Automated Construction of the Basic Dynamic Models of the Atmospheric Photochemical Systems Using the RADM2 Chemical Mechanism as an Example

The method of automated construction of the basic dynamic models, i.e., mathematically correctly simplified models using a minimum possible set of the dynamic variables, which is necessary for correct description of the spatiotemporal behavior of the atmospheric photochemical systems, is developed. The proposed method is based on numerical search for all the chemical components rapidly evolving near their equilibrium states and sorting out slow families among them, whose characteristic lifetimes turn out to be much longer compared with those of their components. The method is applied to the RADM2 chemical mechanism which is used for simulating air contamination in urbanized regions. This allowed us to find almost thirty fast variables of the above-mentioned system, single out three intersecting slow families and reduce the dimension of the corresponding basic dynamic model more than twice. It is shown that the constructed model well reproduces (both qualitatively and quantitatively) the calculated results, which are obtained using the reference full-scale model, and can be used for both direct simulation of the air quality in the atmospheric boundary layer with allowance for the actual transfer processes and retrieval of information on the tropospheric chemical substances from the available observation data bases.