Ensemble-based data assimilation schemes for atmospheric chemistry models

The atmosphere is a complex system which includes physical, chemical and biological processes. Many of these processes affecting the atmosphere are subject to various interactions and can be highly nonlinear. This complexity makes it necessary to apply computer models in order to understand the natural behavior of the atmosphere. In addition to the chemical and physical processes it requires detailed information on the emissions, deposition and transport of trace constituents. Data assimilation represents a crucial tool for estimating and predicting the chemical processes in the atmosphere. It refers to several techniques that aim to combine the information from various sources to provide unified and consistent description of an atmospheric chemical system. The aim of this thesis was to investigate the use of several sequential data assimilation techniques in both ideal and real settings in the context of atmospheric chemistry. The increased complexity of models together with the different types of information about pollutants have the potential of contributing to a better understanding of chemical processes. Therefore, an optimal selection of the data assimilation techniques adapted to the new challenges in the atmospheric chemistry is required.Applied mathematicsElectrical Engineering, Mathematics and Computer Scienc