Work performed over the past 4 years at the Center for Integrating Statistics and Environmental Sciences (CISES) at the University of Chicago and Argonne National Laboratory to deal with various aspects of model uncertainty in numerical models of atmospheric chemistry and transport will be presented. Development of a new method for inverting
a single tracer version of a regional scale chemical transport model to account for uncertainties in emission using measured concentration of the tracer in the atmosphere as a constraint will be discussed. Application of this methodology to Ammonia, a key precursor to atmospheric aerosols will be discussed. Implementation of the adjusted
Kalman Filter approach that accounts for initial condition uncertainty and provides a method for sequential data assimilation of data with large uncertainties will be presented. An application of this methodology for Carbon Monoxide a primary pollutant in continental air will be presented. As a part of these model development work we have modified several aspects of pre-existing community based models of dynamics and chemistry on regional scales. This is expected to lead to the development of an online-coupled model for regional scale aerosol and dynamics using the Argonne Modeling Coupling Toolkit. The current status of this work and expected outcomes will also be discussed.