As a typical water exchange of surface water and groundwater, hyporheic flow widely exists in streambeds and is significantly affected by the characteristics of sediment and surface water. In this study, a low-permeability clay lens was chosen to investigate the influence of the streambed heterogeneity on the hyporheic flow at a river section of the Xin’an River in Anhui Province, China. A 2D sand tank was constructed to simulate the natural streambed including a clay lens under different velocity of surface water velocity. Heat tracing was used in this study. In particular, six analytical solutions based on the amplitude ratio and phase shift of temperatures were applied to calculate the vertical hyporheic flux. The results of the six methods ranged from −102.4 to 137.5 m/day and showed significant spatial differences. In view of the robustness of the calculations and the rationality of the results, the amplitude ratio method was much better than the phase shift method. The existence of the clay lens had a significant influence on the hyporheic flow. Results shows that the vertical hyporheic flux in the model containing a clay lens was lower than that for the blank control, and the discrepancy of the hyporheic flow field on both sides of the lens was obvious. Several abnormal flow velocity zones appeared around the clay lens where the local hyporheic flow was suppressed or generally enhanced. The hyporheic flow fields at three test points had mild changes when the lens was placed in a shallow layer of the model, indicating that the surface water velocity only affect the hyporheic flow slightly. With the increasing depth of the clay lens, the patterns of the hyporheic flow fields at all test points were very close to those of the hyporheic flow field without a clay lens, indicating that the influence of surface water velocity on hyporheic flow appeared gradually. A probable maximum depth of the clay lens was 30 to 40 cm, which approached the bottom of the model and a clay lens buried lower than this maximum would not affect the hyporheic flow any more. Influenced by the clay lens, hyporheic flow was hindered or enhanced in different regions of streambed, which was also depended on the depth of lens and surface water velocity. Introducing a two-dimensional sand tank model in a field test is an attempt to simulate a natural streambed and may positively influence research on hyporheic flow.