ADAPTIVE ESTIMATION OF THE LOG FLUCTUATING CONDUCTIVITY FROM TRACER DATA AT THE CAPE-COD SITE

An adaptive estimation scheme is used to obtain the integral scale and variance of the log-fluctuating conductivity at the Cape Cod site bas ed on the fast Fourier transform/stochastic model of Deng et al. (1993 ) and a Kalmanlike filter. The Iter incorporates prior estimates of th e unkown parameters with tracer moment data to adaptively obtain impro ved estimates as the tracer evolves. The results show that significant improvement in the prior estimates of the conductivity can lead to su bstantial improvement in the ability to predict plume movement. The st ructure of the covariance function of the log-fluctuating conductivity can be identified from the robustness of the estimation. Both the lon gitudinal and transverse spatial moment data are important to the esti mation.