Numerical Simulation of Global Atmospheric Chemical Transport with Three-dimensional Adaptive Wavelet-based Mesh Refinement

Alterations in the chemical composition of the Earth's atmosphere, determined by the global chemical atmospheric transport, can significantly impact air quality, climate, human health and absorption of ultra-violet radiation. Efficient and accurate numerical simulation of global Chemical Transport Models (CTMs) is a difficult task due to several factors: multi-scale nature of the problem, large number of chemical species and complex chemical kinetics. These factors significantly slow down numerical calculations of the global CTMs and impose severe limitations on the spatial resolution for stationary numerical grids. It has been shown, that crude resolutions introduce large amount of numerical diffusion into the system, which in combination with strong flow stretching causes high numerical errors.

In order to resolve the described above difficulty we developed Wavelet-based Adaptive Mesh Refinement (WAMR) method for numerical simulation of atmospheric chemical transport. The WAMR algorithm is a three-dimensional adaptive grid technique implemented for parallel architectures. The method introduces dynamically a fine grid in the regions where small spatial structures occur and a cruder grid in the regions of smooth solution behavior. Therefore, the algorithm allows minimization of the number of degrees of freedom for a prescribed accuracy and results in much more accurate solutions than conventional numerical methods implemented that use stationary uniform or quasi-uniform grids.

The results of three-dimensional numerical simulations of global atmospheric chemical transport obtained with the WAMR algorithm have been compared against conventional CTM computations. It has been shown that the method allows two orders of magnitude or more finer resolution than conventional stationary grids for the same total number of grid points. Therefore, the method provides a realistic opportunity to resolve most challenging multi-scale problems in the global atmospheric chemical transport on existing computers by producing accurate results at a relatively low computational cost.

This work is supported by a National Science Foundation grant under Award No. HRD-1036563.