This research presents the first application of a distributed discrete-continuum numerical model for simultaneous inversion of long-term spring flow in conjunction with solute- and heat-transport signatures, in a large-scale karst system. The MODFLOW-2005 CFPv2 code was adopted for this reason, utilizing PEST for automated inversion. Taking the principal of model parsimony into account, simulations were performed at two stages, testing likely conceptual models based on the hydro-geological understandings. At the first stage, simple yet realistic conceptual models were setup considering different probable variants of conduit geometries. Results of this stage highlights the importance of sinkhole-spring dynamics, in the recorded spring signatures; such that, successful simultaneous calibration of all spring signatures could not be obtained unless high linkage of conduits to the sinkholes all over the catchment was adopted. Subsequently, some additional rational complexity was added in order to achieve a more demonstrable model of the karst system, as the second stage. Model selection criteria, fitting statistics and reasonableness of calibrated parameters were assessed, at this stage. Although one of the model variants could statistically represent all observations with reasonable parameter values and high agreement with catchment hydrogeological perspective, it still cannot be approved as a predictive tool, unless field measurements on conduit system characteristics and matrix hydraulic parameters, as well as long-term observation of aquifer physicochemical signatures be collected for further validation/improvement of conceptual and numerical model. Achieved simultaneous inversion results can serve as a proof of concept for applicability of physically based discrete-continuum modeling approach for flow and transport modeling in karst.