Restoring river ecology — A complex challenge

All over Europe, rivers have been regulated during this century in order to drain and free agricultural land and to provide flood protection. In the process, 95% of the natural wetlands along European rivers have been detached from the rivers and unique ecological systems have been destroyed. The river regulations have many places been far from successful, leading to increased risk of embankment breaching and environmental deterioration. The present trend is that an increasing number of these rivers are restored by moving the embankments and turning them back into a meandering course. The development of such ‘ecological’ rivers is a highly interdisciplinary task, which requires complex tools describing the interaction between hydraulics, sediment transport and morphology, water quality, habitat and ecology and the interaction between surface water and ground water. The paper describes the necessary elements of such study and the implementation of 2 ambitious river restoration projects in Denmark. The restoration of Skjern River is the largest project of its kind in Northern Europe. The restoration project is unique in its size, in the public awareness around it and in the technical difficulties it has to deal with. The key objectives have been to return the river to its original meandering course, to create permanent wetlands, to increase the biodiversity in the river valley, to improve the conditions for trout and salmon, and to increase the nutrient turnover of the system. A key requirement for the restoration project was that no decrease in flood protection standard would be acceptable outside the project area. The paper describes the detailed mathematical modelling study carried out as part of the flood risk, migration conditions for salmon and trout, river morphology, saline intrusion and retention of nutrients and marshes. The second described project concerns the restoration of a section of the Brons creek in the southern part of Jutland, Denmark. The section is located in the lower part of a 100 km2 river catchment mainly covered with farmland and plantation. The project objective is to increase the potential for NO3− removal in the wetland areas along the creek and in the creek itself. Measures include restoration of the old creek course, water level rise in the creek by ‘semi-natural’ submerged weirs, and cutting off all artificial drainage pipes and ditches in the catchment. Mathematical modelling tools have been applied to demonstrate the possible effect of the restoration project. The modelling approach comprises integrated hydrological modelling with MIKE SHE, river modelling with MIKE 11 in combination with agro-chemical modelling based on DAISY. MIKE SHE simulates the flow and transport in different components of the hydrological cycle and the coupled MIKE SHE/MIKE 11 approach is able to simulate the dynamic interaction between water table in the creek and the groundwater table as well as soil water content in the areas along the creek. On the local scale it is possible to obtain detailed information on land-use for each field in the catchment for the past five years both with respect to crop type and applied fertilizers. This information forms the main input to DAISY, which runs as a sub-routine to MIKE SHE simulating both infiltration of water and nitrogen as well as its consumption and degradation. MIKE SHE takes care of transport and degradation of https://www.w3.org/1998/Math/MathML" display="inline">NO3− in the unsaturated zone and in the groundwater. An overall model covering the entire catchment was used to generate time-varying boundary conditions to the local scale model in terms of water and https://www.w3.org/1998/Math/MathML" display="inline">NO3− fluxes and groundwater head variations. In lack of field data of detailed land-use and fertiliser application, a statistical approach was applied to allow DAISY to generate infiltration of water and nitrogen to MIKE SHE. Firstly, the area was divided into main groups — farmland, plantation, urban areas and areas with permanent grass. Secondly, crops and different farming practices with respect to NO3 application were distributed randomly over the farmland according to available statistical information. A simple approach for NO3− degradation in the groundwater zone was introduced and MIKE SHE was used to generate the boundary conditions for local scale modelling. The paper describes the results of the modelling and its implication on the restoration design.