up campaigns have been performed for detected GW signals: GROWTH 2019ApJ…885L..19C ; 2019ApJ…881L…7G ; 2019ApJ…881L..16A ; 2020ApJ…890..131A ; 2020arXiv200305516A , GRANDMA 2020MNRAS.492.3904A ; 2020MNRAS.497.5518A , GOTO 2020MNRAS.497..726G , Magellan 2019ApJ…884L..55G , ZTF 2020NatAs.tmp..179A , MMT and SOAR 2019ApJ…880L…4H .

The Japanese collaboration for Gravitational-wave ElectroMagnetic follow-up (J-GEM) has conducted optical and NIR follow-up campaigns of GW events since 2015 (2016PASJ…68L…9M, ; 2017PASJ…69….9Y, ; 2018PASJ…70….1U, ; 2017PASJ…69..101U, ; 2017PASJ…69..102T, ; 2018PASJ…70…28T, ). In this paper, we introduce J-GEM facilities and infrastructures which realize rapid follow-up observations, and describe details of our galaxy-targeted surveys to GW events in O3 run. Results of wide-field blind surveys by Hyper Suprime-Cam (HSC) on the 8.2-m Subaru telescope and Tomo-e Gozen on the 1.05-m Kiso Schmidt telescope will be reported by Ohgami2020 ; Niino2020 , respectively. We describe our observational methods in Section 2, and the results of the follow-up observations in Section 3. In Section 4, we discuss the results and future prospects of our follow-up campaigns. Finally, we give our conclusions in Section 5.

J-GEM consists of a consortium of facilities that cover optical, NIR, and radio wavelengths. To enable rapid follow-up observations for GW sources detected at arbitrary time from various directions, we form a network of multi-latitudinal and longitudinal observatories (see figure 20 in Tanaka2020 ). Our network enables an effective search for a transient in a large GW localization area by involving multiple telescopes and cameras. The telescopes used for EM follow-up during O3 are listed in Table 1. With these multi-located telescopes, we are able to perform the stable and quick follow-up avoiding the effect of bad weather. Optical data were obtained with nine instruments on eight telescopes, and NIR data were obtained with four instruments on four telescopes.

We conducted both galaxy-targeted and wide-field surveys in response to GW event alerts. These two approaches have different advantages. The wide-field survey can provide a search of the counterpart regardless of having or not having the host galaxy if the FoV of the camera is sufficiently large. The galaxy-targeted survey can also be effective if the counterpart is associated with the host galaxy. To maxmize the efficiency, the cataloged galaxies within 3D localization map can be ranked according to the position in the localization area as well as a luminosity of the galaxy. By coordinating galaxy-targeted observations with multiple telescopes distributed over the world, we can further enhance the capability of our survey by reducing the risk of weather loss. A drawback of galaxy-targetted survey is an incompleteness of the cataloged galaxies, which is severe at $\geq$200 Mpc (Fig. 7). On the other hand, the wide-field surveys do not rely on the information of cataloged galaxies.

During O3, we performed galaxy-targeted follow-up for 23 GW events (Table 2). Depending on the properties of the signals and estimated masses, each event is categorized by LVC into six classes: BBH, BNS, NSBH, MassGap, Burst, and Terrestrial. A Terrestrial event is likely to be an instrumental or environmental origin. In addition, LVC estimates the probabilities, HasNS, in which the GW source includes at least one NS component, and HasRemnant, in which the GW source has a non-zero amount of neutron star matter. Each facilities took into account these parameters to decide observations. Eventually, we observed 11 BBH, five BNS, four NSBH, one MassGap, one Burst, and one Terrestrial (having $\sim$40% probability of BNS) during O3.

Data of our follow-up campaigns were recorded in planner and Image Server, and the summary is shown in Table 3. The maximum number of the observed galaxies for each GW event is 170, which is for GW190425. The total coverage of each GW event is estimated by summing the probabilities of the observed galaxies, where the probability of each galaxy, $P^{(j)}$, is calculated using Eq. (1). The total sky coverage of our galaxy-targeted follow-up campaign for each GW event during O3 reaches a maximum of 9.8% and a median value of 0.012%.

The limiting magnitudes of our observations depends on the telescopes, instruments, and the sky condition. The deepest limiting magnitudes of our follow-up observations in the optical and NIR bands are 21.2 and 20.6, and median limiting magnitudes are 18.7 and 18.6, respectively. Here, the limiting magnitude of the FOCAS observation, 23.5 mag in $R$-band, is not included in the calculation of median values. The peak magnitudes of GW170817 scaled by distances of 40 Mpc ($g=$16.8), 100 Mpc ($g=$18.8), and 200 Mpc ($g=$20.3) are shown together with the distribution of the obtained limiting magnitudes in Fig. 2. Thus, with our observations, we could potentially detect emission from a GW170817-like BNS source at a distance of $\sim$100 Mpc.

Since early detection of the EM counterpart is desired to provide better understanding of the emission mechanisms at this epoch (see Section 4.2), it is worth analyzing the starting time after the GW detection. The starting times of our observations after GW detections are distributed with a range from 0.041 days to 4.841 days, with a median value of 0.558 days. The starting times of the events are shown in Fig. 3. We succeeded in starting follow-up observations within 0.5 days after detection for 10 GW events.

J-GEM is a consortium of astronomical observation facilities that covers at optical, NIR, and radio wavelengths. We developed an efficient data management and candidate screening system that perform the following: (1) ranking cataloged galaxies in the 3D localization map, (2) coordinating and collecting information of observations with different facilities, (3) measuring the characteristics of obtained images, and (4) producing cut-out images of new images as well as reference and subtracted images. The system enables efficient search of a possible transient in the GW localization volume and subsequent follow-up observations of potential candidates.

We performed follow-up for 23 GW events (11 BBHs, five BNSs, four NSBHs, one MassGap, one Burst, and one Terrestrials) during O3. In our follow-up campaigns, the shortest starting time between the detection of the GW signal and the onset of observations was as short as 0.041 days, and observations of 10 GW events started within 0.5 days after GW detection. Some of the limiting magnitudes obtained in the follow-up to BNS and NSBH were deeper than the scaled luminosity of GW170817. We show that our follow-up system exhibits the potential to detect a GW170817-like EM counterpart in the GW event with a localization accuracy of $\sim$500 deg² and at a distance of $\sim$100 Mpc.

LVC will schedule the next observing run that will operate together with KAGRA 2020JPhCS1342a2014A . The expected localization accuracy will be narrower to 110—180 deg² at 115 Mpc in the case of BNS thanks to the increased number of GW detectors (2018LRR….21….3A, ). Out studies demonstrate that J-GEM facilities will contribute to follow up GW events detected by GW detectors during the next run.

This work was supported by MEXT KAKENHI Grant Numbers JP17H06363, and JP24103003, JSPS KAKENHI Grant Numbers JP15H02069, JP16H02158, JP17H06362, JP18H03720, JP19H00694, JP19K14761, JP20H00158, and JP26800103, the U.S. Department of Energy under contract number DE-AC02-76-SF00515, Optical and Near-Infrared Astronomy Inter-University Cooperation Program, and the joint research program of the Institute for Cosmic Ray Research (ICRR).

Based in part on data collected at Subaru Telescope, which is operated by the National Astronomical Observatory of Japan.

The Pan-STARRS1 Surveys (PS1) and the PS1 public science archive have been made possible through contributions by the Institute for Astronomy, the University of Hawaii, the Pan-STARRS Project Office, the Max-Planck Society and its participating institutes, the Max Planck Institute for Astronomy, Heidelberg and the Max Planck Institute for Extraterrestrial Physics, Garching, The Johns Hopkins University, Durham University, the University of Edinburgh, the Queen’s University Belfast, the Harvard-Smithsonian Center for Astrophysics, the Las Cumbres Observatory Global Telescope Network Incorporated, the National Central University of Taiwan, the Space Telescope Science Institute, the National Aeronautics and Space Administration under Grant No. NNX08AR22G issued through the Planetary Science Division of the NASA Science Mission Directorate, the National Science Foundation Grant No. AST-1238877, the University of Maryland, Eotvos Lorand University (ELTE), the Los Alamos National Laboratory, and the Gordon and Betty Moore Foundation.

This publication makes use of data products from the Two Micron All Sky Survey, which is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by the National Aeronautics and Space Administration and the National Science Foundation.