LIGO, Leaks and NGC 4993

No matter what the official policy may be, the more people there are in a collaboration the more likely it is that someone will let their excitement get to their head and start leaking news and starting rumours either directly or indirectly via social media. And so it came to pass last Friday that the following tweet appeared:

I didn’t comment on the time as I thought it might be unreliable – as it indeed it still may be – but now New Scientist has amplified the signal I feel I can’t really be blamed for mentioning it here.

The rumours going round identify the optical counterpart as being in the galaxy NGC 4993 , a red band image of which, from the Second Digitized Sky Survey (DSS2) is shown below:

NGC 4993 is the fuzzy blob slightly above and to the left of the centre of the image. It’s a fairly nondescript lenticular galaxy in a group that can be found in the constellation of Hydra. It lies in the constellation of Hydra, was actually first discovered by William Herschel and it is about 10 arcmin across on the sky. It’s quite nearby, as these things go, with a distance of about 124 million light years (i.e. 40 Mpc or so) and is about 14th magnitude.

If there is an optical counterpart to a gravitational wave event coming from this galaxy then that suggests it may be a coalescence of neutron stars. The black hole mergers that appear to be responsible to the three existing gravitational wave signals that are claimed to have been detected are not expected to release optical light. Confirmation of this interpretation can be found by where the Hubble Space Telescope was pointed yesterday:

Look familiar? HST was, in fact, observing a `BNS-Merger’ (which is short for `Binary Neutron Star’)…


If this rumour is true then it’s obviously exciting, but there are questions to be asked. Chief among these is how sure is the identification of the counterpart? A transient optical source in NGC4993 may have been observed at the same time as a gravitational wave signal was detected,  but the ability of LIGO to resolve positions on the sky is very poor. On the other hand, the European VIRGO experiment joined Advanced LIGO for the ongoing `O2′ observing run (which ends in a couple of days). Although VIRGO is less sensitive than LIGO having a third detector does improve the localization of the source – assuming, of course, that it detects a signal. Even in that case it certainly won’t be possible to pinpoint the GW source to within 10 arc minutes, which is the precision needed to place it definitely within NGC 4993.

Anyway, we wait and see what, if anything, has been found. If it is a claimed detection then I hope that LIGO and VIRGO will release sufficient data to enable the analysis to be checked and verified. That’s what most of the respondents to my poll seem to hope too!

65 Responses to “LIGO, Leaks and NGC 4993”

  1. Joseph Conlon Says:

    If you look at the (public) Chandra short-term schedule,

    you see that it has also pointed on the 19th August for 25ks (at NGC4993) at SGRB170817A, a triggered Target-of-Opportunity for X-ray counterpart of a GW event.

    From the (entirely public) obsid details at

    you can see that

    (a) a requirement for the trigger to be activated was both a GW discovery AND a positive EM counterpart identification by the Dark Energy Camera

    (b) the obsid has a count rate of 11Hz and there are 300k X-ray photons in the observation (which sounds like a HUGE count rate for an X-ray telescope)

    It looks very real.

  2. Tom Shanks Says:

    Since ESO VLT was observing it for a GRB follow-up team over the weekend it looks like it was coincident in time with a GRB source and that gave the position on the outskirts of ngc4993. If you want the exact coord you can look in ESO archive! GRB interpreted as neutron star- neutron star collision. All unconfirmed but…

  3. Anton Garrett Says:

    I thought that three detectors were needed in order to triangulate where in the sky the gravitational radiation was coming from, but there are currently only two LIGO detectors. So how come they can claim that the radiation comes from a specific point in the sky (which happens to correspond to an optical source)? Could anyone explain, please?

  4. How does SGRB170817A (I understand S stands for Short) relate to the FERMI detection GRB170817A, GCN 21520, which is at RA = 176.8, DEC = -39.8, although with a pretty large error. Is there either a separate set of detections, or a set of triangulations of existing detections, which is not reported through GCNs?

    • Tom Shanks Says:

      I would have said they would get the exact position from GRB then the X-ray but the above X-ray obsn was on 19/8 whereas ESO VLT X-shooter obsns started on 18/8. Maybe they did detect it in optical imaging first. Incidentally I see that they have observed the GW/GRB afterglow every night since wit Shooter as well as VISTA VIRCAM etc.

      • telescoper Says:

        The override info at

        suggests a GW detection and a positive EM detection prior to the Chandra observations.

      • The point though Tom is that if it is the same source, then that position given in the GCN is a long, long way from NGC4993 (which is at 13h09m47.7s -23d23m02s). I can’t be bothered to get out my calculator and work out exactly how far, but its probably around 25 degrees. The quoted uncertainty in the FERMI position is 11.6 degrees, so this is just about consistent.

        Getting a gamma ray position then X-ray is how SWIFT usually works, but I see no GCN from SWIFT between August 13th and August 22nd, and I think I get them all.

      • Reading the abstract I think that the position comes from LIGO then DECam, and the identification with the GRB is maybe a bit tentative. DECam field of view is 2.2 degrees diameter, so it cannot have been working from the FERMI position. Maybe LIGO + VIRGO positions are that good now. So I think Tom, you are right that they detected it in the optical first (or second after the GW detectors).

      • telescoper Says:

        I don’t believe the LIGO+VIRGO position can be that good. I think it’s primarily an argument based on timing.

  5. Will Sutherland Says:

    Couple of comments: the Fermi burst info went public 7 hrs after the detection, so if the presumed LIGO event was coincident to ~ 1 sec in time then LIGO would know then or eariler that there’s a very low coincidence probability and hammer the optical/NIR/Xray searches.
    I also doubt the LIGO position can be good to 1 degree; however,
    DECam can simply mosaic ~ 100 pointings to cover a ~ 300 deg^2 burst uncertainty region, though 20 degrees radial offset is getting hard – it looks like they must have had some additional info to narrow the search zone, e.g. combined LIGO/Fermi position , or SWIFT, or possibly a flash from one of the all-sky cameras ?
    An arcmin-or-better position would be required in advance of the Chandra and Hubble followup, so there surely must be some sort of optical/NIR/X-ray transient detected before those.

    • Although the GCN was issued 7 hrs after the burst, there is information on the trigger page here:

      which is issued from a few seconds after the burst. The trigger with what they call the final position (ironically further away from NGC4993 than the initial position) was some three hours later.

      How immediate is a LIGO detection?

  6. Bryn Jones Says:

    The HST has made a series of observations of NGC 4993 recently. This sequence of observations started at 2017 August 22d 07h UTC and was for the following three HST proposals:

    HST Proposal 14804: Rapid ToO observations of the first gravitational wave counterparts

    PI: Andrew Levan (University of Warwick)

    Proposal Abstract:
    We propose a series of disruptive ToO observations of the first electromagnetic counterparts to gravitational wave sources. These observations will track the likely rapidly fading counterparts to levels a factor 10 fainter than possible from the ground. They will determine the spectral and temporal evolution, evaluate their power sources, ascertain their contribution to the production of heavy elements in the Universe, pinpoint them on their host galaxies and provide information to hone further searches. In concert with already award late time (>3 week turnaround) observations these observations will provide a unique and powerful view of a newly discovered, but long awaited class of astronomical object.

    HST Proposal 15346: Verifying a candidate counterpart to gravitational waves

    PI: Mansi Kasliwal (California Institute of Technology)

    Proposal Abstract:

    Proposal Abstract:
    With advances in sensitivity of gravitational wave interferometers, the direct detection of neutron star mergers should be imminent. Identification of an electromagnetic counterpart would enable a wealth of astrophysics and answer the long-standing question of whether neutron star mergers are the missing cosmic mines of heavy elements synthesized by the r-process. We will be searching for a fast-fading optical counterpart with the new Zwicky Transient Facility at Palomar Observatory. Here, we propose to use HST/WFC3 to look for infrared emission from a single, most-promising candidate optical counterpart. The infrared emission would serve as a direct diagnostic of the radioactive decay of heavy elements;

    HST Proposal 15382: UV Spectroscopy of GRB170817A

    PI: Matt Nicholl (Harvard University)

    Proposal Abstract:
    GRB170817A is the most nearby short GRB ever discovered. We detected an optical counterpart using DECam and propose to obtain a UV spectrum using STIS.

  7. Davide Castelvecchi Says:

    It appears that ALMA was looking at NGC 4993 too 🙂

  8. Bryn Jones Says:

    I see from the NOAO Archives that DECam (Dark Energy Camera) on the CTIO 4-metre telescope imaged the NGC 4993 region on 18 August 2017 (around 00h 05m UTC).

  9. Tom Shanks Says:

    This does raise the issue of why all this secrecy is necessary. While it’s good fun working out what happened maybe someone should tell us! After all, these detectors and telescopes are all publicly funded facilities and maybe their follow-up would be even better had there been less “Omertà”. While I appreciate we don’t want to be flooded by false detections, I get the feeling that there is some Astro power politics involved in keeping things private. Thank goodness for twitter or we wouldn’t know anything!

  10. George Dishman Says:

    While LIGO on its own can give only an annulus on the sky from the two detectors, and VIRGO would localise that to a small region where three annuli overlap (3 by any 2-of-3), LIGO also gives luminosity distance from the measured strain. If that was around 40 Mpc, even with a quite large uncertainty, the known range limits the number of candidate galaxy hosts in the search volume.

    • John Peacock Says:

      Although I can’t provide a good explanation, LIGO does better than an annulus, even with just two detectors. Look at the localisation maps from their first event.

      • George Dishman Says:

        I believe that is because the arms at the two sites are slightly offset and of course the curvature of the Earth adds to that. The result is that the sites respond slightly differently and that improves the localisation.

      • Timing from to separated detectors gives an annulus. The slight misalignment of the arms between the two sites makes them sensitive to slightly different combinations of the two wave polarizations. For a binary, the signals in the two polarizations are correlated. Taken together, two detectors can, up to statistics of source orientation and SNR, localize the source. Localization is not at all of the precision characteristic of other astronomical observations (and is inconveniently shaped, to boot), but it is much better than you’d get from the timing annulus alone.

  11. Davide Castelvecchi Says:

    The official LIGO-Virgo end-of-O2 statement has now appeared:
    “The Virgo and LIGO Scientific Collaborations have been observing since November 30, 2016 in the second Advanced Detector Observing Run ‘O2’ , searching for gravitational-wave signals, first with the two LIGO detectors, then with both LIGO and Virgo instruments operating together since August 1, 2017. Some promising gravitational-wave candidates have been identified in data from both LIGO and Virgo during our preliminary analysis, and we have shared what we currently know with astronomical observing partners. We are working hard to assure that the candidates are valid gravitational-wave events, and it will require time to establish the level of confidence needed to bring any results to the scientific community and the greater public. We will let you know as soon we have information ready to share.”

    • Davide Castelvecchi Says:

      From this report on the Virgo duty cycle, it looks very likely that Virgo was in lock at 20:00:07 GMT on 17 August (the time stamp on the GRB170817A circular)

      • Davide Castelvecchi Says:

        Although it if I am not mistaken, at that time (22:00 CET) Hydra had just set, SW below Pisa’s horizon, which looks pretty close to one of Virgo’s blind spots to me. (Its arms point N-NE and W-NW, so the blind spots would be N-NW and W-SW, I think.)

      • Davide Castelvecchi Says:

        Actually I was clearly having a brain melt when I wrote that. The time stamp on the GCN is 20:00:07, but the event itself occurred at 12:41:06.47 UTC! At that time, it seems that Hydra was just above the horizon, S-SE of Cascina. I’m not sure how good that is in Virgo’s antenna pattern.

      • Will Sutherland Says:

        The S-SE of VIRGO looks about right – as an approx estimate,
        at the Fermi 12:41:06 UTC it appears that NGC 4993 was at zenith around longitude 41.6E, 23S i.e. in the Mozambique Channel; while the sub-solar point was in the SW of Mali.

        This implies all Australia was dark, but NGC4993 had almost set from the eastern side with most Australian telescopes. If there was any simultaneous visible/NIR “flash”, Western Australia, S.Pole or Indonesia would be possible sites to have seen it in darkness. From CTIO/ESO, it became observable soon after sunset some 11h later.

      • Agreed. I also checked it.

    • telescoper Says:

      It’s likely we will have to wait a matter of months for an official science update, as that was the timeline for previous announcements of GW detections and there is now potentially a lot of EM data to analyse too.

      • Davide Castelvecchi Says:

        Agreed! Although it’s fun to speculate based on the info we have, isn’t it? 🙂

  12. telescoper Says:

    I’d like to say a very big `thank you’ to all the commenters on this post who have taken the trouble to provide so many interesting and well-informed contributions on this post!

  13. […] there has been speculation on Twitter, in a New Scientist article and on astronomer Peter Coles’s blog that such a link has been found, based on publicly available observation […]

  14. […] there has been speculation on Twitter, in a New Scientist article and on astronomer Peter Coles’s blog that such a link has been found, based on publicly available observation […]

  15. […] there has been speculation on Twitter, in a New Scientist article and on astronomer Peter Coles’s blog that such a link has been found, based on publicly available observation […]

  16. […] 23 August, a commenter on the blog of astrophysicist Peter Coles, of Cardiff University in the UK, noted that NASA’s Chandra X-ray observatory had jumped into the […]

  17. […] there has been speculation on Twitter, in a New Scientist article and on astronomer Peter Coles’s blog that such a link has been found, based on publicly available observation […]

  18. […] Weitere Artikel finden sich hier (Übersetzung via Spektrum), hier, hier sowie hier. […]

  19. […] 23 August, a commenter on the blog of astrophysicist Peter Coles, of Cardiff University in the UK, noted that NASA’s Chandra X-ray observatory had jumped into the […]

  20. […] 23 August, a commenter on the blog of astrophysicist Peter Coles, of Cardiff University in the UK, noted that NASA’s Chandra X-ray observatory had jumped into the […]

  21. […] 23 August, a commenter on the blog of astrophysicist Peter Coles, of Cardiff University in the UK, noted that NASA’s Chandra X-ray observatory had jumped into the […]

  22. […] 23 August, a commenter on the blog of astrophysicist Peter Coles, of Cardiff University in the UK, noted that NASA’s Chandra X-ray observatory had jumped into the […]

  23. Catching up … weird secrecy going on here, only going to give the conspiracy-theorists (if they find out about it) more reasons to distrust astronomers !

  24. Craig Lage Says:

    I think the people that worked on LIGO for more than 20 years without seeing a signal of gravitational waves deserve a few weeks to digest the new data before they publish it. You don’t want raw data anyway, you will just have more questions. You want digested data and conclusions. Give them the time they need to do good science.

  25. […] 23 August, a commenter on the blog of astrophysicist Peter Coles, of Cardiff University in the UK, noted that NASA’s Chandra X-ray observatory had jumped into the […]

  26. […] WordPress blog statistics page, there’s been a lot of traffic here in the past week owing to my post about the rumours of a new gravitational wave source detected by LIGO (and possibly VIRGO). In the interest of completeness I’ll just post a quick […]

  27. […] 25 Aug 2017, doi: 10.1038/nature.2017.22482; Peter Coles, “LIGO, Leaks and NGC 4993,” In the dark, 23 Aug 2017; y muchos otros […]

  28. There might be some astronomical observation of gravitational waves produced by neutron stars; although, I think, this time it is quite improbable, looking at the sheer fortuitousness of the so-called detection makes it untenable — the VIRGO run was too short (just 25 days), LIGO never found any orbiting neutron stars’ gravitational waves in the last 3 years, while there are too many neutron stars nearby to have slipped LIGO’s notice. Let me add, 6 years earlier, minuscule gravitational waves of a wide frequency range (nearly zero to around 3 KHz) were first produced and detected in my lab late in 2010 and were reported in a US patent application which now is a US patent 8521029. You can find the patent detail on the USPTO site as well as on . You can check out gravitational waves and my work on Wikipedia. Let me also add, even if I am letting out a little secret, it is impossible to register any black hole mergers, because of the sheer volume of mergers — I cannot talk more on this subject. So, let me tell you, LIGO actually never detected any black hole mergers in the past too. The least I can say is that the reported mergers were a result of the intense imagination of the LIGO folks, to say the least.

  29. […] über eine LIGO-Beobachtung verschmelzender Neutronensterne – siehe z.B. hier, hier, hier, hier, hier und hier – zu stimmen scheinen, mit der Verkündung am 16. Oktober. Und er erzählt, […]

  30. […] of other observatories strongly suggests that this announcement will address the subject of the rumours that were flying around in August. In other words, it’s likely that on Monday we will hear about the first detecting of a […]

  31. Bryn Jones Says:

    So, it looks as though we deduced the main elements of the story correctly!

    • Davide Castelvecchi Says:

      Indeed! Although how the actual story unfolded in reality was insane, beginning with the fact that they noticed the event based on a single-interferometer trigger

      • telescoper Says:

        Yes, the lack of a detection in Virgo meant they needed quite a bit of scrambling to get the position!

      • Bryn Jones Says:

        Yes. I now regret that one of us didn’t summarise our information to give a most likely scenario. We’d have been close to the truth!

        I had thought the actual story would not have been as clean as we’d have hoped, probably because the signal-to-noise ratio would be poor, or there was a non-trivial difference in time between the GW event and the GRB. In the event, the signal was strong and the match in times was close.

        Instead, the departure from an ideal case involved the initial detection from only one interferometer, something I’d never have guessed!

      • Davide Castelvecchi Says:

        NGC 4993 is indeed very close to Virgo 🙂

  32. […] a hundred people gathered at IUCAA to see this evening’s press conference, which basically confirmed most of the rumours that had been circulating that a Gamma Ray Burst had been detected in both GW and EM radiation. I won’t write in detail […]

  33. Ashley Sears Says:

    MrMBB333 a YouTuber has been reporting and documenting this for over 5 years

  34. Within a non-academic point of view, regarding gravitational waves, we should bear in mind there are various meanings of the expression gravitational waves, and those detected by LIGO experiment are not the cause of gravity force. From another perspective, they are not produced by accelerating mass but by violent merging bodies. Finally, because of the drag effect they produce, they are most probably related to the so-called expansion of the Universe and dark energy.

  35. […] including reblogs). The two most popular posts were about the Bullying Scandal in Zurich and on the rumours surrounding a gravitational waves from a source in NGC 4993 (proved later to be […]

  36. […] of the things I remember about this was the fascinating way in which various `outsiders’ used a comments thread on this blog to piece together the clues to what going […]

  37. […] other sources that promoted the rumor include: The Daily Mail, Nude Socialist, Telescoper, Starts With A Bang, Wired, National Geographic, Quanta Magazine, The Independent, Science Alert, […]

  38. […] started to spread via this blog as people outside the LIGO/transient source community used a comments thread here to share information of where telescopes were looking. Was that really 6 years […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: