Skip to Main content Skip to Navigation
Theses

Imagerie multi-paramètres et multi-résolutions pour l'observation et la caractérisation des mécanismes de glissements-coulées

Abstract : Controlling factors (predisposition, triggering) of landslides as well as their mechanism and kinematic behaviour can be extremely heterogeneous and interact on very versatile time constants. Numerous gaps exist in the evaluation of the hazard associated with this process, in particular for slow landslides which can evolve into faster slides. Indeed, landslides show some behaviour which is very variable in time and space which is characterized by a large range of displacement rates (from less than 1 centimetre per year to a few meters per day) and the presence of irregularities and large heterogeneities in the distribution of their petrophysical properties. Classical techniques of observation and investigation allow us to get only localized information which is not enough to quantify the time and spatial variability. As a result of which landslide numerical models are calibrated and validated over a limited amount of data. Recent developments in multiparameter geophysical imaging allowed some improvements in the direct and indirect acquisition of data on the deformation (photogrammetry, images correlation, laser scanner) and petrophysical parameters (electric tomography and seismic refraction). Although not as precise as the classical techniques, they have the advantage of providing multi-scaled and spatially distributed information. A major difficulty is to combine this spatial information to the ones obtained with classical technologies (GPS, extensometry, piezometry, geotechnology) and integrate them into a conceptual model coherent for the modelisation. Some scientific works have begun to combine different approaches obtained from field observations (geomorphology, geology) and instrumental data (hydrogeophysics, photogrammetry, laser scan). The objective is to develop methodologies which allow one to spatially determine the diverse major characteristics of landslides (intern structuration, hydrologic behaviour, kinematic behaviour, deformation mechanism). A conceptual model of the functioning and modelling of the hydro-mechanical behaviour of landslides with Z-soil and Slow-Mov codes is proposed. The sites selected for our analysis are the marly landslides of Super-Sauze and La Valette in the basin of Barcelonnette (Alpes-de-Haute-Provence). The research is presented in four parts and the objectives are as follows: 1. To characterize the internal structure and 3D geometry of landslides with the help of geophysical, geotechnical and geomorphologic data ; 2. To propose a hydrological conceptual model of the non saturated zone followed by temporal and spatial monitoring of electric resistivity and piezometric data analysis ; 3. To characterize the kinematics of landslides from terrestrial remote sensing platforms (photogrammetry and laser scan) combined with aerial platforms and differential GPS monitoring. 4. To identify the thresholds of hydrological and kinematics regime changes with numerical modelling. The results highlight the vast scope of application of the main imaging technologies for the investigation of landslides and allow us to define their limitations of use. Some future research will develop these various methodologies for surveillance purposes and to improve the predictive capabilities of the hazard.
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-00806806
Contributor : Jean-Philippe Malet <>
Submitted on : Tuesday, April 2, 2013 - 1:59:44 PM
Last modification on : Thursday, March 26, 2020 - 2:20:00 AM
Document(s) archivé(s) le : Sunday, April 2, 2017 - 11:17:43 PM

Identifiers

  • HAL Id : tel-00806806, version 1

Collections

Citation

Julien Travelletti. Imagerie multi-paramètres et multi-résolutions pour l'observation et la caractérisation des mécanismes de glissements-coulées. Géophysique [physics.geo-ph]. Université de Caen, 2011. Français. ⟨tel-00806806⟩

Share

Metrics

Record views

524

Files downloads

2491