Physiographic controls on fractions of new water in 12 nested catchments

In the context of global change, the characterization and quantification of the changing pulse of rivers is a pressing challenge. Over the past decades, rapidly increasing computational capabilities and the related complexity of numerical models have contributed significantly to improve flood forecasting systems. However, our understanding of the mechanistic causality especially of extreme hydrological events remains fragmented. Streamflow responses are notoriously threshold-bound and site-specific, thus making extrapolations to ungauged basins and projections into future climate scenarios difficult without physical evidence. There is thus still a need for inter-catchment studies across contrasted physiographic and climate settings, ideally spanning over large observation time intervals. Here, we rely on a 13 year-long, fortnightly resolved precipitation and stream water 18O isotope record from 12 nested catchments with different bedrock geologies (marls, sandstone, schists) and land cover in the Alzette River basin (Luxembourg). Located on the eastern edge of the sedimentary Paris Basin, our study area has a rather homogeneous semi-oceanic climate. The 18O records varied between catchments exhibiting both seasonal and interannual patterns during the 13 years of observations. The seasonal amplitude of the precipitation 18O signal was strongly damped in stream water of catchments dominated by permeable bedrock geology and large storage volumes. This dampening effect was much less pronounced in catchments dominated by marly (and thus less permeable) bedrock with limited storage capacity. Across the set of 12 nested catchments, stream responses to precipitation were highly variable. Runoff coefficients were typically highest in catchments dominated by less permeable bedrock, as opposed to catchments with permeable bedrock, exhibiting low runoff coefficients. We found that the fractions of new water (Fnew) determined via ensemble hydrograph separation (as per Kirchner, 2019), i.e., water less than two weeks old, were correlated to bedrock geology. In catchments with mixed (i.e., permeable and less permeable) bedrock types, we noticed an increase in Fnew with discharge mirroring the domination of groundwater contributions from areas with permeable bedrock during low to medium discharge and the activation of fast flow paths in sectors dominated by less permeable substrate at higher discharge. Findings shed new light on the role of bedrock geology on fundamental catchment functions of water collection, storage, mixing and release. The latter largely determine the responsiveness of catchments to variability and/or changes in climate. This information is key for better anticipating catchment response to future changes in climate. References: Kirchner, James W. (2019). Quantifying new water fractions and transit time distributions using ensemble hydrograph separation: theory and benchmark tests. Hydrology and Earth System Sciences, 23, 303349.