| Summary: | The paper is devoted to the development of a software complex for predictive modeling of biogeochemical cycles in the Azov Sea on high-performance computer systems in a limited time. The complex development includes the creation of software modules of the computational structure to calculate the concentrations of phosphorus, nitrogen, silicon, sulfur, dissolved oxygen forms and describe in detail phyto- and zooplankton dynamics in the Azov Sea; the integration of various environmental databases, satellite monitoring. The complex is adapted to solve a wide class of predictive problems of water ecology and water resources management. It includes the implementation of developed 3D mathematical model of hydrophysics and biological kinetics on computational grids, consistent with the complex shape of the water coastline. Parallel algorithms have been developed for numerical implementation of water ecology problems, oriented to high-performance computer systems. To increase the calculation efficiency of computational grid fragments assigned to graphics accelerators, an algorithm and its software implementation were developed in the CUDA C language. Based on the developed specialized software tools focused on supercomputers, scenarios for the development of the environmental situation and sustainable development management at biological rehabilitation of the Azov Sea were developed, including forecasts of changes in harmful algae concentration; dynamics of spatial interaction processes between phyto- and zooplankton populations; evolution of biological kinetics processes on the example of plankton interaction.
|
|---|