Data assimilation

Enhancing process models through ensemble-based data assimilation

Even the best process models may have uncertain parameters. The quality of such a model can be improved by data assimilation, that is, using measurements of some of the process variables to reduce uncertainty in the process-model parameters. Sometimes, this is the end product of data assimilation, like when seismic data are used to identify subsurface structures, while sometimes, the real aim is to improve the predictive power of the process model, like in weather forecasting.

Ensemble-based data assimilation is key to operational weather and climate forecasting and has the last decade become the dominating technique for improving the predictive power of reservoir flow models used for decision support in petroleum reservoir management.

The Data Assimilation and Optimization group within NORCE Energy consists of almost 20 researchers, all experts on ensemble-based methods. Personnel from the group invented ensemble-based data assimilation methodology, originally for use within oceanography, and initiated the research on this methodology for petroleum reservoir management. Members of the group have for many years participated actively in method development and testing within these and related areas.