Context and objectives
Karst collapses are important natural hazards for roads, houses, industries and water tables. More than 150 sink holes had been listed until today in the Tournaisis. Generally, they are cylindrical collapses with a maximum diameter of twenty meters and more than ten meters depth. The occurrence of those holes seems to be fortuitous. However, prone karst collapse areas are spatially well correlated with the location of joints in the Tournaisian limestone. It does not exist any signs to foresee the danger. Indeed, the quaternary deposits hide completely the inner soil movements which will become sink holes.
The objective of this feasibility study consists in an attempt of demonstration that it is possible to explain some parts of the topsoil moisture distribution extracted from ERS SAR data by the natural fracturing of the Tournaisian limestone. If the correlation exist, it will be possible to derive hazards and risks maps for decision planner makers. Indeed, we need to solve a regional problem at reasonable cost. Microwave remote sensing is of interest for that scale. However, due to the small size of collapses, we need to find indirect information detectable with satellite radiometers and obviously related to our problem. We assume that topsoil moisture is the data which link up the remote sensing technique and the karst collapses.
Expected scientific results
After discussions with the members of the team, we think that the basic ideas are relatively well-founded for the parts where tertiary and quaternary deposits are relatively thin. Some problems had been encountered with ERS data. We think that the methodology should given accurate results with airborne SAR. Indeed, in that way, some parameters could be more controlled such as the date of the fly in order 1) to free from the bad weather conditions and the roughness of the surface (control the size of the vegetation over the test area); 2) to reduce the size of the pixels; 3) to increase the performance in the georeferencing step.
The cartography of the moisture depends on the roughness parameter. After the bad experience of that feasibility, it seems to be clear that a strong knowledge of the efficiency in the methodology of acquisition is the first factor to master. The prone karst collapse area detection is really worrying in the Tournaisis part because of increasing of pumping per year. At present, there is no state control about the relation between the occurrence of sink holes and the pumping of the water table. Only few scientists say about the problem or media when collapses had damaged infrastructures. The high complexity to forecast such a phenomenon should imply more research than presently.
|ULg - Laboratoire de Géomorphologie et Télédétection