If predictions from process models are combined with remotely sensed data, the process model can be continuously updated with new observations. This is called data assimilation, a technique that is intensively applied in operational weather forecasting.