English
 
Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT

Released

Conference Paper

Operational 14-day-ahead prediction of Earth's effective angular momentum functions with machine learning

Authors

Kiani Shahvandi,  Mostafa
IUGG 2023, General Assemblies, 1 General, International Union of Geodesy and Geophysics (IUGG), External Organizations;

Schartner,  Matthias
IUGG 2023, General Assemblies, 1 General, International Union of Geodesy and Geophysics (IUGG), External Organizations;

Gou,  Junyang
IUGG 2023, General Assemblies, 1 General, International Union of Geodesy and Geophysics (IUGG), External Organizations;

Soja,  Benedikt
IUGG 2023, General Assemblies, 1 General, International Union of Geodesy and Geophysics (IUGG), External Organizations;

External Resource
No external resources are shared
Fulltext (restricted access)
There are currently no full texts shared for your IP range.
Fulltext (public)
There are no public fulltexts stored in PuRe
Supplementary Material (public)
There is no public supplementary material available
Citation

Kiani Shahvandi, M., Schartner, M., Gou, J., Soja, B. (2023): Operational 14-day-ahead prediction of Earth's effective angular momentum functions with machine learning, XXVIII General Assembly of the International Union of Geodesy and Geophysics (IUGG) (Berlin 2023).
https://doi.org/10.57757/IUGG23-0346


Cite as: https://gfzpublic.gfz.de/pubman/item/item_5016114
Abstract
Effective Angular Momentum (EAM) functions of the Earth are connected to the various geophysical processes that perturb the Earth's tensor of inertia, thereby causing excitations that result in variations in Earth's rotation. Therefore, for the analysis of Earth Orientation Parameters (EOP), the determination and prediction of EAM functions are of high importance. The latter would benefit high-accuracy predictions of EOP, an essential task in geodesy for real time applications such as spacecraft navigation. For this reason, the Space Geodesy Group at ETH Zurich uniquely provides 14-day predictions of EAM functions on a daily basis via its Geodetic Prediction Center https://gpc.ethz.ch/EAM. The products cover the domains of the atmosphere, ocean, hydrology, and sea level, including both mass and motion terms. The general approach is based on a variation of Neural ODE, a powerful neural network tool for the prediction of time series of physical origin. A comparison to the results of other institutes reveals that our predictions during the first 6 days are more accurate by 50%. In addition, EAM predictions of days 7 to 14 show a good agreement with observations, thereby enabling us to better predict EOP. We show, for instance, that the 10-day prediction accuracy of length of day is improved on average by 17% by using the improved forecasts of the axial components of EAM.