Hydrogeological Model of Northern Switzerland

  pdf NAB 24-16 Hydrogeological Model of Northern Switzerland(38.45 MB)

A regional steady-state hydrodynamic model was developed to examine the groundwater flow system in Northern Switzerland, focusing on aquifers above and below the Opalinus Clay and the influence of regional-scale faults on groundwater flow. The model geometry was based on existing geological models and extended with hydrogeologically significant layers. Except for the Jura Main Thrust, which was modelled in more detail, faults have been implemented as vertical zones with a constant width of 150 m. Model boundaries were determined using hydraulic data and conceptual considerations and a sensitivity study was performed to test various boundary conditions to determine the best-fit model. The best-fit model was selected based on performance metrics such as mean absolute error and correlation coefficients between modelled and observed head values. Using the best-fit model as the base, a careful calibration of the model parameters was performed using the PEST algorithm. This calibrated model is the base case. The sensitivity of the model was also explored with additional cases.

The base case is marked by a piezometric depression located in the area of the lower Aare Valley and Hochrhein Valley in the region of Kaiserstuhl – Hohentengen (close to the Nördlich Lägern siting region, NL), separating the hydraulic flow system in the northern part of the model domain into an eastern (including the Zürich Nordost siting region, ZNO) and western part (including the Jura Ost siting region, JO). This depression can be observed in all aquifers and is the result of the topography, the surface waters and the location of outcrops. Generally, flow occurs from topo­graphically higher outcrops in the west or northeast towards the piezometric depression.

For the Malm aquifer in the eastern part (ZNO), recharge areas are mainly located along the topo­graphically higher outcrops in the north, discharging into the Rhine Valley. In the western part of the model, flow is directed towards the Aare Valley and other, smaller river valleys. The area between these two (including the NL siting region) is characterised by low hydraulic heads where recharge of the Malm mainly occurs through the thick Molasse units along the northern part of the Lägern Anticline and discharge along the Rhine River in the area of Kaiserstuhl – Hohen­tengen.

The Hauptrogenstein aquifer, only present in the western part of the model, shows recharge areas in the topographical elevation zones (areas of Frickberg and Internal Jura (Folded Jura) to the south of JO) and discharges into the river valleys (lower Aare Valley and Sissle Valley).

The local Keuper aquifer was implemented in the model with different zonations of the hydraulic properties (based on geological and hydrogeological observations). For the NL siting region, the calibration results suggest lower hydraulic conductivities than in ZNO and JO. Recharge areas for the Keuper aquifer are located between the Wutach and the Klettgau Valleys north of the NL and ZNO siting regions, and locally in outcrops north of the JO siting region. Comparable to the Muschelkalk aquifer below the Keuper aquifer, discharge areas for the Keuper aquifer in JO are mainly located in the lower Aare Valley and the Sissle Valley. From the NL and ZNO siting regions, flow is directed southwards. In NL, flow velocities are particularly low, reflecting the low hydraulic conductivities and gradients.

Finally, the Muschelkalk aquifer represents the lowest aquifer implemented in the model. In the western part, recharge is mainly located in topographically higher outcrops to the north and to the south of JO, while flow is directed towards the Aare and Sissle Valleys. In the eastern part, a large-scale flow system occurs extending from the recharge area along the Wutach Valley, across the ZNO and NL siting regions towards the regional discharge area located at the confluence of the Aare and Rhine Rivers.

A set of additional cases addressed the local Keuper aquifer. Lithological and hydrogeological data show that the nature of this aquifer varies between the siting regions with marked variations in hydraulic conductivity. These differences and the related uncertainties were accounted for by different parametrisations: isotropic and variants of anisotropic K_xx/K_yy. The anisotropic cases are based on the observation that in parts of the Keuper aquifer, sand-rich channels seem to be mainly responsible for groundwater flow. To account for the resulting flow direction depending on the orientation of the channels (which are not known), two additional cases, with preferred E-W- or N-S-directed flow, have been simulated. These additional cases show that the details of flow direction and gradients locally change depending on the parametrisation, but the flow field remains stable with respect to the recharge and discharge areas. Generally, the fit of these addi­tional cases to observations is not as good as in the base case.

In the base case, the Malm aquifer consists of the «Felsenkalke» and «Massenkalk», the Schwarz­bach Formation and the Villigen Formation. As an additional case, only the «Felsenkalke» and «Massenkalk» were assigned with a higher transmissivity, however, this has little to no effect on the regional recharge/discharge areas compared to the base case.

The effects of regional-scale faults on the flow system were assessed performing sensitivity analyses. Using various permeability scenarios, conceptually accounting for the orientation of the faults with respect to the current stress field, show that the Baden – Irchel – Herdern Lineament and the Jura Main Thrust (roughly E-W-trending faults) effectively separate the model into a southern and northern hydrogeological system. Roughly N-S-trending faults (especially the Randen Fault) may locally channel flow from the recharge areas in the northern part of the model to the southeast. However, the resulting simulated head distributions of the sensitivity analyses show that faults appear to have only minor and local effects on the regional flow patterns.

Overall, the hydraulic situation in the model area is very robust, changes of parametrisation show only localised impacts. The comparison of simulation results with observation data supports the accuracy of the hydraulic simulations and shows that the model boundaries do not impede the reliability of the model.