On the decomposition in least squares (with examples of its application in satellite geodesy)

Methods of decomposition are compared and analyzed to display their basic principles and properties, and to give more or less general rules for application of the particular method to a given estimation problem. The methods pertain to three types of least squares estimation: adjustment, filtering, and collocation. Phase, group, and sequential adjustments are presented, as are Kalman filtering and sequential collocation. Examples of applications from satellite geodesy are used to illustrate the different methods.