“Base Case”, “Multi-Realizations” and “Multi-Scenarios”, Current Status and Questioning

There is an apparent complementarty between these approaches: (1) the “Base Case” is the best estimation at a given moment, (2) “multi-Realizations” allows associating uncertainties, (3). “Multi-Scenarios” compensates the weaknesses of “Multi-Realizations” when the level of knowledge is insufficient.<br>This ideal vision is largely questionable. Firstly, the status of the “Base case” has changed. Initially, it was a deterministic model. Today, stochastic simulations are intensively used. The choice of stochastic realizations is a key point and still an open question. We want and “average” model but what does “average” mean? Secondly, multi-realizations are well adapted to evaluate uncertainties around one model. However, experience has shown that treating only one model can lead to some unexpected results (increase of uncertainty assessment with increase of knowledge). In addition, what is the validity of the realizations generated (no bijection between geological concepts and numerical model) and how many realizations? Finally, “multi-scenarios” seem to be an alternative. However, practice shows a clear lack of control, especially in the quality of the results. Where is the chosen scenario in the space of uncertainties? The theoretical bases are difficult to establish.<br>Based on case studies, these questions will be developed<br>