Optimal processing of satellite-derived magnetic anomaly data |
| |
| Authors: | Malcolm G McLeod |
| |
| Institution: | Institute of Geophysics and Planetary Physics, University of California, Los Angeles, CA 90024 U.S.A. |
| |
| Abstract: | The statistical concept of the power spectrum has proven to be of great value in the analysis of time series and linear systems for which the inputs and outputs are functions of time. This paper shows how the concept can be extended to two-dimensional spatial power spectra and illustrates, by example, how the concept can be applied to the determination of optimal data processing methods for satellite-derived magnetic anomaly data and to the planning of missions to obtain such data.The analysis techniques indicated are applied to a data set and data processing procedure described by Mayhew et al. (1980). These authors describe magnetic anomaly data for Australia and surrounding ocean obtained by the polar orbit POGO series satellites. This paper shows that the data processing method used by these authors is approximately equivalent to an invariant two-dimensional linear filter and that it is reasonably close to optimal with respect to accuracy, though some possible improvements are suggested. Nevertheless, as is usual when filtering data, some real “signal” is unavoidably removed along with the “noise”, resulting in errors that can be quite large.A method for reducing these errors by using additional data from a medium inclination orbit satellite (for example, 60° inclination) is suggested. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
