Stellar astrophysics

Supernovae in the neighbourhood

Detailed measurements of radioisotopes in deep-sea deposits, plus modelling of how they reached Earth, indicate that many supernovae have occurred near enough to have potentially influenced evolution. See Letters p.69 & p.73

For more than half a century, astronomers have speculated that supernovae have occurred close enough to Earth to affect the planet, possibly contributing to mass extinctions or climate change. In this issue, two studies of isotopes produced by supernovae reveal events thought to have happened a few million years ago: Wallner et al.1 (page 69) report measurements of marine deposits that contain isotopes produced outside the Solar System, whereas Breitschwerdt et al.2 (page 73) model the transport of iron-60 and use this to calculate the explosion times and sites of the likely sources. It is now possible to ask key questions with some precision; for example, could these supernovae have had substantial effects on Earth's climate and organisms — and perhaps even a role in human evolution?

The past 500 million years or so have witnessed mass-extinction events of varying intensity. These include the major end-Cretaceous event that wiped out the dinosaurs and about half of the planet's other species 66 million years ago, and a moderate event at the end of the Jurassic period 145 million years ago. In 1954, it was suggested3 that a nearby supernova caused the greatest mass-extinction event, which occurred about 250 million years ago (end-Permian) and affected more than 90% of species on Earth. For decades afterwards, progress in research centred on estimates of how often such stellar events might happen and what effects, if any, there might be on Earth and the interstellar medium. Unsurprisingly, these studies lacked precision.

Breakthrough computations4 in the 1990s showed that, for moderately close explosions — those occurring within about 100 parsecs (a few hundred light years) of Earth, expected every million years or so — radioactive nuclides can be transported and deposited on our planet. The isotopes are carried by dust grains that are only weakly affected by magnetic fields or gas pressure. Several candidate isotopes are problematic when used as evidence of supernovae: for example, beryllium-10 is also produced by radiation impinging on Earth's atmosphere, and plutonium-244 measurements suffer from background contamination originating from the testing of nuclear weapons. Nevertheless, concrete comparisons of estimates with measurements became possible for the first time.

In 1999, the detection5 of iron-60 in the deep-ocean ferromanganese crust broke the dam and led to a flood of results. Iron-60 is an excellent supernova indicator because it is produced abundantly in many types of supernova; non-supernova channels would produce only up to one-tenth as much. Nonetheless, the detection would not have been possible without the atom-counting capability of accelerator mass spectrometry, which allowed researchers to separate different isotopes by mass. This was followed by the first detailed computation6 of the effects on Earth's atmosphere of photons and cosmic rays from supernova sources. That study6 established a 'kill distance' of roughly 8 parsecs, within which the effects of these high-energy particles on terrestrial biota would be catastrophic. Supernovae this close should, on average, occur as often as once every 800 million years or so7, occasionally interspersed with additional events due to γ-ray bursts.

Breitschwerdt et al. looked at the deposition history of iron-60 and the trajectories of stars from a likely precursor group — the Local Bubble, a region of hot gas that includes both the Solar System and the probable sites of the supernovae that formed the bubble (Fig. 1). They find that the nearest explosions were approximately 100 parsecs away. The authors highlight the main sources of uncertainly in their conclusions as being in the supernova yield of iron-60 and in its transport through the interstellar medium and deposition on Earth. Despite these uncertainties, their findings are an important proof of concept for a unified picture of the dynamics of the interstellar medium, of the dynamics of the stellar group that gave rise to the supernova, and of the subsequent deposition of isotopes on Earth. The general set of parameters described in the study falls within a range suggested in recent work7 on analytical estimates of deposition processes.

Figure 1: Simulations of natural fireworks on a grand scale.

Zentrum Astron. Astrophys., Tu Berlin, Michael Schulreich

Breitschwerdt et al.2 performed high-resolution numerical simulations of supernovae that formed the Local Bubble (foreground, right) and the Loop I superbubble (left). Located in the Milky Way, these bubbles are large regions of hot, low-density gas surrounded by shells of swept-up debris that the Solar System has traversed for the past 5 million to 10 million years. In the image, the central solid green patch near Earth (represented by a white star) shows the mass distribution of the iron-60 isotope associated with the two bubbles 2.3 million years ago. The authors' simulations of the trajectory of iron-60 expelled from the supernovae agree with measurements of the isotope in the deep-sea crust by Wallner and colleagues1.

By using accelerator mass spectrometry on three different deep-sea archives, encompassing four sediment cores, two iron–manganese crusts and two iron–manganese nodules, Wallner et al. greatly expand the amount of available data for iron-60, as well as for aluminium-26 and beryllium-10. They find iron-60 levels averaging at about 40 times background levels. These results favour supernovae as the source, in two main events: one at 1.7 million to 3.2 million years ago, and the other 6.5 million to 8.7 million years ago. Both events were longer in duration than could be explained by the passage of a single blast wave. Wallner and co-workers therefore propose that either a series of supernova events directly affected the Solar System, or that the Solar System passed through an interstellar medium polluted by the mixed products of multiple supernovae.

Either the direct or indirect picture is reasonable, given that the rate of supernovae is expected to vary significantly in the different environments traversed by the Solar System during its orbit about the Galaxy. The data provided by Wallner et al. will allow considerable refinement of models such as those of Breitschwerdt and colleagues. It will also facilitate research into several puzzling characteristics of the cosmic-ray spectrum (cosmic rays are mostly accelerated protons) that can be explained8 if a supernova occurred roughly 2 million years ago at the same distance from Earth (roughly 100 parsecs) as the supernovae detected by Breitschwerdt and colleagues.

What can we discern about Earth's violent past from this broadly unified picture? The recorded supernovae occurred beyond the kill distance, and no major global mass extinctions9,10 coincide with them. However, during the suggested interval there was a general decline in temperature that culminated in the extensive series of glaciations in the Pleistocene epoch (from about 2.6 million to 12,000 years ago), although we do not know if there is a link between supernova activity and colder temperature. This climatic variation may be one of the conditions that led to human evolution. Ionization of the atmosphere by supernovae may also lead to an increase in lightning11 and possibly other climatic effects. The new studies1,2 will open up unexplored avenues of modelling and detailed investigation, providing deeper insight into what might have happened on Earth over the past 10 million years as a result of nearby stellar fireworks.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Wallner, A. et al. Nature 532, 69–72 (2016).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Breitschwerdt, D. et al. Nature 532, 73–76 (2016).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Schindewolf, O. H. Neues Jb. Geol. Paläontol. 10, 457–465 (1954).

    Google Scholar 

  4. 4

    Ellis, J., Fields, B. D. & Schramm, D. N. Astrophys. J. 470, 1227–1236 (1996).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Knie, K. et al. Phys. Rev. Lett. 83, 18–21 (1999).

    ADS  CAS  Article  Google Scholar 

  6. 6

    Gehrels, N. et al. Astrophys. J. 585, 1169–1176 (2003).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Fry, B. J., Fields, B. D. & Ellis, J. R. Astrophys. J. 800, 71 (2015).

    ADS  Article  Google Scholar 

  8. 8

    Kachelrieß, M., Neronov, A. & Semikoz, D. V. Phys. Rev. Lett. 115, 181103 (2015).

    ADS  Article  Google Scholar 

  9. 9

    Bambach, R. K. Annu. Rev. Earth Planet. Sci. 34, 127–155 (2006).

    ADS  CAS  Article  Google Scholar 

  10. 10

    Melott, A. L. & Bambach, R. K. Paleobiology 40, 177–196 (2014).

    Article  Google Scholar 

  11. 11

    Erlykin, A. D. & Wolfendale, A. W. Surv. Geophys. 31, 383–398 (2010).

    ADS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Adrian L. Melott.

Related links

Related links

Related links in Nature Research

Astrophysics: Supernova seen through γ-ray eyes

Astrophysics: Super-luminous supernovae on the rise

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Melott, A. Supernovae in the neighbourhood. Nature 532, 40–41 (2016).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.