Procedimiento y análisis de datos

reached the maximum value of 2 l/min, it was decreasing during 3 - 4 days. The piston

ow arrived with a delay of 4 days after the ux had established a mean value of 0.7 l/min.

Spring 32 is perennial and shows a high mineralisation. Like in PB5, a dual porosity was observed.

The fraction of rainwater can roughly be calculated with a "mixing model" with equations 6.3 and 6.4.

ECrain = Electric conductivity of rain, ≈ 10µS/cm

EC_spring = Average electric conductivity of the spring water [µS/cm]

ECmix = Electric conductivity measured during precipitation

If we assume that the rainwater has a EC of 10 µS/cm, then it contributes with an average ux of 8 - 9 % to the total measured discharge. Table 6.8 shows the calculated values.

Year ECspring, µS/cm ECmix, µS/cm x1, fraction rainwater

2010 1300 1200 0.08

2011 1700 1550 0.09

Table 6.8: Table showing the mean values for the calculation of the fraction of rainwater in the spring water during precipitation events.

To summarise, it can be said that the water from Spring 32 has two origins: Water that is rapidly

owing in fractures close to the surface and deeper groundwater which is strongly mineralized and arrives with a delay of a few days after a rain event.

6.10 Hydrogeochemistry

Water samples were taken from observation wells, springs and ephemeral creeks on and close to the Pont Bourquin landslide from summer 2009 until spring 2011. Water temperature and electrical conductivity (EC) of the water were measured in the eld. For some samples, also the pH and the redox potential Eh were measured. The pH and Eh values shout be taken with caution because the sensors were not very stable. The concentration of major cations (Na⁺, K⁺, Mg²⁺ and Ca²⁺) and major anions (F⁻, Cl⁻, NO3−, SO42− and HCO3−) from each sample and additionally the elements Si and Sr from selected samples were analysed at Geolep, EPFL. δ³⁴S isotopes were analysed at Institute of Mineralogy and Geochemistry, University of Lausanne.

The results are listed in Tables B.2 and B.3 in Annexe B. The location of the samples are shown on Figure 6.29. For the description of the methods see Section 4.11 and 4.13.

6.10.1 Major ions

The balance between the anions and cations was 0.6% in average with a maximum of 3.6% at piezometer 1A. Thus, no analyses had to be rejected (see Section 4.11 for the method). The

6.10. HYDROGEOCHEMISTRY

Figure 6.29: Map showing the locations of water samples from creeks (C), observation wells and springs.

The symbols for the sampling locations are used in the scatter plots on the following gures. Sti diagrams are plotted for eleven water samples from 15. April 2011.

6.10. HYDROGEOCHEMISTRY

groundwater of the Pont Bourquin landslide is composed to 98.5 % of the cations Mg²⁺ and Ca²⁺ and the anions SO42− and HCO3−. Figure 6.29 shows Sti diagrams with these four ions for eleven water samples taken on 15. April 2011. In Figure 6.30.A, the cations Mg²⁺ and Ca²⁺

are plotted versus the anions SO42− and HCO3−.

Highest Ca²⁺ and Mg²⁺ concentrations of 25 meq/l and 10 meq/l, respectively, were measured in PB5 whereas in the water samples from the moraine the Ca²⁺ and Mg²⁺ concentrations were less than 6 and 1 meq/l, respectively. The mean ratio Ca²⁺:Mg²⁺ in all samples is about 3:1 (see Figure 6.30.B).

Figure 6.30: Scatter plot showing the principle cations Mg²⁺ and Ca²⁺ versus the principle anions SO42− and HCO3−.

The concentration of SO42− ranges from 1.5 to 30 meq/l. A few samples from the moraine in the upper part of the landslide (1A, SpA, CreekR, Ch2 and 1C) have not really been in contact with SO42−. The SO42− concentration is highest in PB5 and the springs in the lower part of the landslide and in spring 45L below the upper secondary scarp. HCO3− ranges from 0.5 to 7 meq/l with a mean value of 3.9 meq/l. The samples from the moraine at the top and to the East of the landslide (1A, SpA, CreekR) show the lowest and the springs in the lower part of the landslide (Spring 32, Spring 25 and Spring 18), observation well PB5 and Spring 49 in the upper part of the landslide show the highest HCO3− concentrations.

Temporal variations

On Figure 6.30 it can be seen that there is a large temporal variation in the water chemistry especially for the Spring 32 and observation well PB5. The dates of the samples are shown on Figure 6.31. The groundwater in observation well PB5 (Figure 6.31.A) is higher mineralised dur-ing the winter months. For sprdur-ing 32 no seasonal trend is observed (Figure 6.31.B). In (Figure

6.10. HYDROGEOCHEMISTRY

6.31.C) the temperature measured for each sample is plotted versus the conductivity values. The temperature variation is larger in the piezometers in the moraine ( 2 - 17°C) than in observation well PB5 and spring 32 (5- 13°C). The variation in mineralisation is smaller in the samples from the upper part of the landslide than for PB5 and Spring 32.

The temporal variations at single locations may reect long-term seasonal changes or short-term dissolution during and after rain events.

Figure 6.31: A and B: Temporal variations of the concentrations of Mg²⁺, Ca²⁺, SO42− and HCO3−

in observation well PB5 and Spring 32. C: Temperature plotted against the electric water conductivity of the samples illustrates temporal variations.

Origin of the major elements

Mg²⁺, Ca²⁺, SO42− and HCO3− are typical major ions in water that was in contact with evap-oritic and carbonate rocks. This can be seen in Figure 6.30.A because the slope is 1:1. Ca²⁺

and Mg²⁺ comes most likely to large parts from the dissolution of calcium carbonate CaCO3, dolomite CaMg[CO3]2, and evaporites gypsum Ca[SO4]2H2O, anhydrite Ca(CO3) and magne-sium sulphate Mg(SO4). The HCO3− comes most likely to large parts from calcium carbonate

6.10. HYDROGEOCHEMISTRY

CaCO3 and dolomite CaMg(CO3)2 that reacted with H2CO3. The SO42− could originate from evaporites gypsum Ca[SO4]2H2O, anhydrite Ca(SO4) and magnesium sulphate Mg(SO4) associ-ated with the cellular dolomite. Furthermore it could originate from pyrite FeS2 present in the black shale. The concentration of Sr²⁺ and the δ³⁴S can give important information about the origin of SO42−.

6.10.2 Strontium

Strontium analyses were carried out in order to test the evaporitic impact upon the water chem-istry. High strontium concentration is explained by the dissolution of celestite Sr(SO4) which is associated with evaporites. The highest Sr²⁺ concentration was measured in PB5 (1.5 g/l) and springs 32 and 25 (0.8 mg/l) (see Figure 6.32).

These are rather low values compared to evaporitic springs in the Western Swiss Alps, where strontium concentrations between 1 and 15 mg/l (mean 7.9 mg/l) were described (Mandia, 1991).

The ratio Sr²⁺/Ca²⁺ in PB5 and Spring32 is 0.0014 and 0.0011, respectively. A ratio Sr²⁺/Ca²⁺

> 0.001 (expressed in meq/l) is typical for evaporitic water (De Montety et al., 2007). In Super-Sauze landslide, a ratio between 0.0012 and 0.0078 meq/l has been measured (De Montety et al., 2007). Thus the evaporitic origin of SO42− in the groundwater of the Pont Bourquin landslide is not very pronounced compared with other locations.

Figure 6.32: Strontium versus Calcium concentration.

6.10.3 Sulphur isotopes

Sulphur isotopes from six groundwater samples were analysed to obtain information about the origin of SO42− from the dissolution of evaporites or from pyrite. The results are shown on Figure 6.33 and in Table B.3 in Annexe B. The measured δ³⁴SV CT D values range between 3.8 and 10.8 ‡. The lowest value is observed in Spring 49R in the upper right part of the landslide.

This spring originates in black shale (see Figure 6.10, middle right). Generally, the δ³⁴SV CT D

values are increasing towards the toe of the landslide.

This δ³⁴SV CT D values are interpreted as a mixture of SO42− from the dissolution of pyrite and evaporites. Typical δ³⁴SV CT D values for pyrite and Triassic gypsum/anhydrite are -25 to 2.5 and 15 to 17 ‡, respectively (see Section 4.13). The lower values observed in the upper part of the landslide indicate a larger contribution from pyrite dissolution. This led to the assumption that