One of the impacts of climate changes nowadays is the increase in the frequency of highintensity rainfall events alternating with extreme dry periods, which affect the components of the hydrologic cycle, such as runoff, infiltration, and aquifer recharge. Several experimental investigations and theoretical studies have demonstrated that infiltration flow in fractured media can develop along preferential pathways. However, the prediction of infiltration phenomena in fractured media still remains an open issue. This, together with erratic rainfall patterns due to climate changes,
affects the prediction of aquifer recharge and contaminant transport in fractured aquifers. The present work contributes to reducing this research gap by means of experimental investigation and forecast analysis, with a focus on the geometrical properties of single fractures and their influence
on flow patterns. Several fracture surfaces based on different fractal dimensions, standard deviations, and mismatch lengths were designed using the SynFrac model and were generated by 3D printing technology. The results revealed that the fracture’s fractal dimension has a significant impact on the number of flow paths, while the fracture inclination only increases the number of intermediate preferential channels, and, hence, modifies the flow rate distribution over the fracture outlet. Additionally, the change in the inclination angle of the dry fracture from 55 to 65 degrees resulted in an 8% reduction in the mean width of first flow path. A sensitivity analysis using an M5 tree indicates that there is a linear relationship between flow rate and the exponential form of the fractal dimension. The location of flow channels is a function of fracture fractal dimension, and the influence of mismatch length on their location is negligible. Finally, an accurate prediction algorithm with a Nash value of 0.81 was developed using Wavelet transform in order to estimate the time series of periodic flow rates over the fracture outlet
