Enhancement of Reflood Test Prediction by Integrating Machine Learning and Data Assimilation Technique

Figure 3

The IMD scheme. The prior and posterior parameters in DA are referred to as the multipliers of the physical models before and after the DA process. DA can provide high-quality datasets, including parameter distributions, to train the ML model to generate prospective samples for the following DA process.