For investigating selective withdrawal problems, laboratory experiments were conducted in a horizontal flume with a co-directional two-layer flow of different density into a line sink. Saline water was used as the fluid of the lower layer instead of sediment-laden turbid water for achieving a steady state. In this study, a trend curve was produced using the data of many runs which were taken under varying conditions from aspiration of both layers to only lower-layer aspiration. An adequate parameter for determining the critical condition was obtained from this trend curve. The critical condition is defined as the beginning or ending of aspiration for the upper layer. Suitable variables for the parameter are also discussed. A theoretical formula is suggested and verified, which is better than a traditional empirical formula for calculating the withdrawal concentration. Effects of slot elevation on the lower layer flow and the thickness of the mixing layer are also discussed.

It is observed, during flood events, that bed forms initially grow in height and make the riverbed rougher. But later, under high discharge, the bed forms grow longer with the opposite effect of making the riverbed smoother. After the discharge drops to a lower value, new bed forms regenerate on top of the elongated bed forms. This mechanism leads to a significant variation in the bed roughness and the water stage and hence determines the behavior of floods and the risk of flood disasters. This work presents detailed modeling of bed forms under discharge hydrographs and simulates the conditions under which the bed is flattened out in the upper plane bed regime. The flow was simulated by large‐eddy simulation, and the sediments were considered as rigid spheres and modeled in a Lagrangian framework. The bed morphodynamics were the result of entrainment and deposition of sediment particles. We examined several discharge hydrographs. In the first case, we increased the discharge linearly and then kept it constant after reaching the upper plane bed condition. The dunes were generated and grew during the rising stage of discharge. When the flow conditions reached the upper plane bed regime, high‐frequency ripples were generated and helped to flatten the bed. The results also showed that in contrast with mechanisms in the dune regime, the flattening of the bed was associated with a distinct pattern of sediment transport which deposited sediment mainly in the lee side of the dunes and led to flattening of the bed. After flattening, the sediments were mainly transported in suspension mode. As long as flow conditions stayed in the upper plane bed regime, the bed remained flat with small high‐frequency ripples. We also examined two other scenarios: one with an immediate falling stage of discharge after the rising stage and the other with a period of constant discharge between the rising and falling stages. Dunes were regenerated during the falling stage of discharge for both cases. In the first case, the dunes were regenerated before the bed was flattened because of a time lag between the flow and the bed deformation. Bed flattening was followed by dune regeneration in the second case. Moreover, a significant reduction in the form drag coefficient was observed after the flattening in the second case. The reduction in form drag led to a reduction in the flood intensity. The model showed its ability to reproduce the key phenomena of the development of subaqueous dunes under conditions of a passing flood wave.

Interflow is an important water source contributing to river flow. It directly influences the near-surface water cycles for water resource management. This study focuses on assessing the interflow potential and quantifying the interflow in the downstream area along the Kaoping River in southern Taiwan. The interflow potential is first determined based on the modified index-overlay model, which employs the analytical hierarchy process (AHP) to calculate the ratings and weightings of the selected factors. The groundwater and surface water flow (GSFLOW) numerical model is then used to link the index-overlay model to quantify the interflow potential for practical applications. This study uses the Monte Carlo simulations to assess the influence of rainfall-induced variations on the interflow uncertainty in the study area. Results show that the high potential interflow zones are located in the high to middle elevation regions along the Kaoping River. Numerical simulations of the GSFLOW model show an interflow variation pattern that is similar to the interflow potential results obtained from the index-overlay model. The average interflow rates are approximately 3.5 × 10⁴ (m³/d) in the high elevation zones and 2.0 × 10⁴ (m³/d) near the coastal zones. The rainfall uncertainty strongly influences interflow rates in the wet seasons, especially the peaks of the storms or heavy rainfall events. Interflow rates are relatively stable in the dry seasons, indicating that interflow is a reliable water resource in the study area.