Réf. Malet 2006 - C1

Remarques sur l'organisation de la matrice d'impact, références sur des tentatives de corrélation entre climat passé et activité des mouvements de terrain et références sur des prédictions d'activité pour des scénarios de changement global
Jean-Philippe Malet

Climate change, landslides, impact assessment.

Organisme(s) / Contact(s)
Faculty of geosciences of Utrecht ; j.malet@geo.uu.nl

Références des études sur lesquelles porte le commentaire

(1) - Paramètre(s) atmosphérique(s) modifié(s)
(2) - Elément(s) du milieu modifié(s)
(3) - Type(s) d'aléa modifié(s)
(3) - Sous-type(s) d'aléa
Températures - Précipitations   Mouvements de terrain  

Pays / Zone
Massif / Secteur
Site(s) d'étude
Période d'observation
Monde / Europe / Alpes          


Establishing link between climate & landslides is very difficult:

• because of the relatively few records of landslides events (date, incomplete databases) in the last century, in the Little Ice Age and at least during the Holocene. There have been some attempts to estimate these links, especially for the Italian and Swiss Alps, and to some extents for the French Alps;

• because of the variety of landside types (e.g. the term slope movements or slope instability or slope failures would be preferable to use in the project), shallow or more deep-seated, affecting earth, debris or rock materials, etc, and therefore controlled by a variety of factors. In relation to hydroclimatic events, landslides can be triggered (not exhaustive):
- by intense rainfalls (especially shallow soil slips and debris flows),
- by a sequence of long duration rainfalls and hence a modification of the regional groundwater tables,
- by undercutting related to a rise in the discharge of streams;

• because landslide activity related to climate can be:
- episodic (one-step landslide, e.g. failure -> runout -> definitive stop),
- not only episodic (the same dates/periods of activity are in common each year, for instance for slow-moving persistently active mudslides or rockslides).

There is actually no general clearance or evidence of a relation between coldest periods and less landslide activity or warmest periods and higher landslide activity, even if it is fairly obvious that a period of increased rainfall will increase landslide activity. Matthews et al. (1997) demonstrated (all over Europe) an increase in landslide activity during the Little Ice Age (characterized by increased rainfall combined to a lowering of temperatures).

But our actual climate change scenarios indicate increased rainfall combined to a rise of temperatures and increased rainfall in Winter and less in Summer… We have no idea if this "new" combination will do, and increase or not slope failures. One "empirical" statement is that (for the Alps) wetter Winters and drier Summers is more or less the climate actually experienced by Italy or the Spanish Pyrenees… which are "disaster lands" regarding slope instability.