Реферат: Artificial tracer tests constitute one of the most powerful tools to investigate solutetransport in conduit-dominated karstic aquifers. One can retrieve information about the internalstructure of the aquifer directly by a careful analysis of the residence time distribution (RTD).Moreover, recent studies have shown the strong dependence of solute transport in karstic aquifers onboundary conditions. Information from artificial tracer tests leads us to propose a hypothesis aboutthe internal structure of the aquifers and the e_ect of the boundary conditions (mainly high or lowwater level). So, a multi-tracer test calibration of a model appeared to be more consistent in identifyingthe e_ects of changes to the boundary conditions and to take into consideration their e_ects on solutetransport. In this study, we proposed to run the inverse problem based on artificial tracer tests with anumerical procedure composed of the following three main steps: (1) conduit network geometrieswere simulated using a pseudo-genetic algorithm; (2) the hypothesis about boundary conditionswas imposed in the simulated conduit networks; and (3) flow and solute transport were simulated.Then, using a trial-and-error procedure, the simulated RTDs were compared to the observed RTDon a large range of simulations, allowing identification of the conduit geometries and boundaryconditions that better honor the field data. This constitutes a new approach to better constrain inverseproblems using a multi-tracer test calibration including transient flow.