How close does a supernova has to be to be lethal?

[SubTachyon]

So one guy comes to me and says that he doesn't believe that a supernova 50ly away would be dangerous enough to pose a lethal threat to us. I contradicted him because based on my general reading supernovas (depending on their magnitude) between at least a 100 to 3000 light years away would expose us and our atmosphere to overwhelming levels of radiation. However then I thought of going further and actually doing some calculations myself to prove my point but I soon discovered the task is much more difficult than I would like it to be:

I start with the assumption that the supernova is 50ly away and radiates the energy worth of 1 Foe (10^44j). This gives me the value of 36Mj per metre squared at our distance. However here come the unknowns:
- Over what duration is this energy expelled? Are we looking at most of it shining through in 100 seconds or a month or the whole year?
- What % of that is short-wavelength radiation? Because unless we get kilowats worth of infrared radiation we are really mainly concerned with the harmful stuff right?
- How much can our atmosphere take? How many wats worth of gamma and x-rays do we have to be exposed to in order to deplete our ozone layer and kill us?

I am assuming there will be people here who dealt with a similar issue at some point and can perhaps at least talk from experience if not maths?

 

[Astronuc]

Distance would obviously depend on the energy/type of SN.

Take 2006gy - do the flux out to 100 - 1000 ly.
http://www.lanl.gov/orgs/adtsc/docs/4Heger.pdf

 

[Chronos]

A typical [type Ia or II] supernova would need to be less than 30 light years distant to pose any threat to life on earth. GRB's, on the other hand, could be a threat over a couple thousand light years. Another factor is what lies between Earth and a supernova. A significant quantity of interstellar hydrogen would cause compton scattering, further attenuating the deadlier wavelengths. For some recreational math, see here:
http://www.tass-survey.org/richmond/answers/snrisks.txt

 

[DrStupid]

Distance would obviously depend on the energy/type of SN.

And in case of an anisotropic supernova (e.g. due to bipolar jets) it will depend on its orientation.

 

[KrisOhn]

There are more things than just supernova that can kill us, what about starquakes?

http://en.wikipedia.org/wiki/SGR_1806-20

(Ok, I just posted the link because I think its awesome.)

 

[twofish-quant]

This was a topic of a lunch conversation and the consensus was that the lethal distance was in the tens of light years. It wouldn't be the direct radiation that would kill us but the gamma rays would destroy the ozone layer.

To do a rough calculations

1) it really doesn't matter since you are looking at total energy absorbed
2) you can put in order 10% to 50% in high UV
3) for this you calculate the number of ozone atoms and compare to the number of photons that get generated

If you are impatient, here is someone that has already worked it out... :-) :-) :-)

http://iopscience.iop.org/0004-637X/585/2/1169/56170.text.html