| Summary: | Groundwater modelling can be applied within many fields, such as aid for geotechnical examinations and contaminant transport. In many cases however, groundwater modelling is not used due to the need of large data quantities. A sensitivity analysis has been conducted for hydraulic conductivity, groundwater recharge and data resolution in time and space, to examine what parameters affects the result most. A reference case was calibrated to form the basis of the analysis. The reference case was formed by a ground model scanned with laser from a plane, probes for the level dividing friction soil and clay, a base model composed by interpretations and probes for base level and 19 groundwater pipes. These data were then scaled down where geological and hydrological data were changed in different experiments. It was clear that the amount of information points was not as important as the placement of them. For both types of data it was important to spread out probe points and make sure that peaks in the topography were included. The results generally showed that recharge areas on a higher altitude are the most important recharge areas. These areas have no other water supply unless further boundary conditions apply. A change in the hydraulic conductivity in the friction earth gave a larger difference in model results than when an equally large change in hydraulic conductivity was applied in the clay layer. The largest difference in the model result occurred when the same change was applied for both layers at the same time. The reason the hydraulic conductivity in the friction soil layer matters so much is because it is the layer through which water travels. A change in the clay ’s hydraulic conductivity does not pose an obstacle in the same way it does in the friction soil. A change in the hydraulic conductivity gave a smaller change in model results than an equal percentage change in groundwater recharge. Since higher uncertainties are associated with hydraulic conductivity an extended sensitivity analysis was performed for this parameter. This analysis showed that a change in hydraulic conductivity gives rise to larger differences in model results than for a change in groundwater recharge within their reasonable uncertainty ranges.
|
|---|
