Why the 18.6 year tide cannot explain the change of sign observed in j sub(2)

Recent studies show a change, starting in 1998, in the behavior of the variation of the dynamical flattening of the Earth (J2), supposed to be constant (secular), and mainly due to the post glacial rebound effect. In this paper, we study to what extent this behavior can be correlated or not with the 18.6 year tide: with more than twenty years of tracking data on LAGEOS-1, that is to say more than a period of 18.6 years, this effect can now be separated from the secular variation. We use our theory of mean orbital motion, dedicated to studies of the long period effects on the orbital motion. We build one single arc of LAGEOS-1 from 1980 to 2002, which provides a continuous description of the orbital parameters. This is the great originality of our approach. We focus our attention on the ascending node of LAGEOS-1, and we show that the change observed in j sub(2) cannot be attributed to a statistical error due to a correlation, in short arcs results, between the secular variation of J sub(2) and the 18.6 year tide. The proof is based on the adjustment of amplitudes and phases of the long period tides, and on the shape of the residuals.Key words. secular variation of J sub(2), 18.6 year tide, mean orbital motione