Fast Variance Component Estimation in GOCE Data Processing
- 2k Downloads
For the processing of GOCE (Gravity Field and steady-state Ocean Circulation Explorer) data the program system pcgma (Preconditioned Conjugate Gradient Multiple Adjustment) was designed as a tailored solution strategy for the determination of the Earth’s gravity field in terms of a spherical harmonic analysis. Within GOCE-HPF (High Level Processing Facility) the pcgma algorithm works with the purpose of a tuning machine in that it is used to optimize the filter design and to determine optimal variance components with respect to the combination of satellite-to-satellite tracking (sst) data, satellite gravity gradiometry (sgg) data and additional prior information about the smoothness of the gravity field (the latter especially with regard to the polar regions). pcgma is based on an extended version of the iterative conjugate gradient (CG) algorithm, which allows for data combination in terms of observation and normal equations. A basic prerequisite for handling the nesting of the two iterative methods (variance component estimation (VCE) and parameter estimation using CG) is an efficient and fast implementation, because the VCE requires a repeated solution of the system. In this paper we will show how the nesting can be organized in an optimal way. We will concentrate on the reduction of CG iteration steps.
KeywordsConjugate Gradient Observation Group Geoid Height Conjugate Gradient Algorithm Variance Component Estimation
Part of this work was financially supported by the BMBF Geotechnologien program GOCE–GRAND II (Grant 03F0421B) and the GOCE High-level Processing Facility. Parts of the simulations were performed using JUMP at Jülich Supercomputing Center. The computing time was granted by the John von Neumann Institute for Computing (project 1827).
- Alkhatib, H. (2007). On Monte Carlo methods with applications to the current satellite gravity missions. PhD thesis. Institute of Geodesy and Geoinformation. University of Bonn.Google Scholar
- Boxhammer, C. (2006). Effiziente numerische Verfahren zur sphärischen harmonischen Analyse von Satellitendaten. PhD thesis. Institute of Geodesy and Geoinformation. University of Bonn.Google Scholar
- Boxhammer, C. and W.-D. Schuh (2006). GOCE gravity field modeling: computational aspects – free kite numbering scheme. In: Flury, J., R. Rummel, C. Reigber, G. Boedecker M. Rothacher and U. Schreiber (eds), Observation of the earth system from space. Springer, New York, pp. 207–224.Google Scholar
- Brockmann, J.M. (2008). Effiziente Varianzkomponenentenschätzung ber iterative Techniken bei der GOCE–Daten Prozessierung. Diploma thesis. Institute of Geodesy and Geoinformation. University of Bonn.Google Scholar
- Koch, K.R. (2007). Introduction to bayesian statistics 2nd ed., Springer, Berlin, Heidelberg, New York.Google Scholar
- Koch, K.R. und J. Kusche (2001). Regularization of geopotential determination from satellite data by variance components. J. Geodesy, 76(5), 641–652.Google Scholar
- Schuh, W.-D. (1996). Tailored numerical solution strategies for the global determination of the earth’s gravity field. Number 81 in Mitteilungen der Universität Graz.Google Scholar
- Schuh, W.-D. (2003) The processing of band-limited measurements; filtering techniques in the least squares context and in the presence of data gaps. In: Beutler, G., M.R. Drinkwater, R. Rummel and R. v. Steiger (eds), Earth gravity field from space – from sensors to earth sciences, Band 108, Space Science Reviews, 67–78. ISSI Work-shop, Bern.Google Scholar
- Siemes, C. (2008). Digital Filtering Algorithms for Decorrelation within Large Least Square Problems. PhD thesis. Institute of Geodesy and Geoinformation. University of Bonn.Google Scholar