Pumping-well data for conditioning the realization of the fracture aperture field in groundwater flow models

Stochastic methods were used in a new mathematical modeling approach on a regional scale in order to study the Idume Spring discharge (Salento region, Southern Italy). The groundwater flow was addressed in a set of horizontal fractures with variable apertures in each fracture plane since the aquifer is characterized by a discrete number of poorly interconnected fractures. The new approach uses the fracture conductivity field of the aquifer, derived from field estimations in 49 pumping-wells, for the site characterization of the system. These fracture conductivities were expressed using the mean aperture values of the overall fractures, by directly inverting the groundwater radial flow equation under steady-state conditions. The regional covariance of the mean fracture aperture was considered as input in the proposed stochastic computational method to generate the aperture values in each fracture plane. The continuity equation, on the basis of the solution of Navier–Stokesʹ equation, was integrated into each fracture by using the finite difference method. A computational code performed the evaluation of the outflow in every single fracture of the idealized set. The computed (1071±2.1 L/s) and measured (1068 L/s) Idume Spring discharge were compared, demonstrating positive performances of both the proposed stochastic approach and computational procedures.