Steady-state semi-analytical solutions for assessing the two-dimensional groundwater hydraulic head depletion induced by river dam removal

• Analytical solutions for assessing the GW hydraulic head change due to dam removal. • River-groundwater and dam reservoir – groundwater exchanges. • Method of fundamental solutions. • Robin condition. • Satisfactory comparison with field data. Dams are pervasive features of the world’s river systems. The reservoir associated with the dam modifies the distribution of the hydraulic head in the aquifer and the natural groundwater flow. However, such modifications must be defined to forecast their environmental, economic and/or social impacts. Based on the method of fundamental solutions (MFS), steady-state semi-analytical solutions are proposed for evaluating the long-term spatial distribution of the rise or decline of the hydraulic head in an aquifer caused by a dam reservoir on a watercourse or its removal. MFS was chosen because of its ability to account for different boundary conditions along the banks of the watercourse. The dam reservoir can have a triangular or rectilinear geometry. The aquifer is assumed unconfined, homogeneous, inclined and finite. The solutions take into account both the reservoir-aquifer interaction and those between the stream and the aquifer up- and downstream of the reservoir. The reservoir-stream system fully penetrates the aquifer, and is separated from the aquifer by semi-pervious banks (Robin condition). Analysis of the semi-analytical solution included sensitivity tests, and assessment of the influence of several parameters: reservoir leakance parameter, width of the reservoir and width of the aquifer. The results show that the reservoir and river leakage parameters up- and downstream of the reservoir, as well as the geometry of the reservoir dam, are the main determining parameters of the extent of the induced groundwater mound. They show that low aquifer-river exchanges in the river upstream and downstream of the reservoir increases the spatial extension of the mound. The degree of error introduced by the method of linearization used to solve the governing partial differential equation is also discussed. A solution has been also developed with a pumping well, particularly to assess what the depletion of the water level at the well location might be if the dam is removed. The proposed solutions were applied at two field sites, giving satisfactory results. These semi-analytical solutions will be useful applications for assessing the long-term spatial impact on the aquifer of the emplacement, removal or leveling of a reservoir dam.