The task of providing an optimal analysis of the state of the atmosphere requires the development of novel computational tools that facilitate an efficient integration of observational data into models. We discuss several new computational tools developed for the assimilation of chemical data into atmospheric chemical transport models. The distinguishing feature of these models is the presence of stiff chemical interactions.

The variational tools presented in this talk include automatic code generation of chemical adjoints, properties of adjoints for advection numerical schemes, calculation of energy singular vectors and their use in placing adaptive observations. Data assimilation results using the 4D-Var method are shown for several real test problems to illustrate the power of the proposed methods.

We introduce a new method for modeling the background errors as autoregressive processes. The proposed approach is computationally inexpensive, captures the error correlations along the flow lines, and results in nonsingular background covariance matrices. We also discuss practical aspects of nonlinear ensemble Kalman data assimilation applied to atmospheric chemical transport models.