A MODFLOW model to understand the hydrological characteristics of the Silberleite catchment to eventually correct super conducting gravimeter measurements.
Understanding hydrology around Moxa Observatory to correct superconducting gravimeter measurements.
Since 1999, a superconducting gravimeter records earth gravity with high accuracy at the Moxa Observatory near Jena, Germany. This gravity signal is disturbed by changes in local water movement. These relatively small water masses are large enough to have a significant effect on gravity measurements. In order to correct for these mass movements, the hydrology of the Silberleite catchment should be understood.
The dominant rock type in the catchment is highly consolidated shale with fractures. The top 2 m consists of (highly) weathered shale material with a clear orientation to the south-east and a high hydraulic conductivity. Furthermore there are strong indications of a fault dissecting the study area in a eastern and western part. This fault offers a potential preferential flow through the shale formations, and can probably be considered as a horst-graben system.
A new MODFLOW model based on a MODFLOW model developed by Bergsma (2012) has been constructed to better understand the hydrological behaviour of the Silberleite catchment. In this model interception and evapotranspiration by Scots pine (Pinus Sylvestris) was implemented. For loss due to interception 22% of the precipitation was subtracted from the total precipitation that falls on a forested area. The potential evapotranspiration was calculated using Penman-Monteith with aerodynamic and bulk surface resistances related to coniferous forest. Lastly to have the amount of model layers correspond to the amount of layers observed in the field, two extra model layers have been added to the slopes of the Silberleite valley. To better understand the influence of the hydrological properties a sensitivity analysis was performed, which indicated that the hydraulic conductivity of the deep, consolidated shale layer (occasionally with fractures) was the most influential in the model. This is the result of this layer being the thickest (60 m) and thus a large fraction of the total amount of water can be stored in this layer if the storage capacity is sufficient. Another reason would be that this layer has lowest hydraulic conductivity relative to the other model layers, but since MODFLOW uses a harmonic mean of the hydraulic conductivity when calculating flow between cells, the cell with the lowest hydraulic conductivity is decisive in this calculation. This results in the model being the most sensitive to the layer with the lowest hydraulic conductivity.
Model results were promising. A shallow piezometer at 2 m depth captures the hydrological behaviour of the catchment well. The response time of the model is roughly equal to the measured values, however, the model underestimates the change in head as a result of groundwater recharge by about 0.5 m. The modelled values of the other shallow piezometer is not promising as it shows a very different response to groundwater recharge, even though these two piezometers are located roughly 2 m apart and measure groundwater level at the same depth. A deep piezometer (50 m) captured the response to groundwater fluctuation quite well, but should have a shorter response time as the lag between measured and modelled is about 20 h.
Bergsma (2012) concluded that flow in the Silberleite catchment was surprisingly in the direction of the end of the valley rather than to the valley bottom, following the topography of the (steep) valley slopes. The model results of this study without a fault implemented showed the same results in which water flowed through fractures of the deep consolidated shale material towards the end of the valley and not towards the valley bottom. The model results with a fault implemented following the valley bottom showed a flow direction more towards and following the valley bottom compared to the model results without a fault. This was expected as the hydraulic conductivity in the fault is much higher compared to the surrounding consolidated shale material as a result of the fractured material surrounding the fault, offering a preferential flow path in the direction of the fault. Although modelling results are promising, thorough calibration is required to use to model for filtering of the gravity recordings at Moxa Observatory.