Dynamic Parameters of the Geocenter Produced by GPS and DORIS Navigation Systems by Means of Adaptive Regression Methods

This work focuses on the creation of the digital model of geocenter (DMGC) as well as on the study and prediction of its dynamics. The observations from GPS and DORIS navigation systems and dynamic regression adaptive modeling (DRAM) were used for this purpose. The diagrams of DMGC’s change are presented as polynomial time series (PTS). The dynamic regression adaptive modeling was used to increase the accuracy of DMGC and its prediction. In doing so, regular effects of DRAM were taken into account for the reliable estimation of the desired parameters. Three software complexes were developed: “Automated System of Dynamic Regression Adaptive Modeling” (ASDRAM) and its modifications – ASDRAM-G (for remote sensing of the Earth measurements reduction) and ASDRAM-T (for technogenic processes analysis). ASDRAN is capable optimizing the structure of the regression model, predicting DMGC, determining polyharmonic series’ (PHS) stability, analyzing the presence of absence of significant harmonic components of PHS, and estimating the correlation between them (cross-spectral analysis). The results produced in this paper have shown their approximation and prediction accuracy for DMGC behavior compared to previous studies. For all DMGC built, the common structural elements are obtained.