Benedikt Soja


Last Name

Soja

First Name

Benedikt

ORCID

0000-0002-7010-2147

Organisational unit

09707 - Soja, Benedikt / Soja, Benedikt

Search Results

Publications 1 - 10 of 243
  • On the detection of structural breaks in GNSS station coordinate time series caused by earthquakes using machine learning
    Item type: Other Conference Item
    Crocetti, Laura; Schartner, Matthias; Soja, Benedikt (2021)
    IAG 2021 Abstract Book: Geodesy for a Sustainable Earth
  • Modelling the Troposphere with Global Navigation Satellite Systems, Meteorological Data and Machine Learning
    Item type: Conference Paper
    Crocetti, Laura; Schartner, Matthias; Schindler, Konrad; et al. (2024)
    IGARSS 2024 - 2024 IEEE International Geoscience and Remote Sensing Symposium
    Global Navigation Satellite Systems (GNSS), such as the American Global Positioning System (GPS) and the Euro pean Galileo system, are capable of monitoring tropospheric properties. An important parameter describing the tropo spheric impact on GNSS is zenith wet delay (ZWD), which is highly correlated to the amount of water vapour in the troposphere and thus interesting for atmospheric and climate research. This work demonstrates how GNSS observations help to sense the atmosphere and its dynamics by using a newly developed machine learning-based ZWD model. The model provides ZWD globally for the years 2010 to 2023 with a positive trend in the Northern Hemisphere and a negative trend in the Southern Hemisphere. Furthermore, the global average ZWD anomaly follows alternating trends, strongly correlated with the El Nino Southern Oscillation (ENSO) ˜ index, increasing up to a correlation coefficient of 0.74 when introducing a time lag of two months.
  • High-resolution Greenland ice sheet mass anomalies from data fusion using graph neural networks
    Item type: Other Conference Item
    Gou, Junyang; Willen, Matthias O.; Wilms, Josefine; et al. (2024)
  • Modeling the Differences between Ultra-Rapid and Final Orbit Products of GPS Satellites Using Machine-Learning Approaches
    Item type: Journal Article
    Gou, Junyang; Rösch, Christine; Shehaj, Endrit; et al. (2023)
    Remote Sensing
    The International GNSS Service analysis centers provide orbit products of GPS satellites with weekly, daily, and sub-daily latency. The most frequent ultra-rapid products, which include 24 h of orbits derived from observations and 24 h of orbit predictions, are vital for real-time applications. However, the predicted part of the ultra-rapid orbits is less accurate than the estimated part and has deviations of several decimeters with respect to the final products. In this study, we investigate the potential of applying machine-learning (ML) and deep-learning (DL) algorithms to further enhance physics-based orbit predictions. We employed multiple ML/DL algorithms and comprehensively compared the performances of different models. Since the prediction errors of the physics-based propagators accumulate with time and have sequential characteristics, specific sequential modeling algorithms, such as Long Short-Term Memory (LSTM), show superiority. Our approach shows promising results with average improvements of 47% in 3D RMS within the 24-h prediction interval of the ultra-rapid products. In the end, we applied the orbit predictions improved by LSTM to kinematic precise point positioning and demonstrated the benefits of LSTM-improved orbit predictions for positioning applications. The accuracy of the station coordinates estimated based on these products is improved by 16% on average compared to those using ultra-rapid orbit predictions.
  • VLBI Scheduling and Simulations for the Genesis Mission
    Item type: Conference Poster
    Schartner, Matthias; Soja, Benedikt (2025)
  • A New Deep-Learning-Assisted Global Water Vapor Stratification Model for GNSS Meteorology: Validations and Applications
    Item type: Journal Article
    Zhang, Wenyuan; Gou, Junyang; Möller, Gregor; et al. (2024)
    IEEE Transactions on Geoscience and Remote Sensing
    Layer precipitable water (LPW), a water vapor product similar to precipitable water vapor (PWV), reports partial moisture content within a specified vertical range. Compared with PWV data, the latest LPW products can describe more refined distributions and variations in water vapor in the troposphere. Global Navigation Satellite Systems (GNSSs), as a powerful water vapor sensing tool, only provide the opportunity to retrieve all-weather PWV, not LPW products. To this end, we develop the first deep-learning-assisted, global water vapor stratification (GWVS) model to estimate the GNSS LPW within any given vertical range. The proposed model is trained and tested using the global radiosonde data, with the training and testing root mean square error (RMSE) of 0.94 and 1.10 mm for radiosonde LPW, indicating the excellent generalization of the GWVS model. Furthermore, the model is comprehensively validated using the data from the two regional GNSS networks and one global network. The RMSEs of the predicted GNSS LPW from the three GNSS networks compared with the co-located radiosonde LPW are 1.52, 1.80, and 1.54 mm, respectively. To study potential applications, we use the model-derived GNSS LPW products to calibrate Geostationary Operational Environmental Satellite-16 (GOES-16) LPW products and improve the GNSS water vapor tomography technique. Results show that the accuracy of three GOES-16 LPW products is improved by 31.3%, 23.3%, and 17.9%, respectively, and the RMSE of the tomography results is reduced from 2.28 to 1.67g/m(3). Both validation and application results highlight that the GWVS model retrieves the required GNSS LPW products and provides additional value for water-vapor-related studies.
  • Advances in automatized schedule generation at the VLBI operation center DACH and introduction of the VLBI correlation center at Wettzell
    Item type: Conference Paper
    Plötz, Christian; Schartner, Matthias; Böhm, Johannes; et al. (2021)
    Proceedings of the 25th European VLBI Group for Geodesy and Astrometry Working Meeting
    Two newly established VLBI components, the Operation Center DACH and a new VLBI correlation facility are introduced. The Geodetic Observatory Wettzell (GOW), jointly operated by the Federal Agency for Cartography and Geodesy (BKG) and the Technical University of Munich (TUM), was accepted as an IVS Operation Center (OC) on November 19, 2019. This addition complements BKG’s substantial contribution to the IVS and its continuous and long-term VLBI observation programs. Recently, the OC in cooperation with ETH Zurich (ETHZ) and TU Wien (TUW) added automated ¨ schedule generation to its capabilities. This cooperation within the institutions operates now under the unified Operation Center with the abbreviation DACH, unifying BKG, ETHZ and TUW. The assigned VLBI schedules for this operation center are generated with an innovative approach consisting of a Python framework (VieSched++ AUTO), which allows to fully automate the generation of highly optimized schedules with VieSched++. The complete workflow is designed to gather and evaluate all master files, calling the VieSched++ scheduling application and also upload the selected schedule files to the IVS data centers. Additionally, a new website at BKG shows the complete history of all scheduled sessions, including quality oriented parameters with graphical compa isons. Furthermore, an outline of the new VLBI correlation facility at the Geodetic Observatory Wettzell will be presented. This includes technical specifications, milestones and timelines.
  • Comparison of machine-learning-based predictions of Earth orientation parameters using different input data
    Item type: Other Conference Item
    Soja, Benedikt; Kiani Shahvandi, Mostafa; Schartner, Matthias; et al. (2023)
    Earth orientation parameters (EOP) are needed for precise navigation on Earth and in space and to connect the terrestrial to the celestial reference frame, and for several real-time applications. EOP are typically determined from the observations of different space-geodetic techniques. In order to overcome latencies in the processing and combination of these observations, accurate predictions of EOP are essential. To improve the modeling and prediction of EOP, effective angular momentum (EAM) functions, covering the domains of the atmosphere, ocean, and land hydrology, are typically incorporated to achieve the best performance. The Space Geodesy group at ETH Zurich provides operational predictions of EOP and EAM via its Geodetic Prediction Center (GPC; https://gpc.ethz.ch). We have developed machine learning approaches to predict EOP over different time horizons based on existing EOP and EAM times series provided by other institutions. In this contribution, we analyze the quality of our operational predictions with a focus on comparing the results based on different input time series. In particular, we use different rapid EOP products (IERS, SYRTE, and JPL) and EAM forecasts (GFZ and ETH). In terms of EAM, we study the impact of using only atmospheric angular momentum compared to the sum of all different EAM components. Preliminary findings indicate that rapid EOP data from SYRTE or JPL in combination with a full set of EAM functions leads to the best accuracy when considering IERS final products as reference.
  • VLBI Intensive sessions: the selection of baselines for UT1 estimation
    Item type: Other Conference Item
    Kern, Lisa; Schartner, Matthias; Soja, Benedikt; et al. (2021)
    25th European VLBI Group for Geodesy and Astronomy Working Meeting. Information and Book of Abstracts
  • Atmospheric monitoring with GNSS IoT data fusion based on machine learning
    Item type: Conference Poster
    Soja, Benedikt; Navarro, Vicente; Kłopotek, Grzegorz; et al. (2021)
    AGU Fall Meeting Abstracts
Publications 1 - 10 of 243