Bayesian prediction of uncertainties of Monte Carlo radiative-transfer simulations [IMA]

http://arxiv.org/abs/1705.10340


One of the big challenges in astrophysics is the comparison of complex simulations to observations. As many codes do not directly generate observables (e.g. hydrodynamic simulations), the last step in the modelling process is often a radiative-transfer treatment. For this step, the community relies increasingly on Monte Carlo radiative transfer due to the ease of implementation and scalability with computing power. We show how to estimate the statistical uncertainty for radiative-transfer calculations in which both the number of photon packets and the packet luminosity vary. Our example application is the TARDIS radiative-transfer supernova code. We develop various approximations to the exact expression that are computationally more expedient while nevertheless providing numerically accurate answers for large numbers of packets. Beyond the specific problem addressed here, our Bayesian method is applicable to a wide spectrum of Monte Carlo simulations including particle physics. It is particularly powerful in extracting information when the available data are sparse or quantities are small.

Read this paper on arXiv…

F. Beaujean, H. Eggers and W. Kerzendorf
Wed, 31 May 17
-206/48

Comments: 9 pages, 4 figures