# Intensity increase in Monte Carlo radiative transfer simulations (in case of Rayleigh scattering)?

I came across something weird (at least for me) while looking at the Müller matrix for Rayleigh-scattering. If I see it correctly, for some scattering angles, it can actually increase the intensity up to 1.5 . My question is, how do you treat it in a Monte Carlo simulation? You take the norm after every scattering event, so the first element of the Stokes vector stays <=1, or do you just leave it at that, even if you get a number bigger than 1? Or does it depend on what you are trying to simulate? I thought that I has a maximum of 1 so I'm trying to understand what can be done in this situation.