Astrophysics just seems soooo speculative. Even with margins of error in the millions of years and light years, it just seems that most of it is assumptions supported not by evidence (because we can't get direct evidence) but by previous assumtions which have over time fossilised into accepted theories.
Not saying this paper isn't good or anything, just that extrapolating a supernova from a few atoms of an isotope of a metal seems... a long fetch. Though probably better than nothing.
Well, they have to come from somewhere and supernovae are the only process that makes them. If you find them consistently in a layer of a certain age all over the Earth, that points to a single source. Two different isotopes too, but only created by the same process so far as we know. It’s pretty decent evidence.
As for direct evidence, I know what you mean. There so way to observe the supernova, but how often do the jury get to watch a murder happen on film? Yet we still convict murderers.
There are assumptions in play, but I doubt they are the ones you are thinking of.
The two main ones are that physics is the same everywhere, and physics is the same all the time. We have no way to say if these assumptions are correct, but we don't have evidence that they are wrong and we do have observations that are consistent with them being correct.
If we take those two assumptions as true, then astrophyiscs is squarely in the middle of the model and measure of hard sciences.
Modelling in particular is wide open for speculation, but there is no reason to think speculative models will be maintained if they are not consistent with measurement. Even the simplest speculative models necessarily have falsifiable and testable claims - if they don't then they don't say much about how the world is.
The nature of astrophysics means that we are able to test and measure in so many ways: near and far, in the present and in the past, atomic scale and galaxy scale. As long as we hold those two assumptions above, we have a lot of evidence and reason to believe our current astrophysical models of the universe.
Not saying this paper isn't good or anything, just that extrapolating a supernova from a few atoms of an isotope of a metal seems... a long fetch. Though probably better than nothing.