CAN PARTICLE TRANSPORT BE MODELLED? INDICATIONS FROM THE MILANDRE UNDERGROUND LABORATORY (JURA, SWITZERLAND)

The groundwater catchment area of Milandre cave stream is about 10 km2 large with a purely autogenic recharge. The cave was explored for nearly 5 km of stream passage.

Since 2003 the cave stream was monitored in order to assess the effect of a road construction located directly on top of the cave.

The monitoring included the following devices:

1) 4 stations along the cave stream, recording water turbidity every 15 minutes;

2) 4 stations equipped with sediment traps, i.e. trapping particles of the cave stream over 3 month periods.

3) 1 station with small horizontal plates collecting the deposited sediment observed every 3 to 6 months.

After 10 years of observations the following major aspects could be evidenced:

1) Water turbidity varied between 0.5 and 300 FTU, with an average value of 11.4 FTU.

2) Most turbidity peaks are related to discharge peaks. However, their relative intensity could not be correlated at all. Various models have been attempted to link turbidity intensity to discharge rate, but all remained unsuccessful.

3) Particles coming out of the system at the spring are rarely those eroded from the soil a few hours or days before. In most cases particles are deposited and eroded several times along their flow through the system.

4) The road construction included large areas where soil was removed. This did not significantly increased water turbidity: Natural variations remained larger than the possible increase related to the construction.

5) The type of particles changed along the last 10 years with an increase of clay particles and a decrease of Quartz and Feldspar. This change could be related to a reduction of the average discharge over the last few years.

6) The flux of particle was estimated to about 50 kg/day per km2 in average (nearly 180 tons per year, i.e. 0.045 mm of soil removal per year). This value may probably be underestimated, but it shows that soil erosion is not very high in this catchment.

Most particle deposition is observed in the epiphreatic zone of the aquifer, with rates of several mm per year. One part is removed by erosion along the cave stream, but another part is accumulated on passage sides. After several decades of accumulation the sediment mass slides down into the stream bed, producing a “cloud” of high turbidity at the spring. This may explain the strange relation between discharge rate and turbidity.