Atmospheric data assimilation for scientific and operational purposes.

Data assimilation can be thought of as a mathematical technique for updating the fields of a forecast model in a controlled manner with information obtained directly by observations. The technique was originally developed for numerical weather prediction purposes, and much of the progress obtained in this field over the past few decades has been made possible by advances in data assimilation algorithms. These algorithms are now being applied also in other areas of the atmospheric sciences (e.g. atmospheric chemistry and climate research), as well as in other geo-sciences (most prominently in oceanography). In this presentation, we will give an introduction to the conceptual framework upon which data assimilation systems are built. We will provide a brief overview of some of the main characteristics of the current generation of assimilation systems, and we will attempt to highlight some of the most important applications in the atmospheric sciences.