Abstract: Taking on Wishart distribution function as likelihood function, this paper represents a maximum likelihood estimation of variance components of every coverage aimed at common points in the overlapped coverage. The method can estimate the unknown parameters and variance components independent of remnant errors and iterative computation.