The concept of goodness of geodetic observations restated in view of the requirements of studies of recent crustal movements

The estimation of recent crustal movements from repeated geodetic observations is somewhat hampered by an apprehension of the uncertainties involved in the estimation of the standard errors of the observations and by the lack of correlation between the actual errors in networks and the theoretical estimates of these errors. It is here shown that a more reliable evaluation of the quality of the observations may be reached through a factor analysis of several statistical parameters. The suggested procedure is tentatively applied to some published data and a scheme aiming at a generalized treatment is proposed.