Death by Supernova?

[Skills]

I saw Phil Plait (the bad astronomer) saying the odds of dying in a supernova explosion is 1 in 10,000,000

Sounds a bit high to me... is a death by supernova likely without a complete human extinction? Don't those odds imply a supernova extinction every 600mm years?

What actually causes the deaths? A newspaper article stated that the ozone layer would be stripped away leading to the earth being frazzled. Is it the light from the explosion that would do this? Does that mean that we would have no forewarning?

Thanks

Adrian

[Telrad]

IMO If there is a Supernova it would be too far away, most of its energy would be scattered or absorbed in the vast emptiness of space; so any type of gamma-ray burst, or a high-energy outburst expected with an eruption, would miss the Earth. Cosmic rays would be diffused by magnetic fields, and most of the damaging light would not affect life on Earth.

[George]

Gamma rays would frie you

[Telrad]

Gamma rays would frie you

IMO a supernova from a star prehaps 150 times bigger than our sun could have a devastating effect on our planet, even at 8,000 light years away; but the earth's magnetic field might deflect some of the cosmic rays and high-energy photons (gamma) from such an event?

[lukebl]

Does that mean that on a planet with, say, 8 Billion people, you would expect 800 to die by supernovas, but the rest to die by other means?

Sorry, being a bit mischevious as these odds mean absolutely nothing. Probability theory is all navel-gazing.

[Zakalwe]

Gamma rays would frie you

Depends.

A gamma ray burst is focused into a tight beam, emitted from the magnetic poles of a supernova. So you would have to have a nearby star go nova AND be in line with it's poles. The media have used Betelgeuse as a headline grabber...in reality, AFAIK, it's poles do not point at us, so any GRB from it's eventual supernova would not be directed at us.

[acey]

I saw Phil Plait (the bad astronomer) saying the odds of dying in a supernova explosion is 1 in 10,000,000

Where did you get the number, and what's it supposed to mean? One person in 10 million dies by supernova? (Obviously not). Chances of the Earth being fried at some point in the next hundred billion years is 1 in 10 million? (How would it be calculated?).

Can't find anything at Bad Astronomy - nearest things I could find were these:

No, a nearby supernova won’t wipe us out | Bad Astronomy | Discover Magazine

Is Betelgeuse about to blow? | Bad Astronomy | Discover Magazine

[Telrad]

Depends.

A gamma ray burst is focused into a tight beam, emitted from the magnetic poles of a supernova. So you would have to have a nearby star go nova AND be in line with it's poles. The media have used Betelgeuse as a headline grabber...in reality, AFAIK, it's poles do not point at us, so any GRB from it's eventual supernova would not be directed at us.

I agree Zakalwe, even though it is abundantly clear that Betelgeuse is far enough away to give us a beautiful display and not a roasting when it does go supernova, it seems the temptation is just too great for some doomsday theorists and tabloid writers!

[JulianO]

IMO a supernova from a star prehaps 150 times bigger than our sun could have a devastating effect on our planet, even at 8,000 light years away; but the earth's magnetic field might deflect some of the cosmic rays and high-energy photons (gamma) from such an event?

I think the gammas are the worst - as they travel at the speed of light. So the first we'll no about it is when they hit us. They won't be deflected by magnetic fields either - so will come straight through and hit the atmosphere and us. Basically it will depend on the flux density we receive.

The charged particles will travel far more slowly, so we'll have lots more warning, and we'll be protected by the magnetic fields somewhat, but probably not enough!

[Skills]

Here's where I got it from: NECSS of DEATH! | Bad Astronomy | Discover Magazine

Lyrics to the song if you don't want to listen:

Odds of dying by supernova explosion: 1 in 10,000,000

Supernovae happen about once per century in any given galaxy. Assuming the event would cause a mass extinction killing everyone on earth- the odds of you specifically dying from one during your lifetime are about 1 in 10 million.

[acey]

Cute song but the statistic is obviously wrong - billions of people have lived and died on Earth and no one has died from supernova explosion. The crucial thing is not whether a supernova happens in our galaxy, but whether it's close enough to do any harm - and that depends on there being a big enough star within a small enough distance.

"Recent estimates predict that a Type II supernova would have to be closer than eight parsecs (26 light-years) to destroy half of the Earth's ozone layer... The nearest supernova candidate is IK Pegasi (HR 8210), located at a distance of 150 light-years."

http://en.wikipedia.org/wiki/Supernova

The last known supernova in our galaxy exploded approximately 140 years ago (as seen from Earth) - though nobody noticed it at the time.

http://geology.com/nasa/youngest-supernova.shtml

The last to have been seen going off was in 1604.

http://en.wikipedia.org/wiki/SN_1604

[FraserClarke]

"Recent estimates predict that a Type II supernova would have to be closer than eight parsecs (26 light-years) to destroy half of the Earth's ozone layer... The nearest supernova candidate is IK Pegasi (HR 8210), located at a distance of 150 light-years."

Supernova - Wikipedia, the free encyclopedia

Well that about works out then doesn't it? Very back of the envelope calculations;

Galaxy is a flat disc 20,000 pc in radius. Total area is 5^9 sq pc. To kill us, a SNe has to be within 10pc, so the area is 1^3 sq pc.

If you make an assumption that SNe are homogenously distributed in the galaxy, one in ever 5^6 SNe go off "next" to the Earth.

If a SNe goes off once per hundred years in the galaxy on average, then a SNe goes off next to the Earth ever 5^8 years; i.e. every 500Myr.

So, if you live for 70 years. The chances that you're killed by a SNe is 70/5^8 =~ 1 in 10 million...

Now, it seems like at the minute, we're a in "low risk" SNe environment because no viable candidate is very near us -- but that is just our lucky draw Over the timescales of evolution, that will change as we shift about in the galaxy.

[acey]

a SNe goes off next to the Earth ever 5^8 years; i.e. every 500Myr.

So, if you live for 70 years. The chances that you're killed by a SNe is 70/5^8 =~ 1 in 10 million...

Your 500Myr seems to be a good estimate - Wikipedia's figure is once every 240 million years. But you can't then conclude that the probability of a presently living person dying by supernova is 1 in 10 million.

By dividing by 70 you're saying that the probability of any randomly selected human being, throughout the whole history of the earth, dying by supernova, is roughly 1 in 10 million - assuming that humans have existed throughout the time in question, which is a few hundreds of million of years. But people have only been around for a million or two.

More meaningful is to say that any randomly selected "lifetime" (i.e. 70-year stretch), from the whole of Earth's history, has a roughly 1 in 10 million chance of being a time when a supernova happens. (Though from the Wikipedia figure it's 1 in 3.4 million - and that's not the probability that the "lifetime" will occur during a certainly fatal supernova, only that at some time during the "lifetime" there will be a supernova within 10 parsecs, which could have consequences for the biosphere). Don't think that would go so well in the song, though.

Near-Earth supernova - Wikipedia, the free encyclopedia

[FraserClarke]

But you can't then conclude that the probability of a presently living person dying by supernova is 1 in 10 million.

Yes I agree -- but that's only because we have prior knowledge of the distribution supernovae at present. If we didn't understand which stars caused supernovae, or didn't know how to measure their distances, then you couldn't make that assumption. Without that knowledge, we are -- as far as statistics is concerned -- all just randomly selected humans throughout the whole of Earth's history.

As you say, the 1 in 10 million is only an average probability over a timescale of billions of years -- it's clearly not constant through-out the Earth's lifetime. At the minute it's probably a lot lower, but it's certainly not zero -- we don't know that there isn't a massive star 1500pc away which is about to cook-off and fry us with a highly directed beam of gamma-rays (i.e. a GRB). We might be looking right down it's pole...

Not such a good song lyric indeed

[JulianO]

On the bright side at 1500pc, that gives us nearly 5000 years to get ready, except... if might already have happened!

[acey]

Yes I agree -- but that's only because we have prior knowledge of the distribution supernovae at present. If we didn't understand which stars caused supernovae, or didn't know how to measure their distances, then you couldn't make that assumption. Without that knowledge, we are -- as far as statistics is concerned -- all just randomly selected humans throughout the whole of Earth's history.

What the statistic highlights is that probabilities are not (generally speaking) facts of nature, but reflections of knowledge.

An analogy: a hotel has 100 rooms, and in one of those rooms there is a ghost. Person A is in the haunted room - he knows it because he's kept awake all night. Person B has been staying in another room for weeks and has come to conclude that his room can't be haunted. Person C shows up looking for a room to stay in, aware that one room is haunted but not knowing which one, or whether it's occupied.

What is the probability of encountering the ghost? For Person A it's 100%, Person B reckons it's close to zero, and for Person C it's 1/100.

With regard to supernovae, we are in the position of Person B. The "1 in 10 million" supernova statistic is from the point of view of Person C, and that is highly misleading.

If person C could arrive at an empty hotel 1000 times, being allocated a room each time, we would expect him to meet the ghost approximately ten times. If we could arrive at a random period in history in which to spend a lifetime, and could do the experiment ten million times, we would expect to see a dangerous supernova on roughly 3 occasions.

Are we exceptionally lucky to be free of supernova risk? No - but if we happened to find ourselves in an era with a dangerous supernova we would conclude ourselves to have been prodigiously unlucky.

[Relativista]

how could faraway supernova kill us? if it can,i think it was a very very luminous star before.since the flux of radiation is luminosity/4pi(distance)^2.then,it must be either near or very luminous enough to kill us. i think :-)

[Relativista]

I think the gammas are the worst - as they travel at the speed of light. So the first we'll no about it is when they hit us. They won't be deflected by magnetic fields either - so will come straight through and hit the atmosphere and us. Basically it will depend on the flux density we receive.

The charged particles will travel far more slowly, so we'll have lots more warning, and we'll be protected by the magnetic fields somewhat, but probably not enough!

No, high-energy particles and electromagnetic waves such as gamma will absorb by our atmosphere.

[acey]

Radiation from the supernova could damage the ozone layer, affecting the biosphere. it is reported that a Type II supernova closer than 8 parsecs (26 light years) could destroy half of Earth's ozone layer.

Near-Earth supernova - Wikipedia, the free encyclopedia

[ollypenrice]

At one time I believed that 30 people per year were killed by falling coconuts. It then went up to 150. Closer research reveals the disappointing truth. It seems unlikely that anyone is killed by falling coconuts, though being hit by them can cause injury and loss of consciousness. I only mean that it is disappointing from the Pythonesque humour point of view, of course, since those persons one might reasonably wish to see quit the earth are unlikely to be coconut pickers and more likely to be politicians, rap artists or reality TV show stars.

But can we be sure that no-one has been killed by supernovae? It is 1054. A Chinese sage is taking his habitual walk along the Great Wall of China when he looks up. Goodnesss me, what on earth is that, he says, stepping back to get a better view...

Olly

[Photosbykev]

I think the the odds of dying in a supernova explosion are an absolute certainty. It's the odds of a Supernova happening that are questionable

[HippyH]

Intresting article to read.

Gamma-Ray Burst Caused Mass Extinction?

[newman]

I saw a documentary that claimed that a supernova within 8,000 light years would have a potential to adversely affect the living conditions on Earth. This struck me as a bit odd but I'm assuming they ran the numbers.

Gamma ray bursts that emit only from the star's poles are called hypernovas and getting hit by that would be the worst case scenario, as all of the energy is focused into a (relatively speaking) small beam as opposed to an expanding sphere with normal supernovas. But for a hypernova to hit us we'd still need to be directly in the polar path of such a star. I wouldn't lose sleep over it

[obscura]

Don't think there is star mass of 15xsun+ that close to us to cause us a problem. The sun itself is the end of the world - but I don't think that's in my lifetime. Never considered coconuts! Better steer clear of them in sainsbury's.

[Neilius]

I guess most of us do the Lotto from time to time. Well the odds of winning the jackpot are 14M to 1.

So death by Supernova is more likely than winning the lottery.