News & Views


Cosmic flashing lights

Gravitational lensing is becoming increasingly important to the study of distant galaxies and dark matter. Two groups have recently detected transient events emanating from far-away lensed galaxies, apparently due to extreme magnification of individual stars.

Monitoring of the night sky reveals a dynamic picture of the cosmos. Observed sources may brighten or become dimmer or even completely disappear, or new sources can appear, seemingly from nothing. We know that a wide range of stellar and galactic processes can give rise to variability and the study of such transient events can provide valuable insight on the physics underlying these processes. In this issue of Nature Astronomy, Steven Rodney and collaborators1 and Patrick Kelly and collaborators2 report a peculiar set of transients that appear to defy traditional transient classification. Instead, the authors argue that the observed variability is due to individual stars being lensed by intervening compact masses, an effect known as microlensing.

Both sets of transient events occurred within galaxies that lie behind massive clusters of galaxies whose immense gravitational pull deflects light from the background galaxies. The galaxy clusters thus serve as gravitational lenses, effectively increasing the power of our telescopes and producing multiple, distorted, and greatly magnified images of the more-distant galaxies3,4 (Fig. 1). The newly discovered transients correspond to stars in the lensed galaxies that have themselves been lensed. Lensing is therefore providing a front-row seat to processes we would not otherwise be able to witness, since even stellar outbursts are not generally detectable at these great distances.

Fig. 1: An image of galaxy cluster Abell S1063.
Fig. 1

NASA, ESA, Jennifer Lotz (STScI).

The blue arcs are highly magnified images of galaxies in the background of the cluster. Abell S1063 and the two clusters studied by Rodney et al.1 and Kelly et al.2 were imaged as Hubble Frontier Fields.

Modelling the lensing of stars by galaxy clusters is complicated because the deflection of light alters the position at which we detect the lensed star. Rodney et al.1 discovered two transients in a galaxy that lies behind the MACS J0416.1–2403 cluster of galaxies. The team modelled the effects of lensing by the foreground cluster and found that, even though the two transients — they dub them Spock events — appeared in different parts of the background galaxy, they could have been emitted at different times from one single system. This is one of the special features associated with lensing. To assess the lensing hypothesis, the team explored the alternatives, asking what types of known stellar variables would produce multiple events with the peak luminosity and time durations of the observed events. Supernovae are the most common types of bright transients observable at great distances, but the events observed by Rodney et al. do not fit any supernova model. Several interesting models do work equally well, including possible progenitors of supernovae. For example, the lensed star could be a luminous blue variable, a massive supergiant star on its way to a core-collapse supernova explosion. The end result would be a neutron star or black hole5. Or the star could be a recurrent nova in which occasional explosions on the surface of a white dwarf can presage a later explosion of the entire white dwarf, a type Ia supernova6,7.

Because light from these transients was emitted when the Universe was just over 40% its present age, the detections illustrate that optical surveys alert to variability in lensed galaxies have the potential to explore processes that occurred at early times in the history of the Universe. Beyond comparing early and present-day variability of known types of objects, there is the exciting prospect of detecting new types of events, particularly those that are rare today but are likely to have been more common during early and more active epochs of galaxy assembly and merger. Examples include stellar collisions, planetary collisions with compact objects, accretion-induced collapse, and mergers of compact objects, including black holes8,9,10,11,12.

The two Spock events, however, can also be explained in another way: a background star with more or less constant luminosity could have experienced a gravitational microlensing event, in which light from a single background-galaxy star is deflected by individual masses (or by the combined effect of a relatively small number of masses) in the galaxy cluster. To picture this, imagine that light from the star in the distant galaxy is deflected by the overall pull of the galaxy-cluster lens and that as this deflected light travels through the cluster it happens to pass very close to one or more of the cluster galaxies’ stars, whose gravitational pull introduces additional deflection and magnification. The variability produced by microlensing is induced by the relative motion of the background star and the stars in the cluster galaxies.

A separate team, Kelly et al.2, discovered two transients in a galaxy lying behind a different cluster of galaxies, MACS J1149+2223. The authors found that gravitational microlensing is the most likely explanation for the flashes they discovered. They refer to the single star that has been lensed as MACS J1149 Lensed Star 1 (LS1). LS1 is magnified by an impressive factor larger than 2,000.

The exciting thing about microlensing is that it is sensitive to the small-scale mass distribution within the galaxy cluster responsible for the lensing. Kelly et al. explore the relevance of LS1 to tests of dark matter models, for example, to put limits on the population of primordial black holes. Their work demonstrates that individual short-time-scale microlensing events provide opportunities to probe the galaxy-cluster distribution of compact masses. This measurement is therefore complementary to lensing by the cluster as a whole, which allows us to model its large-scale distribution of mass.

These two investigations are important. Individually, and even more so in tandem, they open up a rich field for future discoveries. They show us that magnified stellar variability and short microlensing events can be realistically expected from galaxies behind clusters. In fact, the situation is even more interesting, because the galaxy whose star was apparently microlensed by MACS J1149 is also known to have produced a supernova that was lensed by the cluster. The supernova is referred to as SN Refsdal, named for Sjur Refsdal, who predicted the occurrence of lensed multiply imaged supernovae13,14,15.

Whatever we call the events discussed here, they point to a new frontier in which both the spatial and time-variability characteristics of gravitational lensing are important. The cosmic flashes discovered by Kelly et al. and Rodney et al. could be identified because the Hubble Space Telescope (HST), with its extraordinary angular resolution, has been observing some galaxy clusters several times. Slated to continue until mid-2021, HST is likely to obtain additional deep images of galaxy clusters. Its successor, the James Webb Space Telescope (JWST), launching in 2019, will observe light at longer wavelengths, and will also be sensitive to events with lower luminosities. HST and JWST will discover more transients, and in theory will establish event rates, extrapolating to compute frequencies of the rare events that are even brighter than those already detected.

Observations that are much more frequent can be made by ground-based all-sky surveys, a number of which are underway. The Large Synoptic Survey Telescope (LSST) is under construction, aiming to create a spatial map of strong and weak lensing effects across the sky, while at the same time sampling time-variability at a variety of cadences16. Although the events described by Rodney et al. and Kelly et al. are too dim to be detected by LSST and present-day surveys, some of the most interesting events associated with young galaxies should be brighter. Some of these events will help us explore the distributions of dark and dim masses in cluster galaxies. Others will give us a ringside view of highly energetic processes in young galaxies. Fascinating!


  1. 1.

    Rodney, S. A. et al. Nat. Astron. (2018).

  2. 2.

    Kelly, P. L. et al. Nat. Astron. (2018).

  3. 3.

    Wong, K. C. et al. Astrophys. J. 844, 127 (2017).

  4. 4.

    Zheng, W. et al. Nature 489, 406–408 (2012).

  5. 5.

    Dwarkadas, V. V. Mon. Not. R. Astron. Soc. 412, 1639–1649 (2011).

  6. 6.

    Thoroughgood, T. D., Dhillon, V. S., Littlefair, S. P., Marsh, T. R. & Smith, D. A. ASP Conf. Proc. 261, 77–78 (2002).

  7. 7.

    Patat, F. et al. Astron. Astrophys. 530, A63 (2011).

  8. 8.

    Abbott, B. P. et al. Phys. Rev. Lett. 116, 241102 (2016).

  9. 9.

    Shara, M. M. Phys. Rep. 311, 363–369 (1999).

  10. 10.

    Tremaine, S. & Zytkow, A. N. Astrophys. J. 301, 155–163 (1986).

  11. 11.

    Di Stefano, R., Fisher, R., Guillochon, J. & Steiner, J. F. Preprint at (2015).

  12. 12.

    Baron, E., Cooperstein, J., Kahana, S. & Nomoto, K. Astrophys. J. 320, 304–307 (1987).

  13. 13.

    Refsdal, S. Mon. Not. R. Astron. Soc. 128, 307–310 (1964).

  14. 14.

    Kelly, P. L. et al. Science 347, 1123–1126 (2015).

  15. 15.

    Kelly, P. L. et al. Astrophys. J. 831, 205 (2016).

  16. 16.

    LSST Science Collaboration: Abell, P. A. et al. Preprint at (2009).

Download references

Author information


  1. Harvard-Smithsonian Center for Astrophysics, Cambridge, MA, USA

    • Rosanne Di Stefano


  1. Search for Rosanne Di Stefano in:

Corresponding author

Correspondence to Rosanne Di Stefano.