Ziel
Recent space-based geodetic measurements of ground deformation suggest a paradigm shift is required in our understanding of the behaviour of active tectonic faults. The classic view of faults classified in two groups – the locked faults prone to generate earthquakes and the creeping faults releasing stress through continuous aseismic slip – is now obscured by more and more studies shedding light on a wide variety of seismic and aseismic slip events of variable duration and size. What physical mechanism controls whether a tectonic fault will generate a dynamic, catastrophic rupture or gently release energy aseismically? Answering such a fundamental question requires a tool for systematic and global detection of all modes of slip along active faults.
The launch of the Sentinel 1 constellation is a game changer as it provides, from now on, systematic Radar mapping of all actively deforming regions in the world with a 6-day return period. Such wealth of data represents an opportunity as well as a challenge we need to meet today. In order to expand the detection and characterization of all slip events to a global scale, I will develop a tool based on machine learning procedures merging the detection capabilities of all data types, including Sentinel 1 data, to build time series of ground motion.
The first step is the development of a geodetic data assimilation method with forecasting ability toward the first re-analysis of active fault motion and tectonic phenomena. The second step is a validation of the method on three faults, including the well-instrumented San Andreas (USA) and Longitudinal Valley faults (Taiwan) and the North Anatolian Fault (NAF, Turkey). I will deploy a specifically designed GPS network along the NAF to compare with outputs of our method. The third step is the intensive use of the algorithm on a global scale to detect slip events of all temporal and spatial scales for a better understanding of the slip behaviour along all active continental faults.
Wissenschaftliches Gebiet
CORDIS klassifiziert Projekte mit EuroSciVoc, einer mehrsprachigen Taxonomie der Wissenschaftsbereiche, durch einen halbautomatischen Prozess, der auf Verfahren der Verarbeitung natürlicher Sprache beruht.
CORDIS klassifiziert Projekte mit EuroSciVoc, einer mehrsprachigen Taxonomie der Wissenschaftsbereiche, durch einen halbautomatischen Prozess, der auf Verfahren der Verarbeitung natürlicher Sprache beruht.
- natural sciencesearth and related environmental sciencesgeologyseismology
- engineering and technologyelectrical engineering, electronic engineering, information engineeringinformation engineeringtelecommunicationsradio technologyradar
- natural sciencescomputer and information sciencesartificial intelligencemachine learning
Programm/Programme
Thema/Themen
Finanzierungsplan
ERC-STG - Starting GrantGastgebende Einrichtung
75230 Paris
Frankreich