Seasonal excitation of polar motion

We estimate geophysical excitations (χ1 and χ2) of polar motion using multiple sources of data, including recent atmospheric, oceanic, and hydrological models, satellite gravity measurements from the Gravity Recovery and Climate Experiment (GRACE), and compare geophysical excitations with observed polar motion excitations from space geodetic techniques. At seasonal time scales, both model-estimated excitations from the geophysical fluids envelope (i.e., atmosphere, ocean, and hydrosphere) and GRACE-observed excitations agree remarkably well with polar motion observations in the χ2 component, and in the χ1 component, model estimates and observed geodetic excitations show significant discrepancies. However, mass excitations estimated from GRACE show significantly better agreement with observed excitations than those from models, especially in χ1, due to better quantification of terrestrial water storage and oceanic mass changes using GRACE data. Furthermore, GRACE satellite gravity measurements offer a unique means for quantifying contributions from cryospheric angular momentum (CAM) change, a component mostly neglected in previous studies due to the lack of adequate observations or reliable ice sheets models. Based on GRACE estimates, CAM excitations appear a minor, but not negligible contributor to seasonal excitations of polar motion. The significantly better agreement in χ2 (than that in χ1) between observations and model excitations is related to the higher sensitivity of χ2 excitations to atmospheric pressure and terrestrial water changes over the Eurasia and North American continents, because of the special relationship between the ΔS21 spherical harmonic coefficient (proportional to χ2) mass model and the locations of the two continents.

► Observed and AOW model excitations of polar motion agree very well in χ2.
► GRACE excitations agree remarkably well with observations in both χ1 and χ2.
► GRACE provides a unique means for estimating cryospheric excitations.
► The significantly better agreement in χ2 (than that in χ1) has been explained.
► The significantly larger variability in χ2 (than that in χ1) has been explained.