www.nature.

com/scientificreports

OPEN The Resilience of Life to

Astrophysical Events
1
David Sloan , Rafael Alves Batista   1 & Abraham Loeb2

Much attention has been given in the literature to the effects of astrophysical events on human and
Received: 18 January 2017
land-based life. However, little has been discussed on the resilience of life itself. Here we instead explore
Accepted: 5 June 2017 the statistics of events that completely sterilise an Earth-like planet with planet radii in the range 0.5–
Published: xx xx xxxx 1.5R⊕ and temperatures of ∼300 K, eradicating all forms of life. We consider the relative likelihood of
complete global sterilisation events from three astrophysical sources – supernovae, gamma-ray bursts,
large asteroid impacts, and passing-by stars. To assess such probabilities we consider what cataclysmic
event could lead to the annihilation of not just human life, but also extremophiles, through the boiling
of all water in Earth’s oceans. Surprisingly we find that although human life is somewhat fragile to
nearby events, the resilience of Ecdysozoa such as Milnesium tardigradum renders global sterilisation an
unlikely event.

Within the coming years spectroscopy will likely be employed to identify molecules that are indicative of life in
the atmosphere of exoplanets1. In the context of the search for extraterrestrial life, it is useful to establish the nec-
essary conditions for life to be present for such observations. Broadly speaking this relies upon two ingredients.
The first is an unknown quantity–the fraction of planets on which life begins. The causes of the emergence of life
on Earth are not understood, and thus we do not have a complete theory for predicting where life may begin else-
where. The second is the probability that life has persisted from its inception to observation. In this work we will
show that this is highly likely, as events which could lead to life being completely eradicated are rare. To establish
this we break from the usual study in the literature2–6 of the possible paths to ending human life, and broaden the
analysis to consider those astrophysical events which could rather remove all life by analysing the most resilient
of species–tardigrades.
Tardigrades can survive for a few minutes at temperatures as low as −272 °C or as high as 150 °C, and −20 °C
for decades7, 8. They withstand pressures from virtually 0 atm in space9 up to 1200 atm at the bottom of the
Marianas Trench10. They are also resistant to radiation levels ∼5000–6200 Gy11. For complete sterilisation we
must establish the necessary event to kill all such creatures.
We consider three types of astrophysical events which could constitute a threat to the continuation of our
chosen life forms: large asteroid impact, supernovae (SNe), and gamma-ray bursts (GRBs). GRBs and SNe can be
deadly due to the lethal doses of radiation and in particular the shock wave associated with the burst. Radiation
can cause the depletion of the ozone layer, removing the shield that protects us from cosmic radiation2, 12, 13.
The effects of gamma-ray bursts (GRBs) on humans and land-based life could be disastrous as the eradication
of the ozone layer would leave us exposed to deadly levels of radiation2. However, in such circumstances life could
continue below the ground. Significantly, several marine species would not be adversely affected, as the large body
of water would provide shielding. Even the complete loss of the atmosphere would not have an effect on species
living at the ocean’s floor. The impact of a large asteroid could lead to an “impact winter”, in which the surface of
the planet receives less sunlight and temperatures drop. This would prove catastrophic for life dependent on sun-
light, but around volcanic vents in the deep ocean life would be unaffected. Similarly, an increase in pressure, or
acidity spread across the entirety of the (deep) ocean is an unlikely scenario for extinction. The physical processes
by which ocean pressure could significantly increase involve increasing planetary mass; such impacts would first
lead to extreme heating. Even following extreme events, spreading acidity through the entire ocean is unlikely.
The removal of the atmosphere would also lead to mass extinction. However, following such an event the remain-
ing ocean water would form a new atmosphere below which oceans could still form. The energy requirements for
total sterilisation of the planet through atmospheric removal are significantly greater than those for boiling the

1
Department of Physics - Astrophysics, University of Oxford, Denys Wilkinson Building, Keble Road, OX1 3RH, Oxford,
UK. 2Astronomy Department, Harvard University, 60 Garden Street, Cambridge, MA, 02138, USA. Correspondence and
requests for materials should be addressed to R.A.B. (email: rafael.alvesbatista@physics.ox.ac.uk)

Scientific Reports | 7: 5419 | DOI:10.1038/s41598-017-05796-x 1

www.nature.com/scientificreports/

oceans, so the threat of atmospheric removal is contained within that of oceans boiling. We are therefore led to
consider death due to heat or radiation.

Analysis
To raise the entire ocean temperature by T requires a deposit of E = MoT wherein Mo is the ocean mass and 
the specific heat capacity of water. In order to increase the temperature of the entirety of the Earth’s oceans we
need to introduce a large amount of thermal energy. The total mass of water in the oceans is around 1.35 × 1021
kg. The specific heat capacity of water is 4184 J kg−1 °C−1 so we require 5.6 × 1024 J to raise the ocean temperature
by 1 °C. Thus the tardigrade with a tolerance of up to 100 °C would survive until around 5.6 × 1026 J were depos-
ited into the ocean. This is a lower bound–such heat would not be evenly distributed, being it most likely to be
deposited in the upper ocean. To provide a conservative bound, we seek to minimise the depth of the deepest
ocean on any planet–a uniform distribution of oceans across the planet’s surface. When ocean mass is small com-
pared to that of the planet, the depth of the ocean is approximately

αρ 2/3M1/3
p
D= .
(36π )1/3 ρw (1)
Here ρ = 3M p /(4πR p 3 ) is the average planet density, α the fraction of the mass in ocean (on Earth
Mo ≈ 2.3 × 10−4M⊕) and ρw the density of water. Most of Earth’s water is contained within rocks. To remain con-
servative, we consider only the mass of liquid water in the oceans. There may exist planets that are almost entirely
water (α ≈ 1), however for life as we know it, we focus on Earth-like planets with oceans on the surface of a rocky
planet. We give these explicitly as we will assume they are broadly unchanged between planets. For the Earth, this
implies that there must be an ocean of at least 2.5 km in depth. This is far shallower than the deepest points, how-
ever it will constitute a lower bound. The intensity of gamma rays is attenuated by interaction with matter by a
factor exp(−μD), wherein D is the depth and μ the attenuation coefficient. This varies based on the material and
the frequency of the incident radiation. The tardigrade is capable of withstanding over 6000 Gy (enough to endow
every kilogram of material with 6000 J of energy). If the ocean depth is greater than log (700)/µ (the latter figure
being the ratio of the energy deposit per unit mass required to boil water to that to kill a tardigrade) the water
above will be boiling. In fact, if we consider a sufficient radiative flux to kill a tardigrade at depth D, the total
energy deposited upon the planet is at least E = 6000πR p2(e µD − 1)/µ. If our oceans are more than a few metres
deep, this exceeds the threshold energy at which the oceans would boil before radiation would kill the tardigrade.
We therefore consider temperature increase as the primary source of sterilisation.
Large asteroids are the leading candidate for causing of the Cretaceous-Tertiary extinction which took place
65 million years ago, annihilating approximately 75% of species on the planet leaving the Chicxulub crater. This
event devastated larger land animals. Of those with masses over 25 kg only a few ectothermic species survived.
However, around 90% of bony fish species survived14 and deep ocean creatures were largely unaffected by the
event. We estimate an upper bound for the energy deposited by an asteroid of mass Ma as being its free-fall energy
from infinity to the surface of the planet E = 1/2Ma(v∞2 + ve2), where ve = 2GMp/R p is the escape velocity of the
planet (ve ≈ 11.2 kms−1 for Earth), and v∞ is given by Öpik’s close encounter theory15. In order to raise the ocean’s
temperature by T, we require an asteroid of mass
2αT
Ma = 2
Mp.
v∞ + ve2 (2)
To annihilate tardigrades on Earth we require a mass over ∼1.7 × 10  kg. The largest observed asteroids in
18

the Solar System are Vesta and Pallas, with masses of 2.7 × 1020 kg and 2.2 × 1020 kg respectively. There are only
17 other known asteroids of sufficient mass, and a few dwarf planets, the most massive ones being Eris and Pluto,
whose masses are 1.7 × 1022 kg and 1.3 × 1022 kg respectively. We reiterate that our estimate of the required energy
is conservative–it is likely that it would take a significantly more massive impact as ocean heat would only be a
fraction of the total energy. Since we consider Earth-like planets, the order of magnitude of this mass does not
vary greatly between the largest and smallest planets–if oceans constitute an equal fraction of mass this changes
by less than an order of magnitude.
In Fig. 1 we present a model for the impact rate of asteroids as a function of the mass. This is based on the
extrapolation of relation between crater diameter and impact rate16. The mass of the object is related to crater
diameter following ref. 17, assuming asteroids with density ρ ~ 5 g cm−3 entering the atmosphere with incidence
angle of 90° with respect to the normal. In reality, most asteroids have ρ ≈ 2 g cm−3, and in the case of comets
this value is even lower, being the value here assumed a conservative assumption. This is highly dependent on
the asteroid distribution in our Solar System–we assume in the absence of other evidence, that other systems are
similar, however this remains to be verified.
The bulk of the energy output of a supernova is carried by the shock wave. To give an upper bound on the
range at which a supernova would remove all life from the planet, we assume that the shock wave carries all the
energy released. The fraction of energy incident on a planet of radius Rp at a distance d from the supernova is
given by the fraction of the sphere of radius d which is covered by the planet’s surface πR p2 /(4πd 2). To raise the
temperature of a planet by T, we would require a supernova within a distance d given:

Scientific Reports | 7: 5419 | DOI:10.1038/s41598-017-05796-x 2

www.nature.com/scientificreports/

Figure 1.  The cumulative impact rate per year for different masses of asteroids. This follows the
parameterisation given in ref. 25 up to M = 1015 kg; for M > 1015 kg the impact rate is inferred following ref. 16.
Dashed vertical lines indicate the minimum mass needed for complete sterilisation assuming a typical asteroid
with density (ρ = 2000 kg/m3). The lower bound (m1) is that which could cause boiling of the oceans if the
entirety of its energy were converted into heat spread homogeneously throughout the Earth’s oceans. The upper
bound (m1) is the mass of an asteroid whose impact crater is equal to the size of the planet, causing complete
destruction. Here we find that even with the most conservative bound, the likelihood of complete sterilisation is
lower than around 10−5 over the lifetime of the planet.

1
 3 3  E 21
   s 
dSN =    
 32πM1/2
p ρ   αT  (3)
For the Earth, this sterilisation distance is around 0.04 pc, far closer than the closest stars, Proxima Centauri.
Were a supernova to occur at that distance, the ocean temperature would only rise by about 0.1 °C. Furthermore,
although there is a dependence on the mass of the planet, this dependence is quite weak. Note that none of the stars
in the Alpha Centauri system are large enough to go supernova. The nearest potential supernova is the IK Pegasi
system, approximately 45 pc away, which is three orders of magnitude farther than the estimated sterilisation radius.
To assess the relative risk faced by any planet in our galaxy, we approximate the odds of a close enough super-
nova happening over a timespan of 109 years. We find the expected number of stars of sufficient mass within
the sterilisation distance of a planet, and the odds that one of these stars goes supernova. The galactic habita-
ble zones, regions wherein complex life may evolve, depends on the occurrence rate of supernovae. A detailed
simulation-based study was done by Lineweaver et al.18. We evaluate the rate of SN at a position (r, z) (cylindrical
coordinates) as follows:
M max
PSN (r, z ) = χ ∫M dm ξ(m)n⁎(r, z )τ −1(m),
min (4)
with n⁎ being the number density of stars . Following ref. 19 we select Mmin = 8M and Mmax = 25M for super-
19

nova progenitors. ξ(m) and τ−1(m) are respectively the initial mass function and the lifetime of a star of mass m.
This is normalised to the global supernova rate in the Milky Way20. The rate of supernovae explosions within the
sterilisation radius (0.04 pc) over 1 billion years is shown in Fig. 2, for differing galactic locations.
Because the nearest star is about 1.3 pc away, we can conclude that Earth is located in a fortunate position. Near
the galactic centre the density of stars increases and the probability of a SN sterilising life is higher. Nonetheless,
this rate is almost insignificant even close to the galactic core, reaching only around 1% of planets being sterilised.
The calculation for GRBs is similar to the one for SNe, but now we have to assume that the energy is collimated
into in jets. As before, we assume the most pessimistic scenario–the smallest jet angle with the largest energy. The
energy is typically the same as that of a supernova, 1044 J, but the jet angles can be as low as 2°, hence the energy
incident on a target of radius r which lies entirely within the beam angle at a distance d is:

πr 2 1062 r 2
E = EGRB 2
= J
Ωd d2 (5)
Hence for an increase of 100 °C in the ocean temperatures, we would need a GRB within about 13.8 pc; again,
this is an upper limit. The rate of occurrence of short GRBs per volume in the universe is 0.04 Gpc−3 yr−1, and
long GRBs is 0.15 Gpc−3 yr−1 6. We will restrict these occurrences to within galactic discs of stars, therefore we
divide this by the product of the comoving number density of galaxies (≈107 Gpc−3) and the volume occupied
of the galactic disc (1011pc3), we find that the rate is around 2 × 10−10 pc−3 Gyr−1, and hence the probability of a
GRB within the a distance at which it would sterilise a planet, aligned such that one of the beams hit the planet is
3.2 × 10−10 Gyr−1. This number is extremely small and we can conclude that such event is unlikely.

Scientific Reports | 7: 5419 | DOI:10.1038/s41598-017-05796-x 3

www.nature.com/scientificreports/

Figure 2.  The expected number of supernovae within the sterilisation sphere of radius 0.04 pc per Gyr as a
function of galactic position. The black circle indicates the position of the Solar System. Closer to the galactic
centre the stellar density is higher, and thus the likelihood of encountering a nearby supernova increases.
However, this density is only sufficient to give a total rate of around 0.01 expected events per billion years, and
thus total sterilisation through supernovae is still an improbable event.

Planetary systems can also be disrupted by passing-by stars. The rate, R, of such encounters for a given stellar
density, n∗, is
R = n⁎σv, (6)
where σ is the cross section to disrupt the orbit of the planet, and v the velocity of the star. Typically, v ≈ 40 km s−1.
In the neighbourhood of the Solar System n⁎ ∼ 10−3 pc−319. In the case of Earth, σ ∼ 10−9 pc221. Therefore, the rate
of interactions of stars and the Earth-Sun system would be R ∼ 3 × 10−8 Gyr−1. If we repeat this calculation for the
average stellar density in the galaxy, which is ∼0.1 pc−3, the rate would be increased to ∼10−6 Gyr−1. This number is
an upper bound–we expect that only a fraction of systems that experience disruption would eject a planet–yet it is
still extremely small and we can conclude that ejection by this mechanism is a very rare event.

Discussion
Our analysis has focussed on providing an absolute upper bound for the rate of complete sterilisation of an
Earth-like planet during its evolution, by considering the required events that would lead to the death of the har-
diest species on Earth. With such assumptions, we find that the probability is less than 10−7 per billion years. The
overall likelihood of complete sterilisation is small even for planets which could exist around dwarf stars for ten
trillion years, the most likely time for life to find itself22. For asteroids the impact rate of deadly objects is
10−5 Gyr−1. Eventually the evolution of the host star will render a planet sterile, either through expansion to the
point of oceans boiling, or through a post-collapse freezing. There is a third scenario, where life continues around
geothermal vents on a rogue planet until capture by a new host system, or the source of heat is extinguished. The
time scale for the former is conservatively bounded by the rate of disruption of planetary systems by wandering
stars, being therefore ∼30 Myr in regions with density of stars comparable to the vicinity of the Solar System; the
time scale for the latter depends on properties of the planet, but are of the order of billion years. Consequently, life
could perdure on a rogue planet long enough for it to be recaptured. We do not fully understand the mechanisms
by which life started, but once it exists on an Earth-like planet, the complete removal of all life (other than through
evolution of the host star) is a very unlikely event.
In this analysis we have considered only direct effects of astrophysical events such as the direct boiling of
oceans; however such events could serve as a trigger for a second change. An example of this is the runaway
greenhouse effect23 in which a less energetic event could evaporate a fraction of the ocean, which in turn leads
to increased temperatures. Such an occurrence would have a lower threshold energy, as total evaporation of the
oceans is no longer necessary; however the precise details of such an event are not known. As an upper bound, we
can presume that this would have to increase global temperatures by a larger amount than seasonal variations and
changes in solar output. In essence this would reduce our energy threshold by about two orders of magnitude, still
leaving such an occurrence very rare.
Throughout this analysis we have made several assumptions, the primary being that life elsewhere will be
similar to that found on Earth. To justify this assumption, we consider that when searches are conducted for
exoplanetary life, the template used is that of life on Earth. We make the further assumption that life will evolve to
adapt to the extreme environments of exoplanets as it has to those on Earth. Again, we justify this by the ubiquity
of life across environmental conditions.
Finally, we note that the type of life we expect to survive all but the most extreme of events is that which could
survive elsewhere in our own Solar System. The history of Mars indicates that it had an atmosphere24 that could
have supported life, albeit in conditions that on Earth would be considered inhospitable, with both the Mars 2020
and ExoMars programs aiming for experimental verification. Organisms with similar tolerances to radiation and
temperature such as tardigrades would be the only kind of life that could survive long-terms in such conditions,

Scientific Reports | 7: 5419 | DOI:10.1038/s41598-017-05796-x 4

www.nature.com/scientificreports/

and even then they would have to be significantly below the surface. The subsurface oceans that are posited to
exist on Europa and Enceladus would have conditions similar to the deep oceans of Earth where taridgrades are
found - volcanic vents providing heat in an environment devoid of light. The presence of extremophiles in the
locations would be a significant step in narrowing the question of life starting on exoplanets.

References
1. Seager, S., Bains, W. & Petkowski, J. J. Toward a List of Molecules as Potential Biosignature Gases for the Search for Life on Exoplanets
and Applications to Terrestrial Biochemistry. Astrobiology 16, 465–485, doi:10.1089/ast.2015.1404 (2016).
2. Ruderman, M. A. Possible Consequences of Nearby Supernova Explosions for Atmospheric Ozone and Terrestrial Life. Science 184,
1079–1081, doi:10.1126/science.184.4141.1079 (1974).
3. Chapman, C. R. Impacts on the Earth by asteroids and comets: assessing the hazard. 367, 33–40 10.1038/367033a0 (1994).
4. Dar, A., Laor, A. & Shaviv, N. J. Life Extinctions by Cosmic Ray Jets. Physical Review Letters 80, 5813–5816, doi:10.1103/
PhysRevLett.80.5813 (1998).
5. Beech, M. The past, present and future supernova threat to Earth’s biosphere. 336, 287–302 doi:10.1007/s10509-011-0873-9 (2011).
6. Piran, T. & Jimenez, R. Possible Role of Gamma Ray Bursts on Life Extinction in the Universe. Physical Review Letters 113, 231102,
doi:10.1103/PhysRevLett.113.231102 (2014).
7. Hengherr, S., Worland, M., Reuner, A., Brümmer, F. & Schill, R. O. High’ temperature tolerance in anhydrobiotic tardigrades is
limited by glass transition. Physiological and Biochemical Zoology 82, 749–755, doi:10.1086/605954 (2009).
8. Tsujimoto, M., Imura, S. & Kanda, H. Recovery and reproduction of an antarctic tardigrade retrieved from a moss sample frozen for
over 30 years. Cryobiology 72, 78–81, doi:10.1016/j.cryobiol.2015.12.003 (2016).
9. Jönsson, K. I., Rabbow, E., Schill, R. O., Harms-Ringdahl, M. & Rettberg, P. Tardigrades survive exposure to space in low earth orbit.
Current Biology 18, R729–R731, doi:10.1016/j.cub.2008.06.048 (2008).
10. Seki, K. & Toyoshima, M. Preserving tardigrades under pressure. Nature 395, 853–854, doi:10.1038/27576 (1998).
11. Hashimoto, T. et al. Extremotolerant tardigrade genome and improved radiotolerance of human cultured cells by tardigrade-unique
protein. Nature Communications 7, 12808, doi:10.1038/ncomms12808 (2016).
12. Ellis, J. & Schramm, D. N. Could a nearby supernova explosion have caused a mass extinction? Proceedings of the National Academy
of Sciences of the United States of America 92, 235 (1995).
13. Thorsett, S. E. Terrestrial implications of cosmological gamma-ray burst models. The Astrophysical Journal Letters 444, L53–L55,
doi:10.1086/187858 (1995).
14. Kriwet, J. & Benton, M. J. Neoselachian (chondrichthyes, elasmobranchii) diversity across the cretaceous–tertiary boundary.
Palaeogeography, Palaeoclimatology, Palaeoecology 214, 181–194 doi:10.1016/j.palaeo.2004.02.049. Mesozoic-Cenozoic Bioevents (2004).
15. Öpik, E. J. Interplanetary encounters: close-range gravitational interactions (Elsevier Scientific Pub. Co, 1976).
16. Hergarten, S. & Kenkmann, T. The number of impact craters on Earth: Any room for further discoveries? Earth and Planetary
Science Letters 425, 187–192, doi:10.1016/j.epsl.2015.06.009 (2015).
17. Collins, G. S., Melosh, H. J. & Marcus, R. A. Earth Impact Effects Program: A Web-based computer program for calculating the
regional environmental consequences of a meteoroid impact on Earth. Meteoritics and Planetary Science 40, 817,
doi:10.1111/j.1945-5100.2005.tb00157.x (2005).
18. Lineweaver, C. H., Fenner, Y. & Gibson, B. K. The Galactic Habitable Zone and the Age Distribution of Complex Life in the Milky
Way. Science 303, 59–62, doi:10.1126/science.1092322 (2004).
19. Jurić, M. et al. The Milky Way Tomography with SDSS. I. Stellar Number Density Distribution 673, 864–914, doi:10.1086/523619 (2008).
20. Tammann, G. A., Loeffler, W. & Schroeder, A. The Galactic supernova rate. The Astrophysical Journal Supplement Series 92, 487–493,
doi:10.1086/192002 (1994).
21. Laughlin, G. & Adams, F. C. The Frozen Earth: Binary Scattering Events and the Fate of the Solar System. Icarus 145, 614–627,
doi:10.1006/icar.2000.6355 (2000).
22. Loeb, A., Batista, R. A. & Sloan, D. Relative likelihood for life as a function of cosmic time. Journal of Cosmology and Astroparticle
Physics 8, 040, doi:10.1088/1475-7516/2016/08/040 (2016).
23. Goldblatt, C. Habitability of Waterworlds: Runaway Greenhouses, Atmospheric Expansion, and Multiple Climate States of Pure
Water Atmospheres. Astrobiology 15, 362–370, doi:10.1089/ast.2014.1268 (2015).
24. Jakosky, B. M. et al. MAVEN observations of the response of Mars to an interplanetary coronal mass ejection. Science 350, 0210,
doi:10.1126/science.aad0210 (2015).
25. Bland, P. A. & Artemieva, N. A. The rate of small impacts on Earth. Meteoritics and Planetary Science 41, 607–631,
doi:10.1111/j.1945-5100.2006.tb00485.x (2006).

Acknowledgements
D.S. and R.A.B. acknowledge the financial support from the John Templeton Foundation.

Author Contributions
The idea of this work was conceived by A.L., D.S. and R.A.B. contributed equally to the analysis of results, with
some input from A.L. D.S. did most of the writing, with aid of R.A.B. Figures were produced by R.A.B., A.L. was
responsible for the scope and accuracy checking of the analysis.

Additional Information
Competing Interests: The authors declare that they have no competing interests.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre-
ative Commons license, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons license and your intended use is not per-
mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

© The Author(s) 2017

Scientific Reports | 7: 5419 | DOI:10.1038/s41598-017-05796-x 5