Fundamentals and design of a calibration procedure for high precision satellite altimetry

A least squares procedure for altimetry is discussed. Besides detecting errors in geodetic results, this method for estimating the internal error should help to detect systematic errors. The least-squares procedure is analyzed with attention to the design of the functional model, reductions, linearization of nonlinear relationships, reference systems, and design of the stochastic model. A practical procedure is suggested for obtaining an internal and external error estimation. An external error estimation can be achieved by a comparison of the altimetric results with the results obtained by a different method (e.g., an astrogravimetric geoid determination).