Research

Geodetic observations allow us to determine precise positions on Earth and the Earth’s orientation in space. High accuracy of geodetic products is essential for applications ranging from navigation to monitoring climate and natural hazards. With advances in space-geodetic techniques, such as GNSS, satellite gravimetry, and VLBI, new opportunities arise to exploit the growing volume and diversity of geodetic data.

Our group focuses on applying artificial intelligence to enhance geodetic monitoring and prediction. Using GNSS and related observations, we model key atmospheric parameters such as tropospheric water vapor and ionospheric electron content, improving weather, climate, and space-weather applications. We also develop machine learning approaches to study surface mass change in the hydrosphere, cryosphere, and solid Earth, creating high-resolution models and linking long-term mass variations to changes in Earth's rotation.

In parallel, we advance space-geodetic methodologies by developing new processing strategies, uncertainty-aware and physics-informed AI models, and algorithms for spatio-temporal prediction, anomaly detection, and data fusion. We further contribute to improved VLBI scheduling and explore innovative approaches for low-cost GNSS data exploitation and orbit determination of large nanosatellite constellations.