I am writing a science fiction novel, where a ship is stranded in a single star system (a red supergiant). One of the plot points is the star becoming supernova in several hours, so the characters have to fix their ship before that happens.

I have basic knowledge of how it works: Iron generated from nuclear fusion gets accumulated in the core, until it reaches a point when iron fusion starts. As iron fusion is an endothermic reaction, the core is no longer able to generate enough energy to hold against its own gravity and external layers pressure, so it collapses, and explode.

I have read that once the iron fusion starts inside the core, the collapse occurs within minutes, that the collapse itself lasts a few seconds (even less than a second), and that the shockwave takes several hours to reach the surface. Is all that correct?

The thing is that I need the characters to bee able to predict the explosion in a short term. A few hours or even minutes. It would be great if they could be aware of the core collapse and start a countdown.

So, are there any external cue of these events, like changes in luminosity or color? Does the star spectrum change when iron fusion starts, or when the core collapses? I know that the core collapse generates a huge amount of neutrinos. Is this amount so intense that it can be easily detectable? (that is, without a huge detector in an underground facility). Can the amount of iron in the core be estimated from tha star spectrum and size, so the aproximate time of the collapse could be predicted?

  • 5
    $\begingroup$ The earliest warning you could get of an impending core-collapse supernovae would come from neutrinos because they react very-weakly with matter, however this is also why such large detectors are needed to measure their presence so its a catch 22 situation. EM clues are there but they are on much shorter timescales. $\endgroup$
    – Dean
    Commented Sep 21, 2016 at 14:37
  • 5
    $\begingroup$ This science of this question is probably fine for this site. However, the associated story issues and most potential follow-on questions may be more appropriate on Worldbuilding. $\endgroup$
    – Makyen
    Commented Sep 21, 2016 at 17:24
  • 5
    $\begingroup$ @Makyen Actually, on Worldbuilding there is a highly related question at How can we extinguish a supernova? The title is a little misleading -- it's more about preventing the star from going supernova late in the process, than about extinguishing or reducing it once it's started -- but I do think that question will be useful for the OP. $\endgroup$
    – user
    Commented Sep 21, 2016 at 19:46
  • 3
    $\begingroup$ @Dean A geiger counter will suffice for your neutrino detector in this case. On the other hand, if you're too close it's simply going to tell you that you're dead: what-if.xkcd.com/73 $\endgroup$ Commented Sep 22, 2016 at 3:39
  • 3
    $\begingroup$ @Mehrdad see physics.stackexchange.com/questions/63558/… but perhaps more specific to your question physics.stackexchange.com/questions/194606/… $\endgroup$
    – ProfRob
    Commented Sep 22, 2016 at 10:06

4 Answers 4


I think your best bet would be detecting neutrinos generated by nuclear burning inside the star (as we do for the Sun). Once the star hits the carbon-burning stage, it's actually putting out more energy in neutrinos than in photons. During the silicon-burning phase, which lasts for a few days and is what creates the degenerate iron core (that collapses once it is massive enough), the neutrino flux increases to about 1047 erg/s a few seconds before core collapse. (The peak flux during core collapse is about 1052 to 1053 erg/s). This paper by Asakura et al. estimates that the Japanese KamLAND detector could detect the pre-supernova neutrino flux for stars at distances of several hundred parsecs, and provide advance warning of a core-collapse supernova several hours or even days in advance. Since your characters are in the same system as the star, they'd hardly need a large underground detector to pick up the neutrinos.

This plot shows an example of neutrino luminosity (for anti-electron neutrinos) versus time for a pre-supernova star (from Asakura et al. 2016, based on Odrzywolek & Heger 2010 and Nakazato et al. 2013); core collapse begins at t = 0s.

Figure 1 from Asakura et al. 2016

By measuring the spectrum of energies for different types of neutrinos and their time evolution, you could probably get a very good idea of how far along the star was, particularly as we can probably assume your characters have much better models for stellar evolution than we currently do. (They'd also want to get accurate measurements of the star's mass, rotation rate, maybe internal structure via astroseismology, etc., in order to fine-tune the stellar-evolution model; these are all things they could do pretty easily.)

The core collapse itself would be signalled by the enormous increase in neutrino flux.

This "What If" article by Randall Munroe estimates that the neutrino flux from a core-collapse supernova would be lethal to a human being at a distance of around 2 AU. Which, as he points out, could actually be inside of a supergiant star, so your characters would probably be a bit further away than that. But it does show that the neutrino flux would be easily detectable, and that your characters might well get radiation poisoning from it if they were closer than 10 AU. (Of course, you'd want to detect it more directly than just waiting around till you started to feel sick, since that might take longer than the shock wave takes to reach the surface of the star.) This is just to bring home the fact that they wouldn't have any problem detecting the neutrinos....

  • 1
    $\begingroup$ Great answer! @Alfonso It might help your realism to address the fact that neutrinos are notoriously hard to detect. Some simple line implying that neutrino detection capabilities have increased ten-fold for your future-tech would help sell the realism that you can actually detect neutrinos on a small space ship without something like current neutrino detectors (which are massive). $\endgroup$
    – zephyr
    Commented Sep 21, 2016 at 15:43
  • $\begingroup$ Thank you very much, Peter. That's exactly the info I needed. $\endgroup$ Commented Sep 21, 2016 at 15:51
  • 5
    $\begingroup$ @zephyr The volume of the neutrino detector can be scaled down by the relative flux of neutrinos you expect. You don't need a large detector to observe the neutrino pulse if you are in orbit around the star that blows up. $\endgroup$
    – ProfRob
    Commented Sep 21, 2016 at 16:37
  • 7
    $\begingroup$ @zephyr From Table 6 of the paper I linked to, let's say KamLAND can detect pre-SN flux 10 hours before core collapse for a star 150 pc away (after a 48 hour integration). The neutrino flux for a spaceship 100 AU from the star is about 100 billion times higher. So, as Rob Jeffries pointed out, you could use a detector 100 billion times smaller than KamLAND. To cut the integration time to a more practical 1 second, you need a detector 500,000 times smaller than KamLAND: about 2 kg of liquid scintillator. (Assuming no better technology.) $\endgroup$ Commented Sep 21, 2016 at 17:01
  • 1
    $\begingroup$ Even without actual neutron detector, that neutrino flux might deposit enough energy to the star that it would be detectable from stars surface layers, or show as specific kind of noise in other detectors (at least cameras, radars and radios) the ship must have. $\endgroup$
    – hyde
    Commented Sep 21, 2016 at 20:13

Other answers are correct; a neutrino pulse is definitely expected as a result of a core-collapse supernova and should occur some hours before a shockwave arrives at the surface.

There essentially would be no visible sign that the star was about to become a supernova and that is because the dynamical timescale of the envelope is relatively long - thus responding slowly with respect to changes in the core. So, even if all support is removed from the centre, the surface can only respond (at best) on a freefall timescale of $\sim (G \rho)^{-1/2}$, where $\rho$ is the average density. If the star is a $10M_{\odot}$ supergiant with a radius of 1 au, then this timescale is tens of days.

Another possibility not mentioned so far is gravitational waves. Assuming that a relatively portable gravitational wave detector was available (!) then you would also expect a sharp gravitational wave pulse on the core collapse timescale (a second or less) that would also presage the supernova blast wave some hours later.

  • $\begingroup$ With respect to the gravitational wave: Would a non-rotating collapsing star emit gravitational waves? From another discussion I understood (perhaps in error) that rotation was a requirement. Something with quadrupole ... $\endgroup$ Commented Sep 23, 2016 at 6:53
  • $\begingroup$ @PeterA.Schneider Supernovae are fully expected to be GW sources, because there is expected to be asymmetry ( the core will be rotating very fast - see pulsars). eg. iopscience.iop.org/article/10.1086/381360/fulltext/… $\endgroup$
    – ProfRob
    Commented Sep 23, 2016 at 6:58
  • $\begingroup$ @PeterA.Schneider Real supernovae turn out to be highly asymmetric, even when their progenitors display a high degree of spherical symmetry. This asymmetry may be sufficient to produce gravitational waves, although I've not crunched the numbers. Finding the source of this asymmetry is an active topic of research. $\endgroup$
    – Calchas
    Commented Sep 23, 2016 at 12:48

As Dean said, supernova progenitors typically release neutrinos prior to full core collapse, remnant formation and the ejection of the outer layers of the star. The process - focused here on the neutrinos - goes something like the following:

  1. At high enough densities ($\rho\sim10^9\text{ g/cm}^3$), electron capture becomes important, where a proton and electron combine to form a neutron and an electron neutrino: $$e^-+p\to n+\nu_e$$ Simultaneously, beta decay may occur, where a neutron decays to a proton, electron and electron antineutrino: $$n\to p+e^-+\bar{\nu}_e$$ However, beta decay becomes less important than electron capture at this point.
  2. Electron capture reduces electron degeneracy pressure in the core, which leads to accelerated core collapse. Degeneracy pressure is important in the cores of many stars, but in extremely massive stars - red supergiants included - it simply isn't enough to stop the collapse.
  3. At densities below $\sim10^{11}\text{ g/cm}^3$, neutrinos can carry away energy, and the initial burst leaves the star within about ten seconds. However, core collapse quickly leads to much greater densities, and when $\rho\sim4\times10^{11}\text{ g/cm}^3$, neutrinos are trapped. They scatter off nuclei, and transfer energy to electrons. Electron-nuclei scattering is also important, and may be dominant at higher energies.
  4. At $\rho\sim2.5\times10^{14}\text{ g/cm}^3$, the core undergoes a "bounce", and the supernova explosion fully begins. A shock wave propagates into the outer core, and more neutrinos are produced via electron capture.
  5. Neutrinos still trapped in/by the stellar remnant are released about ten seconds later. Neutrino pair production, too, leads to rapid cooling. Some of these neutrinos may contribute to a revival of the shockwave.

Neutrinos may arrive hours - or possible days, in some circumstances - before the light from the supernova. The former was the case for SN 1987A, the first supernova from which neutrinos were detected.


  • $\begingroup$ Thank you, @HDE. That detailed answer makes me wonder if the core collapse could be accelerated with the introduction of something "new" in the star. In the novel, there is a battle with an unknown hostile alien ship. That ship is disabled and put into a falling trajectory to the star. If there is a way in with that event could shorten the time to explosion, it would add more drama (the core collapse is detected by the characters before expected). I remark that the falling ship is alien and unknown, so there is no problem to include antimatter, strange matter, or wathever needed. $\endgroup$ Commented Sep 21, 2016 at 16:13
  • 2
    $\begingroup$ @AlfonsodeTerán, Even assuming some type of unobtainium, it is not believable, even in a fictional environment, that the addition of anything that is on the scale of a spaceship would have sufficient effect when unintentionally falling into a star. The difference in scale is just too vast. Writing it such that it did so would almost certainly destroy my suspension of disbelief if I were reading such a story. Even assuming the ship was powered by a micro-black hole, the time it would take for the singularity to significantly affect the star is longer than what you have been describing. $\endgroup$
    – Makyen
    Commented Sep 21, 2016 at 17:23
  • $\begingroup$ @AlfonsodeTerán Mayken's absolutely right. You'd need quite advanced technology to even begin to contemplate having a chance at impacting the evolution of a star like this, if that's even possible. Its effects would be like a drop in the metaphorical bucket. $\endgroup$
    – HDE 226868
    Commented Sep 21, 2016 at 17:25
  • 1
    $\begingroup$ @Joshua unless the black hole is very much not micro (~$1M_{\odot}$) we're talking years at least. black holes are limited in how fast they can eat by interactions between particles in the accretion disk. $\endgroup$
    – Leliel
    Commented Sep 21, 2016 at 23:47
  • 1
    $\begingroup$ @Spike0xff If the lake is supercool and the drop is impure it may trigger crystallization. I don't think such phase-shifts occur in stellar cores in reality but as SF one could have hell freeze over... $\endgroup$ Commented Sep 23, 2016 at 7:02

A superluminous supernova (aka hypernova) can exhibit a double peak to its brightness and some are theorizing that this may be the norm for a superluminous supernova, though as far as I know it's only actually been observed in one case so far (DES14X3taz).

Anyway, in (at least) this case there was an initial substantial increase in brightness. Then the brightness dropped (a couple of magnitudes) for a few days, then ramped back up to considerably brighter than the initial "bump".

You probably are going to need to be careful about the distances involved. The initial burst of light is already large enough that unless your people are quite a ways away, it'll already be enough to fry them to a crisp.

There is one other point that might be interesting for your novel though. After the explosion, what you probably get is a magnetar--which, as you'd guess from the name, is a star with an extremely strong magnetic field--so strong, in fact, that it's likely to cause all sorts of havoc with anything in the vicinity that depends on anything involving electrical activity--not only electronics, but also probably people's nerves as well.

There is an obvious problem here though: a red supergiant is the right type of star as the progenitor for a "normal" supernova. It's probably not the right type as the progenitor for a superluminous supernova. The progenitor of a supernova is typically something like six or eight solar masses. A superluminous supernova is probably (only a few are known, so it's hard to generalize) something like a couple hundred solar masses. Given the amount of energy released, it has to be quite large anyway.

Reference: Smith, et al (2015)


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .