If you are going to describe the fluid motion as probabilities, are you going to assume those probabilities to be independent or not? An independence assumptio

If you are going to describe the fluid motion as probabilities, are you going to assume those probabilities to be independent or not? An independence assumption really helps with computation. However, considering the `eddies' in turbulence, I'd suppose there is large and complex coupling between the probabilities of flow even for positions that are separated. You might deal with the coupling, but at a first guess that feels like it requires tracking many branching paths, which would have exponential memory requirements.