Адаптивное моделирование планетарных процессов на основе спутниковых данных

Improvement of quality of modelling quasi-stationary natural systems slowly varying a kind of statistical distributions, is one of the most actual problems of a modern science. Such property the overwhelming majority of complex natural systems has and most brightly it is shown in atmospheric, and is especial in biospheric processes. Any progress in improvement of display of property of dynamic stability of atmosphere and biosphere promises not only substantial increase of accuracy climatic, meteorological and ecological forecasts, but also improvement of understanding of laws of functioning, self-organizing and interaction complicated the natural complexes forming stable existence of life on our planet. Crucial importance in understanding and modelling of properties of biosphere and atmosphere is played with the information received from artificial satellites of the Earth as a result of long sounding{probing}. However huge volumes and complexity of acting{going} data complicate activity with it{her} conventional mathematical methods. Extremely flexible and effective tool here appear the dynamic neural nets developed within the framework of the Concept of adaptive self-organizing of complex systems (CAS). Activity is devoted to research of their capabilities on modelling quasi-stationary of dynamics a temperature field and net primary productivity of plants.