Are hydro-graphically assessed streamflow components hydro-chemically meaningful?

In recent years, extreme weather events such as intense rainfall and prolonged droughts are occurring with increasing frequency all over Central Europe. To deal with the hydrological consequences an improved understanding of storage and flow dynamics within hydrological catchments might be essential. Hence, the objective of this study is to develop and test a parsimonious work-routine to identify dominant streamflow components and dynamic catchment storages in complex hydrogeological settings.

Based on a two-year field campaign at public streamflow gauging stations in western Germany, we investigated water quality dynamics in two nested catchments with long-term hydrological records. We collected daily water samples and analyzed dissolved organic carbon, nitrate, electrical conductivity, silica, and stable water isotopes. Using these data, we evaluated a hydrographical filter algorithm and a subsequent linear storage detection method, focusing on their hydro-chemical interpretability. We then applied the resulting workflow to published datasets from several nested catchments in Switzerland and the United States. The dynamic hydrographical filter algorithm, DelayedFlowIndex (DFI), is derived from the classical Baseflow Index (IH-UK). It produces Characteristic Delay Curves (CDCs) that describe average catchment drainage behavior with filter widths from 0 to 60 days after a streamflow increase. We improved an existing workflow that divides CDCs into several linearly draining subsets. The new automated routine determines the catchment-specific number of sub-storages. Testing this approach on more than 100 catchments showed that both CDCs and the number of sub-storages can be linked to distinct morphological catchment characteristics. To increase confidence in the hydrological interpretability of hydrographically derived streamflow components, we compared stream hydro-chemical information from our field sites and from published datasets with the resulting flow components. We also examined their roles in streamflow composition throughout the year. At our test sites, rapid flow components showed elevated DOC concentrations. Intermediate components displayed pronounced nitrate peaks. Delayed components had increased silica concentrations, while highly delayed components were associated with higher electrical conductivity. In the downstream sub-catchment, these hydro-chemical signals were additionally shaped by seasonal effects. Water samples collected during a 45-day drought in spring 2025 at both gauges provided valuable information for validating the hydro-chemical signatures of very slow storage components, which are rarely observable.

The streamflow components derived from the DFI show consistent correlations with distinct solute signatures, demonstrating that they are hydro-chemically meaningful. Consequently, DFI analysis combined with the automated storage-delineation algorithm provides a robust, streamflow-based method for identifying catchment-specific flow components. This approach offers valuable insight into storage depletion processes and groundwater dynamics, particularly under extreme low-flow conditions in (mid-)mountain regions.