Regional flow in a complex coastal aquifer system: combining voxel geological modelling with regularized calibration

Journal of Hydrology, Elsevier, ISSN 1879-2707

Volume 562, 2018

DOI:10.1016/j.jhydrol.2018.05.020, Dimensions: pub.1103886214,



  1. (1) University of Copenhagen, grid.5254.6, KU
  2. (2) Geological Survey of Denmark and Greenland (GEUS), Denmark
  3. (3) University of Cagliari, grid.7763.5







Low-lying coastal regions are often highly populated, constitute sensitive habitats and are at the same time exposed to challenging hydrological environments due to surface flooding from storm events and saltwater intrusion, which both may affect drinking water supply from shallow and deeper aquifers. Near the Wadden Sea at the border of Southern Denmark and Northern Germany, the hydraulic system (connecting groundwater, river water, and the sea) was altered over centuries (until the 19th century) by e.g. the construction of dikes and drains to prevent flooding and allow agricultural use. Today, massive saltwater intrusions extend up to 20km inland. In order to understand the regional flow, a methodological approach was developed that combined: (1) a highly-resolved voxel geological model, (2) an ∼1 million node groundwater model with 46 hydrofacies coupled to rivers, drains and the sea, (3) Tikhonov regularization calibration using hydraulic heads and average stream discharges as targets and (4) parameter uncertainty analysis. It is relatively new to use voxel models for constructing geological models that often have been simplified to stacked, pseudo-3D layer geology. The study is therefore one of the first to combine a voxel geological model with state-of-the-art flow calibration techniques. The results show that voxel geological modelling, where lithofacies information are transferred to each volumetric element, is a useful method to preserve 3D geological heterogeneity on a local scale, which is important when distinct geological features such as buried valleys are abundant. Furthermore, it is demonstrated that simpler geological models and simpler calibration methods do not perform as well. The proposed approach is applicable to many other systems, because it combines advanced and flexible geological modelling and flow calibration techniques. This has led to new insights in the regional flow patterns and especially about water cycling in the marsh area near the coast based on the ability to define six predictive scenarios from the linear analysis of parameter uncertainty. The results show that the coastal system near the Danish-German border is mainly controlled by flow in the two aquifers separated by a thick clay layer, and several deep high-permeable buried valleys that connect the sea with the interior and the two aquifers. The drained marsh area acts like a huge regional sink limiting submarine groundwater discharge. With respect to water balance, the greatest sensitivity to parameter uncertainty was observed in the drained marsh area, where some scenarios showed increased flow of sea water into the interior and increased drainage. We speculate that the massive salt water intrusion may be caused by a combination of the preferential pathways provided by the buried valleys, the marsh drainage and relatively high hydraulic conductivities in the two main aquifers as described by one of the scenarios. This is currently under investigation by using a salt water transport model.

Research Categories

Fields of Research

Sustainable Development Goals

Links & Metrics

NORA University Profiles

University of Copenhagen

Danish Open Access Indicator

2018: Blocked

Research area: Science & Technology

Danish Bibliometrics Indicator

2018: Level 2

Research area: Science & Technology

Dimensions Citation Indicators

Times Cited: 5

Field Citation Ratio (FCR): 1.72