Prediction of the Earth Orientation Parameters

Methods overview

The Space Geodesy group at ETH Zurich mainly focuses on applying advanced machine learning and deep learning methods in geodesy. More specifically, we apply recursive and recurrent machine learning methods for the ultra-short-term and short-term prediction of EOP time series. For more information, please check methods.

Products

The predictions are avaliable here

Quality Control

The daily evaluation report is avaliable at EOP prediction accuracy

Reference

Publications

[1] Kiani Shahvandi, M., Dill, R., Dobslaw, H., Kehm, A., Bloßfeld, M., Schartner, M., Mishra, S. & Soja, B. (2023). Geophysically Informed Machine Learning for Improving Rapid Estimation and Short-Term Prediction of Earth Orientation Parameters. Journal of Geophysical Research: Solid Earth, 128, e2023JB026720. https://doi.org/10.1029/2023JB026720

[2] Gou, J., Kiani Shahvandi, M., Hohensinn, R. & Soja, B. (2023). Ultra-short-term prediction of LOD using LSTM neural networks. Journal of Geodesy, 97(52). https://doi.org/10.1007/s00190-023-01745-x

[3] Kiani Shahvandi, M., Schartner, M. & Soja, B. (2022). Neural ODE Differential Learning and Its Application in Polar Motion Prediction. Journal of Geophysical Research: Solid Earth, 127(11). https://doi.org/10.1029/2022JB024775

[4] Kiani Shahvandi, M. & Soja, B. (2021, October). Small Geodetic Datasets and Deep Networks: Attention-Based Residual LSTM Autoencoder Stacking for Geodetic Time Series. In International Conference on Machine Learning, Optimization, and Data Science (pp. 296-307). Springer, Cham. https://doi.org/10.1007/978-3-030-95467-3_22

Presentations

[1] Kiani Shahvandi, M., Schartner, M. & Soja, B. (2022). Differential Learning: A method for polar motion time series prediction (No. EGU22-1101). Copernicus Meetings. https://doi.org/10.5194/egusphere-egu22-1101

[2] Kiani Shahvandi, M., Gou, J. & Soja, B. (2021). Deep quantum learning with long short-term memory for geodetic time series prediction: Application to length of day prediction. In AGU Fall Meeting Abstracts (pp. EP12C-08). American Geophysical Union. https://doi.org/10.1002/essoar.10508301.1

[3] Gou, J., Kiani Shahvandi, M., Hohensinn, R. & Soja, B. (2021, March). Ultra-short-term prediction of LOD using LSTM neural networks. In EGU General Assembly Conference Abstracts (pp. EGU21-2308). https://doi.org/10.5194/egusphere-egu21-2308