Vegetation in rivers often occurs in patches of different sizes playing significant role in form induced, sediment traort, biological and ecological of river systems. In addition, they have a significant role on flow structures. The presence of vegetation patch is different from other obstacles in rivers due to its permeability, however, the presence of an obstacle (e.g., bridge pier) changes the velocity field as well as the vegetation patch do. The effect of vegetation patch on the flow structure is weakly recognized. So investigation of the interaction between submerged vegetation patch and flow structure is of crucial importance. Flow within and just above vegetation behaves similarly to mixing layer rather than the boundary layer, thus at this research the mixing layer theory to quantify the interaction between flow and submerged vegetation patch was evaluated, although, the effect of developing flow over small patches has not been considered in the canonical mixing layer theory. Accordingly, it is essential to combine a canonical mixing layer model and modified equations to quantify evolving area along the patch. Another major aim of this dissertation is to predict vegetation effect on hydraulic resistance in stream by applying Computational Fluid Dynamics (CFD) method to resolve the Reynolds equations. To achieving this goal, The Fluent software, Reynolds Stress model and the Porous zone model were used. To examine the evolving mixing layer equation for the flow along the vegetation patch and the estimated drag coefficient based on numerical solution of Reynolds equations, experimental was conducted in laboratory and river. Laboratory experiments were carried out over an artificial vegetation patch with 4.6 m length (with developing and fully-developed flow) in a rectangular flume in the Hydraulic Laboratory of Isfahan University of Technology for flow discharge 21 and 29 l/s. Field experiments were conducted in Pelasjan River in Chadegan city at Isfahan province and in Beheshtabad River in Ardal city at Chaharmahal Bakhtiari province. The maximum length of the selected vegetation patch in Pelasjan River was 2.7m (with developing flow). The density and vegetation height of canopy increased toward the end of the heterogeneous patch. Moreover, the maximum length of the selected vegetation patch in Beheshtabad River was 0.9 m (with only developing flow). To investigate the impact of canopy density on flow structure, the experiments in Beheshtabad River were carried out with 3 different densities with ahp equal to 4.5, 3.3, and 2.7. The results reveal that there is a reasonable agreement between the measured value of velocity and Reynolds stress profile with the estimated ones by evolving mixing layer equations. The result reveal that there is a reasonable agreement between the measured values of velocity and Reynolds stress profile with the estimated ones by evolving mixing layer equations. However, the spreading coefficient of this model less than ones was reported in literature, 0.06-0.12 thanks to the limitation of vertical development of the mixing layer width by bed and small canopy height or bed and water surface or canopy density. Also, the results show that the mixing layer thickness over vegetation patch with only developing flow (such as selected patches in Pelasjan and Beheshtabad rivers) follows as linear trend, however, it follows logarithmic trend for developing and fully developed flow region (such as vegetation patch in experimental flow). In addition, the flow immediately downstream the vegetation patch behaves similar to a jet and is parameterized by two conventional models, conventional logarithmic low and mixing layer theory. These models present a reasonable agreement with the measured velocity profiles immediately downstream the patch. The predicted drag coefficient based on Reynolds equation and empirical equation has a proper agreement along vegetation patch in experimental flume (with two different discharge) and in Beheshtabad river (with three different density). However, the empirical equation does not show the fluctuation of drag coefficient at the leading edge. Finally, the velocity and stress acquired by numerical solution of Reynolds equation has a rational agreement with the measured ones, but there was no significant difference between predicted profiles based on empirical equation or numerical solving of Reynolds equations. Indeed it can not modify the output of CFD models using porous zone due to the low sensitivity of porous zone to slight fluculation of drag coefficient. Key Words Vegetattion patch, mixing layer theory, Numerical simulation, Drag coefficient, Fluent.