Uncertainties in supernova input rates drive qualitative differences in simulations of galaxy evolution

Feedback from core collapse supernovae (SNe), the final stage of evolution of massive stars, is a key element in galaxy formation theory. The energy budget of SN feedback, as well as the duration over which SNe occur, are constrained by stellar lifetime models and the minimum mass star that ends its life as a SN. Simplifying approximations for this SN rate are ubiquitous in simulation studies. We show here how the choice of SN budget and timings ($t_0$ for the delay between star formation and the first SN, $\tau_{\rm SN}$ for the duration of SN injection, and the minimum SN progenitor mass) drive changes in the regulation of star formation and outflow launching. Extremely long delays for instantaneous injection of SN energy $(t_0 << 20\;\rm{Myr})$ reduces star formation and drive stronger outflows compared smaller delays. This effect is primarily driven by enhanced clustering of young stars. With continuous injection of energy, longer SN durations results in a larger fraction of SN energy deposited in low ambient gas densities, where cooling losses are lower. This is effect is particularly when driven by the choice of the minimum SN progenitor mass, which also sets the total SN energy budget. These underlying uncertainties mean that despite advances in the sub-grid modeling of SN feedback, serious difficulties in constraining the strength of SN feedback remain. We recommend future simulations use realistic SN injection durations, and bound their results using SN energy budgets and durations for minimum SN progenitors of $7M_\odot$ and $9M_\odot$.