In my point of view there is not difference. When you develop a stochastic operational framework, you are considering that some parameters or variables have an stochastic nature. To consider the stochastic or uncertain nature of these variables/parameters you will need a probabilistic approach, i.e., you will have to be prepare to take a decision under an uncertain environment, for that, you will need an approach to help you take a good decision, knowing that there is not optimal (because you dont know the future, you can only estimated). There are plenty of these in literature: stochastic programming, robust optimization, scenario-based models, just to name a few.