Application of variance reduction techniques in marine radioactivity simulation

Abstract

Monte Carlo method plays an important role in simulating the passage of particles through materials. However, in the simulation of rare events such as the marine radioactivity, the method performs inefficiently, owing to its low convergence rate. To solve this problem, we employ three variance reduction techniques in our study of marine radioactivity simulations: importance sampling (IS) of particle position and momentum direction, biased Compton scattering process, and equivalent plane source. The three techniques are all essentially based on importance sampling. The results of biased simulation in comparison with analog simulation indicate that the first two techniques speed up the convergence rate under certain conditions, whereas the combined application of the three techniques achieved accurate results within a much shorter computation time. The effective application of variance reduction techniques makes it possible to simulate rare events. Meanwhile the study supplied useful reference data for the design of a marine radioactivity detector.