The science of "uncertainty quantification" tries to determine the likelihood of different scenarios in a given situation even if knowledge of the situation is incomplete. For instance, when analyzing an oil spill in an ocean, one wants to predict the extent of the contaminated region even when the initial distribution is not known precisely.
This sequence of images shows the evolution (top to bottom) of a theoretical oil spill. The black blob in the top figure represents the oil as it likely spread during the initial spill. The oil is then transported and mixed by two large opposite rotating structures called gyres. This 'stirring' leads to fine and intricate structures as the oil and water mix (bottom figure). In contrast to the oil (black) and water (orange), which swirl in complicated patterns, the blue and yellow "tracers" are transported randomly.