| Abstract |
Interferometric Synthetic Aperture Radar (InSAR) allows, by the acquisition of successive radar data in time, for monitoring the ground movements with millimeter precision. The measurement quality and spatial coverage is strongly influenced by the presence of permanent reflectors. Geothermal activity is particularly developed in the southern suburbs of Paris, with a high density of geothermal doublets. The area is characterized by an extensive human activity; the area shows a limited vegetation cover and numerous anthropic reflectors (buildings, roads, urban materials), which is particularly favorable for the application of the InSAR technique. Geothermal exploitation in this region involves mainly the deep geological formation of the Dogger, a 1 700 meter deep limestone aquifer. Due to the thickness and nature of the overlying terrains, the impact of deep pressure variations is expected to be limited at surface. In fact, a simple geomechanical analysis has concluded that the expected effects are millimetric, which is the same order than the radar interferometry measurement uncertainty. The present study takes advantage of the large stack of High Resolution TerraSAR-X images archived over Paris, over a 65 month-period (75 measurement dates). The objective is to evaluate the InSAR approach suitability for the assessment of ground surface movement in Paris suburbs. An innovative methodology has been implemented for the exploitation of the InSAR measurement. It is based on traditional statistical and geostatistical methods to characterize the spatio-temporal distribution of the InSAR displacements, and to identify local extrema. The process is entirely blind and no a priori information on the location nor on the exploitation rates of the doublets is integrated in the calculations. The applied methodology highlighted several important phenomena over the studied area and the different statistical attributes computed over the entire period, such as velocity and acceleration, revealed large regional events, partly correlated with topography. The temporal analysis of the displacements highlighted strong seasonal effects, with a tendency to a regional ground elevation in winter and a relative subsidence in summer. However, the mechanism causing these periodical variations was not specifically identified: thermal effects, groundwater level variations, or soil properties, or even an association of different factors. A spatial decomposition of the displacements was carried out, at each measurement date, by successive factorial kriging. This decomposition allowed for the dissociation of ground movements at the regional scale, probably related to topographical, geological or climatic factors, from local movements. The analysis of local displacement maps (mean and standard deviation maps over the period studied), revealed some local anomalies, mainly affecting roads, railways and specific buildings. In conclusion, the InSAR analysis conducted did not clearly highlight shifts, subsidence or elevation, specifically related to geothermal activity, showing that the Dogger geothermal exploitation has less impact at surface than regular anthropic activities and natural events. This approach could be very useful in case ground deformation in urbanized environments would need to be further investigated. |