Conceptual hydrogeological model of flow and transport of dissolved organic carbon in a small Jura karst system
Résumé
This paper aims at characterizing infiltration and solute transport processes in a karst system during flood events in order to build a conceptual model of hydrogeological functioning. The study site is the small karst system of Fertans in the French Jura. Soil cover, rock matrix, and a small spring were monitored to measure hydrological and hydrochemical parameters, and particularly a continuous dissolved organic carbon (DOC) signal, which is a relevant environmental tracer of fast infiltration. We used two combined approaches. First, from hydrodynamic and hydrochemical data, we identified the system structure (consistent with a dual porosity scheme) and characterized the main processes occurring during flood events. Second, on the basis of this scheme, we built a new conceptual hydrogeological model coupled with DOC transport to numerically validate the hydrological functioning. This modelling approach is based on a rainfall-discharge model to simulate spring flow. Solute transport is modelled using mixing equations, including an empirical retardation factor, as well as a first order solute decay. The model was calibrated and validated on a set of nineteen flood events, showing its performances in simulating spring hydrographs and delayed DOC signals during flood events with various rainfall intensities. We showed that the recharge area of the karst system varied largely with low and high groundwater periods, which was attributed to the state of hydraulic connectivity in the unsaturated zone. The soil cover appeared to play an important role of mixing and transfer for the recharge water. The model simulated the contributions of pre-event and event waters during flood events and allowed a better quantification of the available resource. It showed, in particular, that total discharge of some flood events during low water periods is mainly composed of pre-event water via piston flow-type processes. Finally, this study shows that the mixing model can simulate solute transport correctly, taking into account degradation and retardation processes. It highlights the need for a quantitative approach on hydrochemical studies of karst systems in order to understand them better.