Résumé
In this work results of hydraulic borehole tests on different spatial scales are attributed to the geometrical properties of karst aquifers. The results are important for parameterization of numerical models dealing with karst hydraulics. Karst aquifers can be described as dual flow systems consisting of low permeability fissured matrix blocks and highly permeable conduit systems. Over a large volume the fissured matrix responds as a continuum to hydraulic events and a representative elementary volume (REV) can be defined. However, this REV is only valid on local scale. On regional scale the drainage of the karst aquifer is controlled by the conduit system which introduces a highly anisotropic geometry into a karst system. Conventional methods for characterisation of aquifer properties are hydraulic borehole tests. Slug-tests, for example, can be applied in deep small diameter boreholes as is often the case in karst systems with thick unsaturated zones. Because of the small volume of integration, test results depend on the location of the investigated borehole and the applied displacement depth. The spectrum of responses may range from oscillating water levels in high conductivity parts of the aquifer to very slow responding water levels in the porous matrix. Estimated hydraulic conductivities vary many orders of magnitude and the representative elementary volume of the fissured matrix is only difficult to estimate. Larger aquifer volumes surrounding a borehole can be characterized applying conventional pumping tests. Pumping tests with pumping rates higher than spring discharge allow a regional drawdown of the water level in karst aquifers. The high pumping rates require that the pumping well is directly connected to the conduit system. The drawdown of the water level at the beginning of the pumping test is controlled by the hydraulic and geometric properties of the conduit system. However, with increasing duration of pumping the drainage of the fissured matrix will occur. The pumping tests can be analyzed with modeling approaches which treat the fissured matrix as a continuum and implement the conduit system as a discrete network or second continuum. Large scale pumping tests therefore allow the estimation of hydraulic parameters of the conduit system and of parameters representing the REV of the fissured matrix.