Environmetric strategies to classify, interpret and model risk assessment and quality of environmental systems

The present paper deals with the application of some environmetric approaches such as cluster analysis, projection pursuit, Kohonen maps, neuron gas, principal components analysis, chemical mass balance modelling, multiple regression on principal components, and time-series analysis in environmental data mining. Several ecologically important objects including marine sediments, wet and dry atmospheric precipitation are treated in order to obtain relevant information about monitoring data set structure and relationships. Multivariate statistical models are offered, which could help in decision making and problem solution of the local environment. In this way the case studies are a contribution to the idea of the increasing role of the environmetrics in the concept of sustainable development.