Polar motion measurements: subdecimeter accuracy verified by intercomparison.

An important bound on the accuracy of modern techniques for monitoring polar motion is established by intercomparison of measurement series from two different observing techniques, very long baseline interferometry and satellite laser ranging. The root-mean-square differences between the estimates of the pole position from both techniques are shown to be only 2 milliseconds of arc (about 6 centimeters at one Earth radius). In the absence of common systematic errors, these differences bound the total errors in both sets of estimates. An initial investigation did not reveal any clear signature in the pole position that seems to be associated with major earthquakes. Continued measurements at this level of accuracy hold promise for resolving long-standing arguments over such questions as the nature of the excitation mechanism required to maintain the motion of the pole.

Variations in the rotation of the Earth.

Variations in the earth's rotation (UT1) and length of day have been tracked at the submillisecond level by astronomical radio interferometry and laser ranging to the LAGEOS satellite. Three years of regular measurements reveal complex patterns of variations including UT1 fluctuations as large as 5 milliseconds in a few weeks. Comparison of the observed changes in length of day with variations in the global atmospheric angular momentum indicates that the dominant cause of changes in the earth's spin rate, on time scales from a week to several years, is the exchange of angular momentum between the atmosphere and the mantle. The unusually intense El Niño of 1982-1983 was marked by a strong peak in the length of day.