The main goal of a joint project undertaken by the geophysical and hydrolog
ical research units of Kiel University is to study preferential flow in a l
arge open-air, full-scale model, looking in particular at near-surface pene
tration and flow of water through the unsaturated vadose zone. An artificia
l irrigation device is installed in place of natural rain, and a homogeneou
s sand body is used instead of natural soil. This provides a reference mode
l for future field experiments. Inside the sand body there are a large numb
er of geophysical and hydrological sensors to measure DC resistivity (using
various electrode configurations), water content and water potential (usin
g TDR and tensiometer instruments, respectively). A ground-penetrating rada
r (GPR) system is installed at the surface, whereas at the bottom several c
ontainers and a thin gravel layer are embedded to measure the flow arrival
and the discharge of water. Irrigation is varied in intensity, time, area,
and salt content (tracer).
Results of the first six experiments show that the percolation of intruding
water can be followed by all techniques and percolation is finally control
led by the discharge measurements. These display some undulations and varia
tions of the water 'front' and agree with the measurements of all other sen
sors. The redundancy achieved by the use of multiple methods was intended t
o enable an assessment of the reliability of the techniques used. The true
values of electrical resistivity before and after irrigation reflect the di
stribution of water saturation within the sand body. A numerical 3-D invers
ion of the apparent resistivity provides information regarding future field
experiments, in which it will be possible to install only some of the sens
ors in order to preserve the natural structure of the soil.