Search

An integrated ecohydrological modeling approach was deployed in three long‐term critical zone (CZ) observatories of the French CZ network (CZ Observatories—Application and Research) to better understand how the CZ heterogeneity modulates the water cycle within territories. Ecohydrological simulations with the physically based model EcH2O‐iso constrained by a wide range of observations crossing several disciplines (meteorology, hydrology, geomorphology, geophysics, soil sciences, and satellite imagery) are able to capture stream water discharges, evapotranspiration fluxes, and piezometric levels in the Naizin, Auradé, and Strengbach watersheds. In Naizin, an agricultural watershed in northwestern France with a schist bedrock underlying deep weathered materials (5–15 m) along gentle slopes, modeling results reveal a deep aquifer with a large total water storage (1,080–1,150 mm), an important fraction of inactive water storage (94%), and relatively long stream water transit times (0.5–2.5 years). In the Auradé watershed, representative of agricultural landscapes of the southwestern France developed on molasse, a relatively shallow regolith (1–8 m) is observed along hilly slopes. Simulations indicate a shallow aquifer with moderate total water storage (590–630 mm), an important fraction of inactive water storage (91%), and shorter stream water transit times (0.1–1.3 years). In the Strengbach watershed, typical of mid‐mountain forested landscapes developed on granite, CZ evolution implies a shallow regolith (1–5 m) along steep slopes. Modeling results infer a shallow aquifer with the smallest total water storage (475–575 mm), the shortest stream water transit times (0.1–0.7 years), but also the highest fraction of active water storage (18%). Plain Language Summary: Understanding how water is stored and released in landscapes is essential for predicting water availability in a changing world. It is of course driven by the local climate, but also by the landscape settings, from the vegetation to the belowground structure inherited from the geological history. In three intensively studied observatories across France, we used numerous field measurements and a numerical model of water‐vegetation‐subsurface interactions to better depict how differences between landscapes shape the water cycle. We found that the geological history determines how much water is stored in watersheds, through the thickness of water‐holding rocks and soils. How this storage contributes to surface processes (including river flow) is modulated by more recent changes, from valley morphology and tectonics to agricultural practices and precipitation patterns. This study highlights that considering the continuity between long and short‐term landscape processes is key to a detailed understanding of the water cycle. Key Points: The long‐term critical zone (CZ) evolution controlling the regolith thickness strongly impacts the total water storage in watersheds, while the Quaternary geomorphological evolution influences the current hydrological partitioning and the separation of hydrologically active and inactive water storageBoth internal watershed characteristics and external forcings, such as current atmospheric forcing and recent land use need to be considered to infer stream persistence and to understand hydrological diversityThe observed hydrological diversity cannot be fully understood without considering a continuum of time scales in CZ evolution [ABSTRACT FROM AUTHOR]

U–Th–Ra disequilibria, cosmogenic in situ 10 Be concentrations and major and trace element concentrations have been analyzed in a 2 m-deep weathering profile sampled at the summit of the granitic Strengbach catchment (France). The data have been used to independently estimate both the long-term regolith production and denudation rates and the weathering and erosion rates. Modeling of the 238 U– 234 U– 230 Th– 226 Ra disequilibrium variations in the lower part of the profile yields a regolith production rate of 12 ± 4 mm/kyr (30 ± 10 T/km 2 /yr), while modeling of the high-resolution 10 Be concentration profile leads to an exposure age of 19.7 ± 2.2 kyr , an inherited concentration of 15,000 ± 1,000 at/g in quartz and a mean denudation rate of 22 ± 10 mm/kyr (37 ± 15 T/km 2 /yr). The consistency between production and denudation rates suggests that, on a millennial timescale, the regolith mass balance at the summit of the catchment is close to a steady state, even if the watershed may have been impacted by Quaternary climatic changes and by recent anthropogenic perturbations (e.g., 20th century acid rain and recent afforestation efforts). The results also indicate that physical erosion is likely the dominant long-term process of regolith denudation in the catchment. Furthermore, the comparison of the long-term production and denudation rates and of weathering and erosion rates determined from the depth profile analyses with the current weathering and erosion rates estimated at the outlet of the watershed based on monitoring of the water chemistry and sediment fluxes suggests that physical erosion may have varied more than the chemical weathering flux during the last 150 kyr. Although very few other sites with U-series, in situ 10 Be and stream monitoring data are available for comparison, the current data suggest that (1) the mass balance steady state of regolith might be commonly achieved in soil mantled landscapes, and (2) physical erosion has varied much more than chemical weathering in mid-mountain catchments over the last 10–150 kyr. These results highlight the importance of the combined analysis of U-series nuclides and in situ 10 Be in the same weathering profile for the determination of key geomorphic parameters, which are important to constraining landscape stability and the responses of landscapes to natural or anthropogenic forcing. [ABSTRACT FROM AUTHOR]

This study presents the results of the coupled hydrogeochemical modeling of the geochemical compositions of spring and borehole waters from the Ringelbach catchment, which is located in the Vosges Mountains (France). This site has been equipped with 150-m-deep boreholes, facilitating the sampling of both rock and groundwater in the granitic bedrock. The data point to very contrasting chemical compositions between spring and borehole waters, which are discussed and explained in this study by the application of the coupled hydrogeochemical code KIRMAT. Using hydrological and geochemical data, simulations were performed through two different water pathways, which crossed different types of rocks within the Ringelbach massif: a subsurface and fast (>2.5 m H2O .yr −1 ) water flow, which is more or less parallel to the slope, for waters supplying the springs, and a rather vertical and slower flow (0.5–0.1 m H2O .yr −1 ) for the borehole waters. The KIRMAT simulations make it possible to account for not only the geochemical differences between the spring and borehole waters but also the geochemical variations observed in waters in both contexts. For borehole waters, the model confirms the importance of the dissolution of minor mineralogical phases that are present in the granite (here, carbonates/dolomites) on the chemical budget of waters. It also shows that the chemical differences between the waters collected in the two studied boreholes result from differences in the water flow in the granitic bedrock, i.e., the difference between water flow in a regular porous medium and water flow in a porous medium crossed by a fracture. This result likely highlights the role of geological inheritance on the hydrodynamical rock properties and the chemical compositions of waters circulating within the granitic bedrock. For spring waters, this model enabled us to constrain the nature of the rock in the pathway, which is neither saprolite nor fresh granite but is instead weathered granite with a weathering age of several tens of thousands of years. Spatial and seasonal variations in the chemical compositions of spring water can be explained as the result of the same circulation pattern for which the water-rock interaction time is determined by the length of the pathway and the water velocity. Especially in cases in which this interaction time is long enough, the precipitation of clay phases is enabled, which plays a major role in determining the chemical composition of the water. Despite the only one-dimensional approach and the uncertainties linked to the geochemical complexity and the associated kinetic data, the results obtained in this study demonstrate the effectiveness of using coupled hydrogeochemical modeling to better understand and quantify the weathering processes and the coupling that exists between water circulation dynamics and water-rock interactions at the catchment scale. [ABSTRACT FROM AUTHOR]