The model of production and expansion of solid, liquid and gaseous pollutants. The optimal parallel solution

The three-phase model of production and expansion of solid, liquid and gaseous pollutants is described. Some numerical investigations of smog appearance and atmosphere cleaning are performed. The scheme of optimal parallelizing rules selection is proposed. The strategies of exchanges reducing and optimization are proposed for geometrical parallelizing. Some part of data exchanges is excluded using periodical extrapolation: rigid regression with coefficients adapting is applied. The special problem of nonlinear integer programming is solved to optimize the rest of data exchanges.