Assessment of hydraulic conductivity from the hydrographic network in shallow crystalline aquifers

Hydrological predictions for ungauged basins at catchment and regional scales still faces the challenge of lack of available data. To meet this challenge, we propose a new method relying on the structure of the stream network. Under the assumption that the perennial stream network is mostly fed by groundwaters, its structure derives from the underlying aquifer properties. It is especially the case for shallow crystalline aquifers under temperate climates where the surface and subsurface hydrological systems are directly connected. The groundwater table remains close to the topography and the spatial extent of the stream network is then controlled by the magnitude of the subsurface hydraulic conductivity (K) with respect to the actual recharge rates (R).

 

Using a parsimonious 3D groundwater flow model, we propose a novel performance criterion to assess the similarity between the modelled seepage areas and the observed stream network. We investigate the sensitivity of our methodology to different digital elevation models (DEM) and stream network products from different databases that may impact the estimates through their different spatial resolutions. We use this method to determine the equivalent hydraulic conductivity for 25 crystalline catchments in western France.

 

The results show that our methodology allows predicting the spatial patterns of the stream network with a high sensitivity to the hydraulic conductivity. We found that estimated hydraulic conductivities vary over two orders of magnitude [10^-5 to 10^-4 m/s] across the 25 investigated catchments and are well correlated to the lithology. While the DEM resolution has no major effect on the results, we found that the proportion of described low-order streams significantly controls the estimations.

 

The proposed approach constitutes a paradigm shift in current methodologies designed to assess catchment-scale hydraulic properties with great perspectives regarding the emergence of remote sensing techniques for the mapping of wetlands and soil moisture. Our method might bring up new opportunities to provide predictions for ungauged basins such as the hydrographic network dynamics in a changing climate.