Inter-​Commission Committee on Theory (ICCT)

Due to the exponential increase in computing power over the last decades, machine learning has grown in importance for several applications. In particular, deep learning, i.e., machine learning based on deep neural networks, typically performed on extensive data sets (“big data”), has become very successful in tackling various challenges, for example, image interpretation, language recognition, autonomous decision making or stock market predictions. Several scientific disciplines have embraced the capability of modern machine learning algorithms, including astronomy and many fields of geosciences.

The field of geodesy has seen a significant increase in observational data in recent years, in particular from Global Navigation Satellite Systems (GNSS) with tens of thousands of high-quality permanent stations, multiple constellations, and increasing data rates. With the upcoming NISAR mission, the InSAR community needs to prepare for handling daily products exceeding 50 GB. In the future, the next-generation Very Long Baseline Interferometry (VLBI) Global Observing System (VGOS) will deliver unprecedented amounts of data compared to legacy VLBI operations. Traditional data processing and analysis techniques that rely largely on human input are not well suited to harvest such rich data sets to their full potential. Still, machine learning techniques are not yet widely adopted in geodesy.

Machine learning in geodesy has the potential to facilitate the automation of data processing, detection of anomalies in time series and image data, their classification into different categories and prediction of parameters into the future. Machine learning and, in recent years, deep learning methods can successfully model complex spatiotemporal data through the creation of powerful representations at hierarchical levels of abstraction. Furthermore, machine learning techniques provide promising results in addressing the challenges that arise when handling multi-resolution, multi-temporal, multi-sensor, multi-modal data. The information contained in GNSS station position time series is essential as it can help derive important conclusions related to hydrology, earthquakes, or volcanism using machine learning. Other important applications are tropospheric and ionospheric parameters derived from GNSS where automated detection and prediction could be beneficial for improved severe weather forecasting and space weather monitoring, respectively. InSAR data will benefit in particular from efficient image processing algorithms based on machine learning, facilitating the detection of regions of interest. In several of these cases, the development of scalable deep learning schemes can contribute to more effectively handling and processing of large-scale spatiotemporal data.

Traditional machine learning techniques for geodetic tasks include convolutional neural networks for image data and recurrent neural networks for time series data. Typically, these networks are trained by supervised learning approaches, but certain applications related to autonomous processing will benefit from reinforcement learning.

The field of machine learning has expanded rapidly in recent years and algorithms are constantly evolving. It is the aim of this JSG to identify best practices, methods, and algorithms when applying machine learning to geodetic tasks. In particular, due to the “black box” nature of many machine learning techniques, it is very important to focus on appropriate ways to assess the accuracy and precision of the results, as well as to correctly interpret them.

• Identify geodetic applications that could benefit from machine learning techniques, both in terms of which data sets to use and which issues to investigate.
• Create an inventory of suitable machine learning algorithms to address these problems, highlighting their strengths and weaknesses.
  Perform comparisons between machine learning methods and traditional data analysis approaches, e.g., for time series analysis and prediction.
• Focus on error assessment of results produced by machine learning algorithms.
• Identify open problems that come with the automation of data processing and generation of geodetic products, including issues of reliability.
• Develop best practices when applying machine learning methods in geodesy and establishing standardized terminology.

Since the establishment of the JSG T.29 “Machine learning in geodesy” (from here on, simply “JSG”), several meetings have been organized to coordinate the activities of the JSG and promote cooperation and interactions of the group members. In the context of the JSG, several activities – as envisioned in the Terms of Reference – have been pursued as highlighted below.

Scientific session organization:

A new series of sessions at the European Geosciences Union (EGU) General Assemblies has been established by members of the JSG. The sessions focus on the very topic of the JSG, i.e., machine learning in geodesy. Since their first inception at the virtual EGU 2021 conference, the sessions have attracted a sufficient number of abstracts to fill oral slots in the EGU Geodesy Division program. Concretely, the following sessions were convened by JSG members:

• vEGU21: “G1.4 Data science and machine learning in geodesy”
• EGU22: “G1.3 Data science and machine learning in geodesy”
• EGU23: “G1.3 New developments in mathematical methods in geodesy, with a focus on machine learning”

The sessions attracted a wide variety of topics related to the application of machine learning in geodesy, featuring various types of data, methods, and applications. In terms of geodetic observation techniques, a majority of the presentations focused on GNSS, followed by InSAR and satellite gravimetry. Concerning the type of machine learning algorithm, most authors utilized some form of deep artificial neural networks, although tree-based ensemble algorithms were also popular choices. The sessions were very well attended and among the most popular sessions in the program of the EGU Geodesy Division.

Editorial activities:

A Special Issue in the Journal Remote Sensing was organized by members of the JSG. The title of the Special Issue was “Data Science and Machine Learning for Geodetic Earth Observation” and submissions were accepted from mid-2021 to February 2023. Eight papers were published as part of the Special Issue. The papers covered the application of machine learning to the prediction of Earth rotation, tropospheric and ionospheric parameters, among other topics.

Website:

A website with a description of the JSG and its activities was created and has since been maintained:
https://space.igp.ethz.ch/services/services-to-iag/icct-study-group.html

It includes general information on the JSG, such as the Terms of Reference and member list, as well as a description of the activities of the JSG.

Repository:

A major objective of the JSG was the establishment of a platform to share code examples concerning the application of machine learning in geodesy. This would allow interested geodesists with no or just little expertise in machine learning to find examples to get started. On the other hand, also experienced users can benefit from the code, for example concerning specific implementation details.

For this purpose, a public github repository has been created by the JSG: external page https://github.com/ICCT-ML-in-geodesy

So far, it features working machine learning examples for:

Earth orientation parameter prediction,
Ionospheric vertical total electron content prediction,
CyGNSS-based windspeed retrieval.
Each example is based on a Jupyter Notebook, which is useful for education purposes as it facilitates the interaction with the code and visualization of the results. In addition to the code, the required data is included in the repository. An additional example concerning InSAR-based pixel selection is in development.

Achievements and results

The members have been actively researching topics related to the JSG. Related publications and presentations are listed in the last section of this report. The publications can be grouped into:

Time series modeling and prediction. While mostly based on GNSS data, very different types of parameters are investigated, including station positions, tropospheric and ionospheric parameters as well as Earth orientation (e.g., Gou et al. 2023, Halbheer 2021, Kaselimi et al. 2020ff, Natras et al. 2020ff, Kiani et al. 2021ff, Ruttner et al. 2022).

InSAR-related investigations, mostly utilizing deep convolutional neural networks (Meganadh et al. 2021, Srivastava et al. 2022, Tiwari et al. 2020ff).

Gravity field and mass change modeling, primarily to increase the spatial or temporal resolution (Agarwal et al. 2023, Uz et al. 2022f)
Other topics include the use of artificial intelligence and machine learning for improved VLBI scheduling (Schartner et al. 2021a,b, Wicki 2021), Earthquake classification (Crocetti et al. 2021), seismology (Pan et al. 2020, Shujian 2021, Wu et al. 2022f), wind detection based on GNSS (Aichinger-Rosenberger et al. 2022), high-resolution refractivity field modeling (Shehaj et al. 2023), as well as remote-sensing-related topics (Marsocci et al. 2023a,b).

A variety of machine learning algorithms were utilized, from decision tree ensembles (random forest, boosting trees, ...) to artificial neural networks (convolutional, recurrent, graph, transformers, …). The investigation of uncertainties (e.g., Natras et al. 2022, Kiani and Soja 2022) is becoming increasingly important. This is also an important principle related to ethical use of artificial intelligence and machine learning as identified in Shelley et al. (2023), a comprehensive report that was co-authored by Ryan McGranaghan, member of the JSG.

Interactions with the IAG Commissions and GGOS

The JSG is affiliated with IAG Commissions 2, 3, 4, as well as GGOS. Additional engagements by the JSG with other entities are mentioned as well.

GGOS Focus Area on Geodetic Space Weather Research

It has been identified that several members of the JSG work on the topics related to the ionosphere and space weather (M. Kaselimi, R. McGranaghan, and R. Natras). Machine learning has become an important tool for the prediction of ionospheric parameters, typically utilizing not only geodetic measurements, but also solar data as features. This fits well to the scope of the GGOS Focus Area about the relationships between ionosphere/thermosphere and space weather. B. Soja, chair of the ICCT JSG, is vice-chair of the JWG “Improved understanding of space weather events and their monitoring by satellite missions”. Synergies between these groups are fostered.

IAG Working Group 4.3.2 Ionosphere Prediction

In a collaboration between members of the JSG and IAG Working Group 4.3.2. “Ionosphere Prediction”, predictions of global ionospheric maps (GIMs) have been investigated. The contributions from the JSG (concretely, ETH Zurich) were based on deep learning (convLSTM). In a hindcast experiment, the 1-hour and 1-day forecasts provided by the different institutions were compared with each other for both quiet days in terms of ionospheric activity as well as storm days. More details on the comparison are provided in the report of the IAG WG 4.3.2 as part of the IAG Travaux 2023.

IERS Second Earth Orientation Parameter Prediction Comparison Campaign (2nd EOP PCC)

The chair of the JSG was involved in the committee for the organization of the Second EOP PCC and represented the interests of the JSG in this context. The goal of the EOP PCC was to compare operational EOP prediction products provided by various institutions. The first EOP PCC finished more than ten years ago and was considered a success. A repetition was important due to the significant changes in data quality and availability as well as the new developments in prediction algorithms since then.

Several institutions participated in the 2nd EOP PCC, including ETH Zurich as a member of the JSG. Overall, there was an increase in popularity of machine learning/deep learning for EOP prediction. The 2nd EOP PCC was completed at the end of 2022 and the results are currently under investigation.

ITU-T Focus Group on Artificial Intelligence for Natural Disaster Management

The chair of the JSG has been involved in the ITU-T Focus Group on “Artificial Intelligence for Natural Disaster Management” (FG-AI4NDM), in particular in the Topic Group on “AI for Tsunami data monitoring” that heavily relies on GNSS data. Synergies in this context have been identified, as studies related to the detection of earthquakes and tsunamis with machine learning are in progress in different groups, including those of JSG members.

GGOS Focus Area on Artificial Intelligence for Geodesy (AI4G)

On May 12, 2023, the GGOS Coordinating Board accepted the proposal to establish a new GGOS Focus Area on Artificial Intelligence for Geodesy (AI4G). In general, it will utilize methods from the field of Artificial Intelligence (AI), including machine learning techniques, to improve geodetic observations and products. This new GGOS Focus Area will be chaired by Benedikt Soja together with his vice-chair Maria Kaselimi, both members of the JSG. The activities of the JSG have thus commenced in the successful establishment of this new GGOS component, which anchors the topic of machine learning in geodesy at a higher level within the structure of IAG. While the GGOS Focus Area can be seen as a follow-up activity to a certain degree, the focus of the GGOS Focus Area AI4G is on the actual improvement of geodetic data and products, whereas this JSG addresses the theoretical and methodological aspects. Within the GGOS Focus Area AI4G, there will be three Joint Study Groups with close relations to other IAG components:

• AI for GNSS Remote Sensing
• AI for Gravity Field and Mass Change
• AI for Earth Orientation Parameter Prediction
   

Agarwal, V., Akyilmaz, O., Shum, C.K., Feng, W., Yang, T.-Y., Forootan, E., Syed, T.H., Haritashya, U.K., Uz, M. (2023) Machine learning based downscaling of GRACE-estimated groundwater in Central Valley, California, Science of the Total Environment, 865: 161138, doi: external page https://doi.org/10.1016/j.scitotenv.2022.161138.

Aichinger-Rosenberger, M., E. Brockmann, L. Crocetti, B. Soja, G. Moeller (2022): "Machine learning-based prediction of Alpine foehn events using GNSS troposphere products: first results for Altdorf, Switzerland"; Atmospheric Measurement Techniques, 15(19):5821–5839, external page https://doi.org/10.5194/amt-15-5821-2022

Crocetti L, Schartner M, Soja B (2021) Detecting earthquakes in GNSS station coordinate time series using machine learning algorithms. EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1975, external page https://doi.org/10.5194/egusphere-egu21-1975.

Crocetti, L., M. Schartner, B. Soja (2021): "Discontinuity Detection in GNSS Station Coordinate Time Series using Machine Learning"; Remote Sensing, 13, 3906, external page https://doi.org/10.3390/rs13193906

Gou J, Kiani Shahvandi M, Hohensinn R, Soja B (2021) Ultra-short-term prediction of LOD using LSTM neural networks, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-2308, external page https://doi.org/10.5194/egusphere-egu21-2308.

Gou, J., M. Kiani Shahvandi, R. Hohensinn, B. Soja (2023): "Ultra-short-term prediction of LOD using LSTM neural networks"; Journal of Geodesy, external page https://doi.org/10.1007/s00190-023-01745-x

Halbheer M (2021) Prediction of atmospheric parameters from GNSS observations and weather models with machine learning. Bachelor Thesis, ETH Zurich, 2021. external page https://doi.org/10.3929/ethz-b-000573133

Kaselimi M, Doulamis N, Doulamis A, Delikaraoglou D (2020) a sequence-to-sequence temporal convolutional neural network for ionosphere prediction using GNSS observations. The International Archives of Photogrammetry. Remote Sensing and Spatial Information Sciences 43: 813-820

Kaselimi M, Voulodimos A, Doulamis N, Doulamis A, Delikaraoglou (2021) Deep recurrent neural networks for ionospheric variations estimation using GNSS measurements. IEEE Transactions on Geoscience and Remote Sensing vol. 60, pp. 1-15, 2022, Art no. 5800715, external page https://doi.org/10.1109/TGRS.2021.3090856.

Kaselimi M, Voulodimos A, Doulamis N, Doulamis A, Delikaraoglou D (2020) A causal long short-term memory sequence to sequence model for tec prediction using GNSS observations. Remote Sensing 12(9): 1354.

Kiani Shahvandi M, Soja B (2021) A new spatio-temporal graph neural network method for the analysis of GNSS geodetic data. EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-545, external page https://doi.org/10.5194/egusphere-egu21-545.

Kiani Shahvandi, M., B. Soja (2022): "Inclusion of data uncertainty in machine learning and its application in geodetic data science, with case studies for the prediction of Earth orientation parameters and GNSS station coordinate time series"; Advances in Space Research 70, 3, pp. 563-575, external page https://doi.org/10.1016/j.asr.2022.05.042

Kiani Shahvandi, M., B. Soja (2022): "Small Geodetic Datasets and Deep Networks: Attention-Based Residual LSTM Autoencoder Stacking for Geodetic Time Series"; In: "Machine Learning, Optimization, and Data Science. LOD 2021", G. Nicosia, V. Ojha, E. La Malfa, G. La Malfa, G. Jansen, P.M. Pardalos, G. Giuffrida, R. Umeton (eds.), Lecture Notes in Computer Science, vol 13163, pp. 296-307, Springer, Cham. external page https://doi.org/10.1109/10.1007/978-3-030-95467-3_22

Kiani Shahvandi, M., J. Gou, M. Schartner, B. Soja (2022): "Data Driven Approaches for the Prediction of Earth's Effective Angular Momentum Functions"; In: "IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium", pp. 6550-6553, external page https://doi.org/10.1109/IGARSS46834.2022.9883545

Kiani Shahvandi, M., M. Schartner, B. Soja (2022): "Neural ODE differential learning and its application in polar motion prediction"; Journal of Geophysical Research: Solid Earth, 127(11), external page https://doi.org/10.1029/2022JB024775

Marsocci, V., Coletta, V., Ravanelli, R., Scardapane, S., & Crespi, M., Inferring 3D change detection from bitemporal optical images, ISPRS Journal of Photogrammetry and Remote Sensing, 196, pp. 325-339, 2023.

Marsocci, V., Scardapane, S., Continual Barlow Twins: continual self-supervised learning for remote sensing semantic segmentation, IEEE Journal on Selected Topics in Applied Earth Observation and Remote Sensing, in press, 2023.

Meganadh, D., Vipin Maurya, Ashutosh Tiwari, Ramji Dwivedi, 2021. A multi-criteria landslide susceptibility mapping using deep neural networks, Advances in Space Research, DOI: external page 10.1016/j.asr.2021.10.02.

Natras R, Schmidt M (2020) Relationship between ionosphere VTEC and space weather indices for machine learning-based model development EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18978, external page https://doi.org/10.5194/egusphere-egu2020-18978.

Natras R, Schmidt M (2021) Ensemble machine learning for geodetic space weather forecasting. Scientific Assembly of the International Association of Geodesy 2021, June 28 – July 2, 2021.

Natras R, Schmidt M (2021) Ionospheric VTEC Forecasting using machine learning. EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8907, external page https://doi.org/10.5194/ egusphere-egu21-8907.

Natras R., Goss A., Halilovic D., Magnet N., Mulic M., Schmidt M., Weber R.: Regional Ionosphere Delay Models Based on CORS Data and Machine Learning. NAVIGATION: Journal of the Institute of Navigation, 70(3), navi.577, external page https://doi.org/10.33012/navi.577, 2023

Natras R., Schmidt M.: Machine Learning Model Development for Space Weather Forecasting in the Ionosphere. CEUR Proceedings of the CIKM 2021 Workshops co-located with the 30th ACM International Conference on Information and Knowledge Management (CIKM 2021), Online, 2021, external page https://ceur-ws.org/Vol-3052/short10.pdf

Natras R., Schmidt M.: Machine Learning Model Development for Space Weather Forecast. Workshop on Complex Data Challenges in Earth Observation (CDCEO) 2021 at the 30th ACM International Conference on Information and Knowledge Management (CIKM), Online, 2021 (Poster)

Natras R., Schmidt M.: Time-series Forecasting of Ionospheric Space Weather using Ensemble Machine Learning. Workshop Women in Machine Learning (WiML) at the Thirty-eighth International Conference on Machine Learning (ICML) 2021, online, 2021 (Poster)

Natras R., Soja B., Schmidt M., Dominique M., Türkmen A.: Machine Learning Approach for Forecasting Space Weather Effects in the Ionosphere with Uncertainty Quantification. European Geosciences Union (EGU) General Assembly, Vienna, Austria, external page https://doi.org/10.5194/egusphere-egu22-5408, 2022

Natras R., Soja B., Schmidt M.: Ensemble Machine Learning of Random Forest, AdaBoost and XGBoost for Vertical Total Electron Content Forecasting. Remote Sensing, 14(15), 3547, external page https://doi.org/10.3390/rs14153547, 2022

Natras R., Soja B., Schmidt M.: Interpretable Machine Learning for Ionosphere Forecasting with Uncertainty Quantification. D4G: 1st Workshop on Data Science for GNSS Remote Sensing, Potsdam, Germany, 2022

Natras R., Soja B., Schmidt M.: Machine Learning Ensemble Approach for Ionosphere and Space Weather Forecasting with Uncertainty Quantification. 2022 3rd URSI Atlantic and Asia Pacific Radio Science Meeting (AT-AP-RASC), 1-4, external page https://doi.org/10.23919/AT-AP-RASC54737.2022.9814334, 2022

Natras R., Soja B., Schmidt M.: Machine Learning Ensemble Approach for Ionosphere and Space Weather Forecasting with Uncertainty Quantification. 3rd URSI Atlantic / Asia-Pacific Radio Science Conference (URSI AT-AP-RASC 2022), Gran Canaria, Spain, 2022

Natras R., Soja B., Schmidt M.: Uncertainty Quantification for Ionosphere Forecasting with Machine Learning. International Workshop on GNSS Ionosphere (IWGI2022) - Observations, Modelling and Applications, Institute for Solar-Terrestrial Physics, German Aerospace Center (DLR), Neustrelitz, Germany and online, 2022

Pan S., Chen K., Chen J., Qin Z., Cui Q. and Li J. (2020). A partial convolution-based deep-learning network for seismic data regularization. Computers and Geosciences, 145, 104609.

Poian V., Theiling B., Clough L., McKinney B., Major J., Chen J. and Horst S. (2023). A machine learning approach for ocean worlds analog mass spectrometry: Exploratory data analysis (EDA). Frontiers, accepted.

Ruttner P (2021) Analysis and prediction of long term GNSS height time series and environmental loading effects. Master Thesis, ETH Zurich.

Ruttner, P., R. Hohensinn, S. D’Aronco, J.D. Wegner, B. Soja (2022): "Modeling of Residual GNSS Station Motions through Meteorological Data in a Machine Learning Approach"; Remote Sensing, 2022, 14, 17, external page https://doi.org/10.3390/rs14010017

Schartner M, Plötz C, Soja B (2021) Automated VLBI scheduling using AI-based parameter optimization. Journal of Geodesy 95: 58, external page https://doi.org/10.1007/s00190-021-01512-w

Schartner M, Plötz C, Soja B (2021) Improved VLBI scheduling through evolutionary strategies, EGU General Assembly 2021, 19–30 Apr 2021, EGU21-1250, external page https://doi.org/10.5194/egusphere-egu21-1250.

Shehaj, E., L. Miotti, A. Geiger, S. D'Aronco, J. D. Wegner, G. Moeller, B. Soja, M. Rothacher (2023): "High-Resolution Tropospheric Refractivity Fields by Combining Machine Learning and Collocation Methods to Correct Earth Observation Data"; Acta Astronautica, 204:591-598, external page https://doi.org/10.1016/j.actaastro.2022.10.007

Shelley Stall, Guido Cervone, Caroline Coward, et al. Ethical and Responsible Use of AI/ML in the Earth, Space, and Environmental Sciences . ESS Open Archive . April 12, 2023. external page https://doi.org/10.22541/essoar.168132856.66485758/v1

Shujian S (2021) Inversion of rock properties using the fully connected neural network. Master Thesis, The University of Tulsa.
Shum, C. K., Zhang, Y., Jia, Y., Ding, Y., Guo, J., Akyılmaz, O., Uz, M., Zhang, C. and Atman, K.: Geodesy as the Sentinel for Climate-induced Hazards Monitoring and Response. In AGU Fall Meeting Abstracts, Vol. 2022, pp. G16A-07, December 2022.

Soja, B., G. Kłopotek, Y. Pan, L. Crocetti, S. Mao, M. Awadaljeed, M. Rothacher, L. See, T. Sturn, R. Weinacker, I. McCallum, V. Navarro (accepted): "Machine learning-based exploitation of crowdsourced GNSS data for atmospheric studies"; 2023 IEEE International Geoscience and Remote Sensing Symposium IGARSS

Srivastava, A., Ashutosh Tiwari, Avadh Bihari Narayan, and Onkar Dikshit, 2022. InSAR phase unwrapping using Graph neural networks, EGU General Assembly 2022, Vienna, 23-27 May 2022, DOI: external page 10.5194/egusphere-egu22-11010

Tiwari A, Avadh Bihari N, Onkar D (2020) Deep learning networks for selection of measurement pixels in multi-temporal SAR interferometric processing, ISPRS Journal of Photogrammetry and Remote Sensing 166: 169-182.3

Tiwari A, Avadh Bihari N, Onkar D (2021) A deep learning approach for efficient multi-temporal interferometric synthetic aperture radar (MT-InSAR) processing. EGU General Assembly 2021, 19–30 Apr 2021, EGU21-12784, external page https://doi.org/10.5194/egusphere-egu21-12784.

Tiwari, A. and Manoochehr Shirzaei, 2023. A novel machine learning and deep learning based semi-supervised learning approach for information extraction from InSAR-derived deformation maps. In preparation.

Tiwari, A., Jonathan Lucy and Manoochehr Shirzaei, 2023. A novel unsupervised LSTM-Autoencoder network for information extraction from InSAR derived deformation. In preparation.

Uz, M., Akyılmaz, O., and Shum, C.: Deep Learning-aided Temporal Downscaling of Satellite Gravimetry Terrestrial Water Storage Anomalies Across the Contiguous United States (CONUS), EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-632, external page https://doi.org/10.5194/egusphere-egu23-632, 2023.

Uz, M., Atman, K.G., Akyilmaz, O., Shum, C.K., Keleş, M., Ay, T., Tandoğdu, B., Zhang, Y. and Mercan, H., (2022) Bridging the gap between GRACE and GRACE-FO missions with deep learning aided water storage simulations, Science. of the Total Environment, 830: 154701, external page https://doi.org/10.1016/j.scitotenv.2022.154701.

Uz, M., Akyilmaz, O., Shum, C., Atman, K.G., Olgun, S., Güneş, Ö., 2023. Deep Learning-Aided High-Resolution Temporal Gravity Field Simulations: Monthly global mass grids and Spherical Harmonics from 1994 to 2021, IUGG 28th General Assembly, Berlin, 11-20 July 2023.

Wicki J (2021) Optimizing geodetic VLBI simulation parameters based on swarm intelligence. Bachelor Thesis, ETH Zurich.
Wu Y, Pan S., Chen J., Song G. and Gou Q. (2022). A Surface-wave Inversion Method Based on FHLV Loss Function in LSTM. IEEE Transactions on Geoscience and Remote Sensing Letters. external page 10.1109/LGRS.2022.3187020.

Wu Y., Pan S., Chen Y., Chen J., Yi S., Zhang D. and Song G. (2023). An unsupervised inversion method for seismic brittleness parameters driven by the physical equation. IEEE Transactions on Geoscience and Remote Sensing Letters. external page 10.1109/TGRS.2023.3273302.