Conditional simulation of spatiotemporal random fields of environmental contamination

The paper considers a method of conditional simulation of spatiotemporal scalar random fields of certain environmental phenomena. The method can be used to predict field values at given space points at specified time, on the basis of field values at other locations and data on second order moment functions in the domain. This approach has been applied to a space-time prognosis of soil contamination fields. The assessment of the spatiotemporal variability of heavy metals' concentrations provides the knowledge needed to monitor and control soil contamination. Empirical data of heavy metal (viz. chromium) concentration in the soil of northern Poland have been used in the study. The acceptance-rejection method has been applied to generate covariance matrices and vectors of discrete field values, taking into account conditional probability distributions. The results of the study show that the considered method can be successfully used to model conditional, spatiotemporal random fields of contamination with relatively small simulation errors.