Polar motion excitation – A broad-band perspective

Many recent studies of polar motion have demonstrated that climate-related variations (air and water motion and mass redistribution) provide the dominant source of polar motion excitation over virtually the entire band of frequencies that can be studied with modern space geodetic data. In spite of this general conclusion, when numerical model estimates of climate sources are actually compared with observed polar motion data, coherence is not necessarily high at all frequencies, and there are often differences in power spectral levels. In a qualitative sense this is understood to be due to imperfect climate model estimates, but it is generally difficult to judge relative or absolute quality of any particular climate series among the many available from a variety of numerical models. Here we examine the full frequency band accessible with daily data in order to understand causes of coherence and power spectral differences. A preliminary step is to introduce a least-squares broad-band data-driven method to form an optimum combined climate excitation time series, using as examples, series derived from the NCAR/NCEP reanalysis fields. This yields a coefficient for each series that is useful in judging its relative quality. The combined climate series is used to examine discrepancies evident in the spectra. At low frequencies the cause is likely to be inadequacies of individual climate component estimates, while at high frequencies alias contamination is probably the cause of diminished coherence and power spectrum differences.

► Development of a least squares data driven method to form an optimum linear combination of climate model estimates.
► Climate time series suffer from aliasing due to undersampling.
► Early polar motion data (1993–2000) suffer from aliasing, but later data seem to be properly sampled.
► True coherence appears to be very broad-band, but estimates are reduced at high frequencies by aliasing.
► The least squares combination method shows that terrestrial hydrology and ocean current estimates require the most improvement.