Observatoire Multidisciplinaire des Instabilités de Versants

Scientific and Technical Project

Scientific interest of landslide monitoring observation in short and long terms and current international state of research

Landslides are a major threat to human life, property and constructed facilities, infrastructure and natural environment in most mountainous and hilly regions of the world (e.g. Petley, geology, 2012). In the last century, Europe has experienced the second highest number of fatalities and the highest economic losses caused by landslides compared to other continents (EM-DAT, 2003): at least 16,000 people have lost their lives because of landslides and the material losses amounted to over € 1400 M in Europe during the 20th century. Apart from human casualties, most of landslides impact widespread disruption to roads and buildings. It corresponds to direct and indirect economic costs. Actual losses produced by landslides in France highlight the lack of a risk mitigation program including the short-term prediction of rapid landslides and integrated early-warning systems.

As a consequence of climate change and increase in exposure, the risk associated with landslides is growing. In areas with high demographic density, protection works when possible to be built are not demonstrated to be efficient because of either economic or environmental constraints (e.g. gorges de la Bourne, Vercors Massif).

The mechanisms that drive the landslide triggering and their dynamics are numerous. Their couplings result in complex patterns that induced large uncertainty in the prediction of both the landslide size and their time of occurrence. This context induces worldwide a misfit between the scientific community knowledge, the responsibility of technical persons in charge of the risk management and the expectation of societal citizens. Contrarily to other natural hazards (earthquake, river flood, volcano eruptions…) there are a few data sets that relate the evolution toward collapse for landslides. This lack of available data set worldwide is more striking in the continental French context where landslides induce economical and societal costs that largely overpass the one induced by earthquakes.

(a) La Salle en Beaumont, soft rock (clay) landslide (Isère) 1,3 * 10 6 m3, 1994, 4 casualties (b) Vercors, 2004, rockslide, 2000 m3, 2 casualties. In both cases, retrofitting the roads were major societal costs.

Therefore the objectives of the OMIV observatory is to provide the scientific community the data sets that will allow:

  1. to identify the physical processes that control the slope instabilities,
  2. to be able to model them,
  3. and to extract from the observables the patterns that may characterize a change in the land-sliding mode.

There is now an important gap, both at the French level and worldwide, between the ability to simulate numerically or physically conceptual slope movements and to specifically reproduce data from past case studies or ongoing landslides.

Moreover, when a given empirical law is a-posteriori adjusted on pre-instability observations, it is rarely validated by multi-parameter observations. As example, the power law divergence of displacements suggested by Voight, (1988, 1989) are reported for displacement values only, without possible tests on the corresponding seismicity data. Similarly, the power law divergence of seismicity rate and energy observed by Amitrano et al. (2005) before a cliff collapse cannot be tested against displacement data, because these latter were not monitored on this site.

Back in 2007’s, both at the French national and at the worldwide level, there was no active slope sliding on which long lasting continuous records, for sensitive range of osbservables (displacement, seismology, hydro-geochemistry) were either recorded or available.

At the national level during the late 1990’s and the early 2000’s, the major projects merged towards 4 ongoing landslides. The projects, which were recurrently funded by regional and national funding (e.g. PGRN, PNRN, ACI ‘CatNat’, ACI ‘Changement Climatique’, ACI ‘Reliefs de la Terre’, ANR ECCO, ANR CTT), gathered investment of several teams on La Clapière, Super-Sauze et Séchilienne sites. These projects result (i) to a progressive long term monitoring of the landslide displacements, (ii) to built up short term experiment which constrain the geometry and the kinematics and the water circulation within the moving landslide.

To stabilize recurrent long lasting data recording and to homogenize the measurements on these 4 test sites, 3 teams merge to federate through the OMIV observatory since 2007. Two other groups will join the OMIV federation starting January 2013 (Chrono-Envirronnement, Besançon and EMMAH, Avignon) to be in charge of the hydro-geochemistry measurement on hard rock and soft rock site, respectively.

Objectives for the landslide observatory:

Within the context described above, the SO-OMIV targets are defined as:

  1. to stabilize and sustain over time the recording of 3 observable classes (kinematics, seismology, hydro-geochemistry of slope) on representative ongoing landslides;
  2. to homogenize measurements on the four landslides (parameter types and space and time sampling);
  3. to archive the data and to provide free real time open access to the data for the wide scientific community;
  4. at a longer term target, to provide the users specific software and algorithms through a WEB base platform. These softwares are tools to analyze and process OMIV data as well as numerical models to simulate the geo-mechanical processes and associated signals that are able to reproduce the observed signals.

These observations aims to contribute to (i) identify the key control parameters that allow to analyze different type of slope instabilities (i.e. soft/hard rock, cohesion/friction, slip/fracture, localized/diffuse damage,….).and (ii) to follow its evolutions through time and space (slowing down or accelerating up, sliding-flow transition,…)

Going beyond some existing case studies in Europe, the key aspect of the French OMIV observatory is embedded in the multi-parameters monitoring and the open access data for the worldwide community. More specifically, when compared to ongoing monitoring in Italy (Rosone, Tessina, Corvara, sites among others), Spain (e.g.Vallcebre, Casas), Switzerland (e.g. La Frasse, Falli-Hölli) and England (e.g. Roughs, St Catherine Point), the French case studies selected by OMIV shares on the four sites, (i) a long history of data monitoring (ii) for landslides typical of mechanisms that affect alpine versants. These points are relevant for the involved material as well for the volume and velocity of the moving masses.

Location of the 4 landslides monitored by OMIV observatory.

The OMIV observatory monitors 4 active landslides in the French Alps (Avignonet, Super-Sauze, La Clapière, Séchilienne), each of them having a minimum of 10 years of data history, in 2007 when the SO-OMIV was first INSU labeled for a 2 year period. Possible extension to monitor the peculiar case study of volcano slope, non available within inland France, will be discussed with the (SC) Scientific Council for OMIV as soon as the monitoring and data accesses for the 4 sites will be in a routine phase. One may not as possible, yet well monitored, site the cliff rock-falls at Piton de la Fournaise as monitored by OVPF, Réunion Island, networks and the response of volcano slope to strong earthquake at La Guadeloupe, Island). The logic that builded up the OMIV observatory converges with the aims of recently national and EC funded projects to observe and to understand and to model slope instabilities (e.g. ANR ECCO PNRH ‘Ecou-Pref: Ecoulements Préférentiels dans les versants marneux’ (2006-2008)  ; ANR CTT TRIGGERLAND ‘Triggering mechanisms of Landslides: analysis and modelling’ (2006-2009); 6ème PCRDT NEST Tackling complexity in Science ‘Triggering of Instabilities in Materials and Geosystems : earthquakes, landslides and snow avalanches’ (2006-2010); 6ème PCRDT Research Training Network Marie Curie ‘Mountain Risk’ (2007-2010), ANR SISCA, ANR Slams 2009-2012).

“Added value” for SO-OMIV:

The major input of SNO-OMIV is to provide a global homogeneous monitoring that apply to both the soft - and the hard- rock landslide classes. This approach was first successfully tested during ACI ‘CatNat’ SAMOA (2002-2004, C. Delacourt, O. Maquaire & D. Amitrano) project. The OMIV approach also gains momentum from bringing together 3 types of instrumental monitoring expertise with the scientific know-how of communities who did not used to interacts on landslide study, before SO-OMIV started. Accordingly the SO-OMIV bring together geo-morphologists, geologists and hydro- geologists, geo-physicists and geo-chemists, rock, soil and fluid physicists working either at lab scale or in-situ or as modeling experiments. Such a multidisciplinary group is well fitted to understand the complex patterns of slope instabilities where soft and hard rocks interact with fluid within a wide range of fracture patterns.

From a broad scientific perspective, the knowledge derived from the complex coupling between fluid pressure – air and rock temperature – stress/strain within heterogeneous media with measurement close to the sliding surface and away from it may also be of major interest to understand the mechanics of rupture and sliding which is the key issues at all scale within the earth crust. As an example the shallow sliding surface can be a proxy for slow slip faulting as advertise on numerous deeper fault in the recent years.

As a more practical implication, groups in charge of alerts and risk assessment for landslides (RTM , Restauration des Terrains en Montagne (RTM), and regional IFSTAR-CETE, local representatives) will appreciate new data and tools (methodology and models) that are efficient and that can be used to improve the decision process for risk management. Most of the empirical methods used by these groups, in France and abroad, are expert based tools which build up on a large number of case studies. For all these current techniques to quantify the uncertainties remains a very difficult task. One the major contributions of the SO-OMIV is to transfer new technical and quantitative approaches to the person in charge of the risk mitigation.

Originality of a French approach

As compared to other region worldwide the landslides in France occurs at a weak rate within moderate climate and low seismicity forcing, respectively. This context allow for slow slope movement which map the route for:

  1. accurate and continuous sampling of the pre-catastrophic phase;
  2. several interactions and feedback between model outputs and updated data inputs;
These two points are the basis for a comprehensive analysis of the physical processes that drive the transition from slow stable slips to fast catastrophic failure. The analysis of such a pre-failure phase are impossible in area for the strong forcing (rain, earthquake) squeezes to acceleration phases beyond most monitoring techniques worldwide. The advantages of the slow moving slopes, as summarized above, may also turn to possible drawback when the time to failure is never reached or is too far away for any scientific feedback. Whenever the catastrophic slip velocity is reached, the follow up on the slipping dynamics allows for a long lasting monitoring of geophysical slips in a context of weaker confining pressure that the one where seismogenic slip localizes. This way the moving slopes emerge as in situ analog to, more easy to observe than, seismic slips.

"Added value" of OMIV federation

As a new National service for INSU observatories the specificities of the SO- OMIV is grounded in the multi-lab team structure that merge as a federation. The 3 lab involved since 2007 imply 3 observatory and 2 new lab enter the OMIV federation next January. It allow for a unique consortium to work together as merging together multidisciplinary data for multidisciplinary groups. The cross correlated analysis of the data will enhance any of the single analysis of one observable.