Stochastic approaches are increasingly important in atmosphere, ocean, and climate science.  Forecasting the state of the atmosphere-ocean-ice-land system in the presence of uncertainty is an important aspect of geoscience. One challenge in Earth system forecasting is providing accurate probabilistic predictions. Stochasticity can be applied across all levels of model complexity, from simple models to general circulation models and Earth-system models. In these approaches, critical but unresolved or poorly understood processes are modeled stochastically. Progress has been made in stochastic modeling of precipitation, clouds, gravity waves, ocean eddies, land-ice and sea-ice processes, as well as in hydromechanics and sub-surface hydrology. However, applications of increasing complexity require new and sophisticated computational tools and techniques. This session aims at bringing together the mathematical, statistical, and geophysical communities. It welcomes submissions related to stochastic processes in meteorology, oceanography, and climate science, including discussion of new modeling techniques, stochastic parameterization schemes, and data-driven stochastic modeling.