THE INTRINSIC RANDOM-FIELD MODEL IN THE STUDY OF SULFATE DEPOSITION PROCESSES

This work is concerned with the application of advanced stochastic analysis in atmospheric environment problems. In particular, the focus is the study of spatially non-homogeneous sulfate deposition processes in the conterminous U.S.A. by means of intrinsic random fields. The intrinsic model used here is significantly more general in theory and also more powerful in practice than many of the conventional methods. It provides spatial correlation tools which can efficiently represent the complex non-homogeneous features of the sulfate deposition data. The intrinsic approach, which has been developed in response to experimental findings, utilizes the spatial variability characteristics of the environmental process and the mathematical characteristics of the random field models to fit these models to the observed data. The results are represented in terms of maps which provide a quantitative description of the spatial distribution of both the degree of non-homogeneity and the correlation features of sulfate deposition over the conterminous U.S.A. An optimal spatial estimation method for non-homogeneous sulfate deposition processes is also discussed. The estimation inputs reflect important variability patterns of sulfate deposition and are operationally executable in the real world. The estimation outputs are detailed maps of sulfate deposition together with the associated maps of 95% confidence interval width. These maps offer valuable information regarding the non-homogeneous distribution of the sulfate deposition values in the various parts of the U.S.A. The practical implementation of the intrinsic model involves an automatic procedure which operates at a local scale providing a detailed description of the spatial structure of sulfate deposition.