Nuclear data uncertainties in the production of p nuclei in massive
stars have been quantified in a Monte Carlo procedure. Bespoke
temperature-dependent uncertainties were assigned to different types of
reactions involving nuclei from Fe to Bi. Their simultaneous impact was
studied in post-processing explosive trajectories for three different
stellar models. It was found that the grid of mass zones in the model of
a 25 M⊙ star, which is widely used for investigations of
p nucleosynthesis, is too crude to properly resolve the detailed
temperature changes required for describing the production of p nuclei.
Using models with finer grids for 15 and 25 M⊙ stars with
initial solar metallicity, it was found that most of the production
uncertainties introduced by nuclear reaction uncertainties are smaller
than a factor of 2. Since a large number of rates were varied at the
same time in the Monte Carlo procedure, possible cancellation effects of
several uncertainties could be taken into account. Key rates were
identified for each p nucleus, which provide the dominant contribution
to the production uncertainty. These key rates were found by examining
correlations between rate variations and resulting abundance changes.
This method is superior to studying flow patterns, especially when the
flows are complex, and to individual, sequential variation of a few
rates.
