Some Proposals About The Vector And Parallel Implementations Of STEM-II

STEM-II is an Eulerian air quality model which includes different pollutant emissions, transport by advection, convection and turbulence, with dynamic meteorological conditions; chemical transformation by gas-phase and aqueousphase, and pollutants removal by dry and wet deposition, complete the main air pollution processes. This model is being applied to simulate local and regional acid deposition (episodic and annual basis) from As Pontes Power Plant, considering both the local power plant environment and the EMEP 17,6 cell (150x150 km). In this work, a study of the large computational requirements of STEM-II, both in terms of storage and execution times, is done. First, the possibilities of vectorization of the code (with minor changes) is discussed and tested; secondly, different issues about the parallelization of the STEM-II code over distributed memory machines are considered.