ESTIMATION OF CONDITIONAL NON-GAUSSIAN TRANSLATION STOCHASTIC FIELDS

A theoretical formulation is presented to estimate conditional non-Gau ssian translation stochastic fields when observation is made at some d iscrete points. The formulation is based on the conditional probabilit y density function incorporated with the transformation of non-Gaussia n random variables into Gaussian variables. A class of translation sto chastic fields is considered to satisfy the requirement of nonnegative definite for the correlation matrix. A method of conditional simulati on of a sample field at an unobservation point is also proposed. Numer ical examples were carried out to illustrate the accuracy and efficien cy of the proposed method. It was found that: 1) the optimum estimator at an unobserved point based on the least-mean-square estimation is e qual to the conditional mean; 2) the estimated error variance is depen dent on the locations of sample observation, but independent of the va lues of observed data; and 3) the conditional variance does not coinci de with the estimated error variance. These findings, which have alrea dy been confirmed for a lognormal stochastic field by the Kriging tech nique are clearly different from the results of conditional Gaussian s tochastic fields.