APPROXIMATE SOLUTION FOR STOCHASTIC INTERPOLATION OF CONDITIONAL NON-GAUSSIAN FIELDS

The approximate algorithm for stochastic interpolation and extrapolation of conditional non-Gaussian fields proposed in this paper produces a nonlinear unbiased estimator with the characteristic minimum variance of errors. The extended Kalman filtering procedure is then used for solving the conditional estimation problem transformed into the Gaussian stochastic field. A lognormal stochastic field is taken up as an example. The accuracy and efficiency of the proposed method is discussed compared with the theoretical solutions of Simple Kriging. It is found that the proposed method does not suffer from the difficulties associated with computing the optimum estimator and estimated error variance.