This paper deals with spectral techniques applied to geodetic problems. The solutions of the Inverse Stokes problem and of the Overdetermined Boundary Value Problem have been obtained applying The Wiener principle directly in the spectral domain. The resulting estimator for the Inverse Stokes problem is a low pass filter which is tuned by the covariance structure of the data while the one solving the Overdetermined Boundary Value Problem is nearly a weighted mean of the two spectra of the boundary data. Numerical examples on simulated data have been carried out to test the derived estimators.