# How often do supernovae occur?

Type Ia supernova are used as standard candles. But they also are transient events. This means that to determine the distance of a galaxy using supernovae, you have to wait for one to occur. How often do type Ia supernovae typically occur in a galaxy ?

• Some things really..... en.wikipedia.org/wiki/History_of_supernova_observation Jan 21 '20 at 18:49
• Don't think the Wikipedia article answers the OP's question, it seems to mostly discuss very local supernovae. I read it as asking for the supernova rate in galaxies generally (which is a function of redshift, metallicity and galaxy type) rather than in the Milky Way specifically Jan 21 '20 at 19:48

There is a lot of scope to provide a very detailed answer here. The rate depends on what sort of a galaxy you are considering and when, what its star formation rate is (or was) and what its total stellar mass is.

A good reference is the Annual Review of Astrophysics article by Maoz et al. (2019). This says that for a Sbc galaxy like the Milky Way, the specific rate ( the rate per unit of stellar mass in the galaxy) of type Ia supernovae is $$10^{-13}$$ per year per solar mass of stars. For a Milky Way stellar mass of $$\sim 6\times 10^{10} M_\odot$$, this means a rate of 0.006 per year, or one type Ia SNe every $$\sim 200$$ years.

In the local universe this rate scales as about $$M^{-0.5}$$, where $$M$$ is the galaxy mass (valid over about 5 orders of magnitude for $$M>10^7 M_\odot$$; Brown et al. 2018). So the specific rate of type Ia supernovae is larger in smaller galaxies.

The rate also decays after a galaxy is formed, or after a burst of star formation as something like $$t^{-1}$$.

• This answer hurt my brain with all the math and astronomy going on here, especially when contemplating that we have actually calculated this sort of thing and can use it to calculate subsequent events. I like it. Jan 22 '20 at 4:38
• Interestingly, nasa.gov says "Astronomers believe that about two or three supernovas occur each century in galaxies like our own Milky Way." Jan 22 '20 at 23:53
• @VictorStone the rate of type Ia supernovae must of course be smaller than the rate of (all) supernovae And is probably smaller than the rate of core collapse (type II) supernovae. Jan 23 '20 at 0:03

The rate is evolving with time (and consequently with redshift). According to a paper in 2005 by Evan Scannapieco & Lars Bildsten (https://arxiv.org/abs/astro-ph/0507456), the rate of SNe Ia is 1 to 3.5×104 per yr per Mpc3 at z=2, which is the peak rate.

You can very roughly consider it as 1 to 3.5 events in a galaxy in 10,000 years. Considering there are so many galaxies in our observable universe, you can still observe some in each year. With more powerful telescopes and strategies, there will more and more SNe Ia be observed. They are dimmer, but more. Probably, more than once per day.

I have a much different answer. I'm taking a course in Astrophysics and the instructors took the approach of using the amount of iron in our galaxy to calculate the number of SN since galaxy was formed. They found how many SN it would take to produce that much iron. According to their calculations you need about 1 SN every 50 years from the Milky Way.

• I don't think that's "much different" from the other answers. Can you show the calculation?
– pela
Jan 23 '20 at 20:11
• This is indeed a valid and more direct cross-check of results for the Milky Way (since we have not been able to observe enough Type Ia supernovae in the MW to count directly). However, could you give a reference for the calculation and be clear whether this results applies to Type Ia supernovae (which is what the question is asking about) or all supernovae. Another point is that this presumably then is an average rate over the lifetime of the Galaxy? If so, the current rate is likely to be significantly lower. Jan 23 '20 at 21:06
• The course is a series of video lectures. they are located here: courses.edx.org/courses/course-v1:ANUx+ANU-ASTRO1x+2T2016/…. This links takes you to the video before the one of interest. So scroll to the top of page and select the next lecture. Jan 24 '20 at 18:46
• The calculation is for type II sn which they state accounts for 1/3 of all the iron, and yes they use the lifetime of the universe. Jan 24 '20 at 18:54
• So @jmh, it isn't an answer to this question. Jan 25 '20 at 17:57