Studying earth’s surface deformations using machine learning and GNSS-integrated data

Crustal motions refer to the movement and deformation of the Earth’s outermost solid layer, its crust. These motions can occur due to various geological processes and forces, such as tectonic activities and isostatic adjustments. In addition to natural processes, human activities can also aggravate deformation. Understanding crustal motions is essential for studying geodynamics, seismic hazards, and long-term forecasts of the Earth’s crust changes. Germany’s geology is characterized by diverse structures, including mountain ranges (such as the Harz and Black Forest) and sedimentary basins (such as the North German Basin). This diverse geology influences crustal movements in different regions. Moderate seismic activity occurs in Germany, particularly in areas such as the Rhine Graben and the Swabian Jura. These regions are characterized by their extensional forces and faulting, which lead to crustal deformation. Also, human activities such as mining, fluid extraction, and reservoir-induced seismicity can contribute to localized crustal deformation. Therefore, this study is expected to provide valuable information to improve understanding and inform decisions on infrastructure planning, hazard mitigation, and resource management in Germany. Various space geodetic techniques are utilized to monitor crustal motions, including Satellite Laser Ranging (SLR), Doppler Orbitography and Radiopositioning Integrated by Satellites (DORIS), Very Long Baseline Interferometry (VLBI), Global Navigation Satellite Systems (GNSS), Interferometric Synthetic Aperture Radar (InSAR), and seismometers. These techniques provide precise measurements of movements. In this study, I introduce dedicated algorithms for precision time series processing to analyze movements and detect deformation risks and potential hazards. These algorithms utilize advanced data processing techniques and sophisticated models to identify patterns, trends, and anomalies in deformation analyses. Furthermore, by leveraging Machine Learning (ML) techniques, I developed methodologies to train forecast models that contribute to the development of early warning systems, determine the dynamic mechanics of seismic events, and observe interactions among the Earth’s spheres. Data from the disciplinary domains of space geodesy, space geophysics, and seismology are utilized to build ML models for forecasting and detecting atmospheric anomalies caused by geohazards. My methodologies have been tested through case studies on the Ring of Fire region, including the 2011 Tohoku earthquake in Japan, the 2015 St.Patrick’s Day geomagnetic storms, the 2017 nuclear explosions in North Korea, and the 2022 Hunga Tonga-Hunga Ha’apai eruption in the Southern Pacific Ocean, as well as other globally typical seismic events. These investigations also identify key drivers and factors that impact Earth’s spheres, enabling the development of action strategies to address climate change and provide early warnings for the sustainable management of natural resources.