Data Assimiliation in Science and Engineering

Catalog description

Description: Data assimilation is a well-established scientific discipline that combines computational models observations. It is geoscience terminology and refers to the estimation of the state of a physical system given a model and measurements. In other words, it is the process of fitting models to data. In engineering fields the terms filtering, estimation, and smoothing are often used. In the last decades data assimilation has gained popularity in many computational disciplines at both universities and research centers. In this course, starting from mathematical preliminaries (e.g., numerical linear algebra, model reduction, optimization techniques, etc), common methods of data assimilation (both sequential and variational methods) are introduced and derived in the context of both variational and estimation theory with emphasis on computational aspects.