Have Inertial Waves Been Identified from the Earth'S Core?
- D. E. Smylie and
- Raymond Hide
Published Online: 29 MAR 2013
Copyright 1988 by the American Geophysical Union
Structure and Dynamics of Earth's Deep Interior
How to Cite
Melchior, P. J., Crossley, D. J., Dehant, V. P. and Ducarme, B. (1988) Have Inertial Waves Been Identified from the Earth'S Core?, in Structure and Dynamics of Earth's Deep Interior (eds D. E. Smylie and R. Hide), American Geophysical Union, Washington, D. C.. doi: 10.1029/GM046p0001
- Published Online: 29 MAR 2013
- Published Print: 1 JAN 1988
Print ISBN: 9780875904504
Online ISBN: 9781118666562
1986 saw two significant developments in long-period gravimetry. The first was the publication by Melchior and Ducarme [1986; MD] of a spectrum calculated from vertical gravity measurements from the super-conducting gravimeter in Brussels. The record was analysed following two large, deep earthquakes and showed several peaks in the period range 13 to 16 hr. [MD] speculated that the largest peak at 13.9 hr. could be due to either an internal gravity wave in the outer core, or translational motion of the inner core, excited by earthquakes. The second development came when Aldridge and Lumb [1987; AL] claimed that the [MD] spectrum in fact contained 8 peaks which were close to the known period of inertial waves in a rotating sphere filled with homogeneous inviscid fluid.
We first review some useful results from the superconducting gravimeter as related to the Earth's core and discuss observations of the free core nutation. We also present the data reduction methods used to produce a new spectrum of the complete 4-year superconducting gravimeter data set (from May 1982 to December 1986) which includes the previous[ MD] data as a subset. We identify 10 peaks in the record which seem to be above the spectral noise level of 4 ngalsf rom frequencies 0 .058 to 0.077 per hr (periods 17.24 to 13.00 hr), and select 7 earthquakes in that time, with seismic moments above 10T MN .m. We find enhanced spectra following the two large Fijii earthquakes (M ay, 1985), very similar in style to the previous results[ MD], whereas following the large Mexicane arthquake ( September 1, 985) the spectrum is much more disturbed with large peaks which show evidence of drift with time. The spectral peaks are not at the same frequencies from one earthquake to the next.
We attempt to interpret the new spectrum in terms of the inertial wave model. Evidence to date suggests that the largest computational error in the theoretical inertial-wavee igenperiodus sed by [AL] arises from neglect of the inner core. Laboratory observationsa nd theoretical eigenperiods published by Aidridge  for a sphericals hell show consider able uncertainty about the sense and magnitude of the difference from the (exact) calculations for a full sphere and have to be treated as unreliable. More recent computations of eigenperiod so internal gravity waves in a strongly- stable core (where convergenceis considerably more certain than the weak or neutral-stratificatiocna se) can be used to infer the effect of buoyancy and an inner core on the eigenperiods obtained from Poincar's theory.
Comparison of the 10 peaks in the 4-years pectrum with the calculated periods of inertial waves with simple spatial structure in a full sphere do not strike us a showing clear evidence of correlation. We discuss the extent to which the agreement claimed by [AL] may still exist (e.g. the gravimeter time series is highly non-stationary which may negate the spectral analysis). We also present the case that the gravimeterr esultsc an be otherwise explained by ocean tidal resonances in the North sea or, alternatively, as the long-sought-for translational motion of the Earth's inner core.
Resolution of the mechanism causing the spectral peaks we observe clearly requires a global network of high quality measurements using superconducting gravimeters