Revisiting the Hydrograph Separation Issue Using High-Frequency Chemical Information

Hydrograph separation procedures aim to separate the baseflow and quickflow components of streamflow. Although widely used, these procedures are considered arbitrary, as they are mostly based on sheer geometrical or mathematical concepts and involve only streamflow data. Because the in-stream chemical concentrations are known to be valuable indicators of the actual flow-generating processes, we propose here a calibration approach for one of the most common hydrograph separation methods: the Recursive Digital Filter (RDF). The proposed approach uses continuous in-stream chloride concentrations, measured at the ORACLE-Orgeval Observatory (France). For this, we combine the RDF method with the chloride solute through the mass balance method (MB) and a multi-objective optimization function. Results show that the calibration approach allows a better estimation of the chemical concentration time series tested at flood event and inter-annual scale.