Geomorphological Controls on Stream Baseflow Recession

International audience; The understanding of the physical processes controlling water fluxes from hillslope aquifers to surface waters is critically needed for predicting stream discharge and establishing sustainable water resources management strategies. Two major hypotheses are currently being debated: the first considers that climate, stream network morphology and topography have a first-order control on groundwater flow and stream recession. Thus the prediction of stream baseflow can be approximated by considering idealized homogeneous groundwater reservoirs with effective properties. The second hypothesis suggests that the structural heterogeneity of the landscape (lithological variations, geological discontinuities, spatial distribution of topographic and tectonic stresses) might significantly control the partitioning of water within the landscape, and in turn impact how streamflow recession behaves.In this contribution we provide a comprehensive analysis of stream discharge and geomorphological databases across various geological domains distributed throughout the globe in order to identify the main factors that might influence recession constants at baseflow. A novel computational approach for single recession analysis revealed that recession constants might strongly deviate from what is predicted from idealized theories, with power law recession constants ranging from 1.5 to more than 10. A correlation analysis allows us to identify the control of key features of the landscape on groundwater storage and streamflow recession constants, with depth to bedrock, lithology and maturity - i.e. degree of weathering and fracturing - of the landscape being the main drivers. We tested these observations through a numerical modelling framework which allowed us to quantify the variability of streamflow recession behaviors raising from the drainage of hillslopes with different structural and geometrical properties. We identify the respective roles of hydraulic conductivity contrasts between the upper and lower compartments, the depth of the shallower compartment and the steepness of the slope on streamflow recession parameters. We finally discuss the implications of these findings in establishing future strategies for the improvement of groundwater storage and stream discharge predictions.