Karstologie et analyse des systèmes aquifères
Conseiller du Ministre de la région wallonne pour l'Eau, l'Environnement et la Vie rurale.
Abstract
Karstology and hydrogeological systems analysis.
During the last decades enormous progresses were realized in groundwater models and aquifer systems analysis. The research on groundwater results in a large body of knowledge among which the mathematical tools of hydrogeological modelisation. These tools were able to give the refined quantitative answers needed for the utilisation and protection of groundwater resources. Obviously the use of these models was more successful in pore aquifers and more difficult in carbonate rocks.
Nevertheless it was shown that the concepts of coefficients of permeability and coefficients of transmissivity are acceptable in carbonate rocks when working at the right scale.
This article analyses why there is still an opposition between some "naturalists" who reject the mathematical tools and some "users" who apply them sometimes without enough knowledge of the natural karst system.
By the analyse of case histories in limestones and chalkrocks the requirements for application of modelisation tools to those aquifers are described. Among the requirements the need for field methods for determination of dispersion coefficients is emphasized.
The case histories, choosen as examples of the need of a detailed system analysis before modelisation, are the induced infiltration of the Scheldt in the Carboniferous Limestone and the radial dispersion model of the chalk cretaceous aquifer at Havré.
The analogy between global hydrographic models and karst aquifer models is pointed out. The last example, the study of the protection area of a water production gallery, illustrates the use of simple physical models in karst aquifers to simulate fissuration and karst channels.
In conclusion the progress in karst aquifers simulation will be highly stimulated by the research of widely diversified groups of scientists in which the system descriptions aspects will be confrontated with the mathematical tools. The resulting models must be as close as possible to the system itself and based essentially on answers of the real system to field tests.