The field of numerical weather prediction has played a major role in germinating the field of data assimilation. The complex numerical weather prediction models immediately demanded algorithms for determining their initial states from available observations. At the same time, and largely disconnected from numerical weather prediction, the field of oil reservoir simulations has led to the development of data assimilation methods with a stronger focus on combined estimation of model states and parameters.
In recent years, one could witness an increasing need for the extension of data assimilation methodologies to dynamical phenomena outside the range of these well-established areas. The projects within Research Area B respond to this need by focusing on application areas for which detailed physics-based models are either unavailable or computationally out of reach. These include applications from biophysics, cognitive neurosciences, seismology and space physics.