Abstarct: Fracture modeling is essential as a part of the simulation and descxription in the broken natural reservoirs for multiple applications of the reservoirs. A fracture network model should include all the elements and properties of fractures that play an important role in a field. Due to the effects of field faults on fractures and their properties, it is a challenging topic to establish a meaningful relationship between all these properties. There are different kinds of methods for determining the characteristics of the natural fractured reservoirs and fluid flow simulation, which are divided into three groups of continuous models, discrete models, and hybrid models. Producing hybrid modeling is done by statistical information derived from measured data in a field. In this study, hybrid modeling was performed using Image logs and other petrophysic logs of three wells and data related to the seismic cube of the studied field. The fracture zones with different fracture intensities were identified by the Image log's data. The distribution of the fracture intensity in the field is associated with uncertainties. For this purpose, secondary data which are fracture drivers can be used. In this study, seismic attributes were used as secondary data to improve the fracture intensity distribution using the GRFS algorithm in petrophysical modeling. By using neural network, different types of improved fractures were integrated as inputs and a good fracture intensity was obtained as output. Correlation coefficient between neural network output and fracture drivers was 0.93, which indicates a very good relation between fracture drivers and fracture intensity from neural network. Seismic attributes were used to determine, interpret and model the field faults. baxsed on validation, fault interpretation and modeling are highly accurate. Then by Fracture and fault classification into 4 fracture zones, the discrete fracture model was created. In the next step, the implicit fracture model was created by specifying the threshold of 250 ft for the length of the fractures. In the end, the hybrid model was obtained by upscaling the fracture network models (DFN and IFM) using a statistical method (oda). The outputs of the hybrid model are the fracture porosity ϕ f, the fracture permeability Kf, and the sigma factor σ, which can be used to simulate the reservoirs. The fracture porosity is highest in the fault zone and the highest permeability is observed in the curve of the reservoir and in the western direction. Also, the Sigma factor is the highest in the curve of the reservoir and to the west of the field. In the northeastern field, the sigma factor reaches its lowest level, indicating that there is no relation between the matrix and fractures in this part of the field.