Wiener filters and collocation in satellite gradiometry
Many years ago the problem was addressed in satellite geodesy whether the collocation scheme based on covariances estimated on the rotation group, under the hypothesis of stochastic invariance of the underlying random field, could be directly applied to the time-wise data analysis, which should be typically based on a prior hypothesis of time stationarity.
The answer was indeed in the negative, for the very simple reason that couples of points separated by constant time differences do not have equal spherical distances, not even in the case of a perfectly circular motion, due to earth rotation.
Nevertheless a direct computation based on simple examples shows that the variations of such distances are at most of a few percent up to half a cycle and that the covariances based on spherical models do agree quite well with those computed along the orbit if the estimation time is long enough, e.g. 1 month.
This has been verified in particular for functionals of the anomalous potential that are of direct interest for a gradiometric mission like GOCE; for instance this is true for T itself, for T r and T rr .
KeywordsCollocation covariance function Wiener filter.
Unable to display preview. Download preview PDF.
- Betti B. and Sansò F. (1989). “The integrated approach to satellite geodesy”, in: Lecture notes in earth sciences, vol. 25 “Theory of satellite geodesy and gravity field determination”, F. Sansò and R. Rummel eds., Springer-Verlag, Berlin.Google Scholar
- ESA (1999). “Gravity Field and Steady-State Ocean Circulation Mission”, ESA SP-1233 (1).Google Scholar
- Moritz H. (1989). “Advanced physical geodesy”, Wichmann, Karlsruhe.Google Scholar
- Papoulis A. (1984). “Signal analysis”, McGraw-Hill, New York.Google Scholar
- Reguzzoni M. (2002). “From the time-wise to space-wise GOCE observables”, EGS General Assembly, Nice, 21–26 April 2002.Google Scholar
- Sansò F., Venuti G. and Tscherning C.C. (1999). “A theorem of insensitivity of the collocation solution to variations of the metric of the interpolation space”, in: International Association of Geodesy Symposia, vol. 121 “Geodesy beyond 2000. The challenges of the first decade”, K.P. Schwarz ed., Springer-Verlag, Berlin, pag. 233–240.Google Scholar