CONSTRAINTS ON DEEP-EARTH PROPERTIES FROM SPACE-GEODETIC DATA

The amplitude of the Earth's nutation driven by a given component of t he tidal potential is governed primarily by three parameters pi which are composites of a larger number of 'basic' Earth parameters (ellipti cities, compliances, moments of inertia, etc., of the Earth and its co re regions). We obtain estimates of the pi by least-squares fitting of nutation amplitudes estimated from very long baseline interferometry (VLBI) data to theoretical expressions based on an analytical formulat ion of nutation theory which incorporates the role of the solid inner core. We show how the estimates obtained, as well as the overall fit, vary with the ellipticity assumed for the inner core, and examine how the results are affected when otherwise unmodelled effects of ocean ti des and mantle anelasticity are taken into account. Considering two an elasticity models, we find that the fit obtained with the use of one o f them is noticeably worse than if the other is used or if no anelasti city correction is made. Independent of the corrections applied, the X (2) Of the fit is found to be smallest if the ellipticity e(s) of the solid inner core is taken as about half that of the Preliminary Refere nce Earth Model (PREM). Independent estimates of the ellipticity ef of the fluid core and other basic parameters on which the pi depend cann ot be obtained from the estimates of the pi alone. Nevertheless, with certain assumptions that are less restrictive than those hitherto empl oyed in the literature, we find ef to be 5.0% higher than the PREM val ue if the best-fit value is assigned to e,, and 4.7% higher if e(s) = e(s(PREM)); these values for ep are in accord with the estimate of Gwi nn et al. (J. Geophys. Res., 91: 4755-4765, 1986), and correspond to a nonhydrostatic flattening of about 465 m and 435 m, respectively, of the core-mantle boundary. Our parameter estimates have implications fo r the value of the static part k(0) of the second-degree Love number k which seem to be hard to reconcile with information from other source s. Observational estimates of the amplitudes of the 18.6 year nutation s are also found to be not satisfactorily matched with theoretical exp ectations. A careful re-examination of data analysis and theory is nee ded to resolve these problems.