Application of the Moving Average Method of Simplifying Simulation of Random Fields

A method of simulating random fields using a moving average is proposed in this paper. Random fields were simulated for different sizes of sub-field and different numbers of cycles of calculations of the moving average. For the fields obtained the covariance function was analyzed. In order to estimate the efficiency of the proposed simulation method of random field based on the method of diagonal covariance matrix was performed. It is shown that two of the simulation methods presented are able to generate a multidimensional random variable with required correlation function. However, the method of diagonal covariance matrix has some limitations caused by the size of the simulated random field, which result from the necessity of converting a relatively large matrix. Using the proposed simulation method it is possible to simulate, in a comparatively quick and simple manner, a random field with a large number of nodes on PC-s. The presented method can be useful in the stochastic analysis of transport phenomena in soil.