ANALYSIS OF SOME ENVIRONMENTAL DATA OF THE BAIKAL REGION

Abstract

Examples of processing and verifying the reliability of some statistical environmental data of the Baikal region are presented in this work. Specific algorithm is proposed for small volume statistical sampling for determining reliability using a linear probabilistic model, which is converted into a logit model to obtain a result with a given level of significance. Some examples of large volume statistics are given. These statistics are converted by regression analysis and algorithms of interpolation formulas. The main task is formed of main dependence between some main parameters of wildfires in the selected region with a given reliability. In this article the field covered will be recommended to one such class of statistical procedures, namely that found in covariance analysis. Considerable progress can be made in modelling natural anomalies and anthropogenic-technogenic factors if a degree of constraint is introduced the covariance analysis. The presented statistical procedures together with other methods of forecasting natural anomalies, for example, methods of dendrochronology, correlation and spectral analysis, etc., allow predicting the main parameters of fires in the studied region with a given reliability. Such these methods can be applied in range of environmental real programs containing the necessary volumes and concentration areas fire protection in the logistics centers of the Ministry of Emergency Situations of Russia. The current realization of presented statistical procedures relies on the computer methods of Mathcad.