Journal of Hydrology, vol.46, no.3-4, pp.239-249, 1980 (Scopus)
General formulation of adaptive parameter and state estimations has been fully developed on the basis of Kalman filters, given a periodic-stochastic model coupled with an observation model. It has been successfully applied to decompose a given monthly flow sequence into periodic and stochastic components. The results show a definite advantage of using adaptive Fourier analysis over conventional analysis. © 1980.