Application of non-parametric modeling in solving problems of environmental monitoring

The analysis of the air quality in Russia showed the invalidity of existing methods of the air control and observation systems. The analysis of the main objectives of environmental monitoring can be solved by using satellite remote sensing data. The system of environmental monitoring in Krasnoyarsk city is developing. But nowadays it is important to apply modeling techniques in the evaluation of the environmental situation to integrate a number of factors (parameters of the emissions sources, current meteorological situation, the air dispersion conditions, the pollutant properties and others.). This paper shows the availability of H-models using for creating a model of ecological situation in Krasnoyarsk city. The using of appropriate non-parametric indicators is necessary to construct the parametric models with using “tubular” processes. This research analyses the issue of identification in the “narrow” and “broad” sense. Also, this paper discusses the standard way of modeling the environmental situation and describes the complexities in the modeling process with “tubular” structure by using of parametric methods. The model of the environmental situation forecast in Krasnoyarsk city was developed with using the methods of non-parametric modeling by the help of nuclear grade. The results of modeling characterize the spatial distribution of carbon monoxide. The using of this approach is efficient for the environment management in real time. The necessity of development of regional monitoring networks demands to solve the problem of choosing observation points and optimization program of air quality monitoring. For this purpose it is necessary to consider all influencing factors in totality, and also to use new technologies (including the remote environmental monitoring methods). The proposed approach will optimize the decision-making procedure in the field of environmental management. The improvement of the model can be connected with calculating of the time and space distribution of pollutants, and increasing of the accuracy of environmental quality forecasts.