The day in conversation for 2026-10-23

discussion

Transcript

The discussion: Mira: That Bayesian reanalysis of 19C breakup seems promising because yielding a new one-neutron separation energy directly challenges some older phenomenological models we used last week. Kai: I agree it's a strong result, but I’m curious about how they handled the systematic input uncertainties; did they treat the uncertainty in the experimental cross-section measurement or is that purely absorbed into the posterior distribution? Lev: From a modeling perspective, if those uncertainties aren't rigorously propagated through the likelihood function correctly, even a new separation energy can be misleading when comparing it to theoretical predictions derived from Hartree-Fock calculations. Mira: Exactly; if the input uncertainty margin is too large, we lose the sensitivity to subtle structural changes that might indicate collectivity versus simple single-particle configurations. Do you think their error propagation accounted for correlations between those input parameters? Kai: They seem to have treated them somewhat separately initially, but I recall from our work on Ne-2

More episodes

← Home