Discusión

The classication proposed in this thesis is a tool to simplify the geology and hydrogeology of slopes and landslides in order to construct conceptual models. The presented classication is based on the assumption that the slip surface of the landslide develops between two geological layers. In some cases, an intermediate third layer (for example weathered bedrock) has to be taken into account. The classication is not suitable for rockfalls. It is applicable for sliding and owing type landslides in any geological and hydrogeological setting. The classication may be used for shallow, medium and deep seated landslides. Generally, shallow landslides can be classied easier. This is because the hydrogeology of large and deep-seated landslides, for ex-ample La Frasse or Triesenberg landslides tends to be more complex than of shallow landslides.

Even though in the present study the classication was applied for landslides in the Alps, it can be used in all parts of the world where landslides occur. This is because the parameters used in the classication (permeability and saturation) and the resulting triggering mechanism can be described for any geological, hydrogeological and climate conditions in the world. Further-more, the classication may be applied for natural landslides as well as for triggering experiments.

Based on the two parameters "saturation" and "permeability contrast", 18 "permeability contrast-saturation types" have been dened for a two layer model and additional 11 types for a three layer model. 29 types might seem a lot, but several types are similar with regard to the po-tential triggering mechanisms. Several types could be grouped, for example the unsaturated triggered landslides or the landslides where overpressure may play a role. More than half of the analysed landslides occur above conned aquifers and none of the analysed landslide occurs in unsaturated conditions. One might ask why unsaturated conditions are taken into account in the classication. This is because it is known that in some cases of shallow landslides, the decrease of suction is sucient for the triggering (Iverson et al., 1997; Fourie et al., 1999; Godt et al., 2009).

Furthermore, the classication can be used to discuss the changing saturation of a slope during a rainfall or snowmelt event. Therefore it is important to also consider unsaturated conditions (as initial stage).

9.1. APPLICATION OF THE CLASSIFICATION

The application of the presented hydrogeological classication for dierent landslides has shown that the denition of a permeability-saturation type may depend on:

• the part of the landslide (upper part or lower part of the slope)

• the considered moment (before or after a rainfall event)

• the scale and detailedness (mechanism of entire landslide or local processes, especially for large and deep-seated landslides)

Even though a landslide may be composed of several permeability-saturation types, in most cases one type can be dened as the most relevant for the triggering.

The classication can be applied for:

1. The construction of a conceptual hydrogeological model of a landslide. The classication helps to create an overview of the determining hydrogeological characteristics and potential triggering mechanisms of a landslide. This can yield the base for numerical modelling and the planning of mitigation measures.

2. Comparing dierent parts of a landslide. For example on the La Frasse and the Pont Bourquin landslide, it can be seen that the slope-upward part and the middle part, respec-tively, are more important for the triggering and the slope downward parts have rather stabilising eects.

3. Discussing the saturation before and after a rain or snowmelt event. It should be clear whether the saturation after a usual event or after an extraordinary event is analysed.

It is possible to evaluate rain events with dierent return periods. Furthermore, due to variations of the saturation, a slope may change the position in the classication in time.

Nevertheless, one saturation-permeability contrast type is determinant for the moment of the triggering.

4. Comparing the hydrogeology of dierent landslides. For example in Rüdlingen, the

classi-cation may be used to compare the natural triggering mechanisms occurring in the region with the triggering mechanisms observed during the experiment. An other example are the two landslides La Frasse and Triesenberg, which are both deep seated and slow moving.

The La Frasse landslide is more active in the upper part where bedrock exltration and overpressure are important processes. The Triesenberg landslide is more active in the lower part which is underlain by low permeable bedrock. Thus, the application of the classica-tion helps to see the dierent mechanisms. The comparison of the Rüdlingen and the Wiler experiments shows that the permeability contrast class is the same. But the saturation was dierent and thus the potential triggering mechanisms and the criticality were dierent.

5. Estimating the criticality of a slope or an entire region to shallow, medium and deep seated landslides.

The hydrogeological classication can be applied with dierent amount of details. This means that depending on the available knowledge about the hydrogeology and geology of a slope, the parameters permeability, saturation and plasticity as well as the number and thickness of the geological layers have to be estimated to create a basic conceptual model. In such a case, large uncertainty may exist. If a more precise conceptual model is needed, for example for com-plex landslides, several investigations may be necessary to evaluate the parameters of interest.

This allows to apply the hydrogeological classication in more detail, for example to distinguish dierent parts of a landslide. The permeability-saturation type rstly dened (before detailed