11:45 AM - 12:00 PM
[ACG43-10] Development of Rating Curve Using Hydraulic Simulation with 3D Topography Data in Mountain Stream
Keywords:mountain river, hydraulic model, rating curve, step-pool
River discharge monitoring is important for understanding the rainfall-runoff process, water resource management, and flood forecasting. However, stream rivers with a catchment area of 1-10 km2 are in an observation gap area where discharge monitoring is difficult due to complex topography such as step-pool structures and turbulent flow, which make it difficult to install weirs and measure flow velocity and channel cross-sections. Although there are examples of continuous observation of water levels and temporary discharge monitoring at multiple points within a watershed, a method for monitoring discharge at high spatial and temporal densities has not yet been established. The objective of this study was to construct a rating curve to convert the time-series variation of water level into discharge in a mountain stream area.
The method for constructing a rating curve is to reproduce the flow of water in a stream area using a hydraulic model, and we attempted to demonstrate this method in the upstream area of the Aono River in the Aono Research Forest of the University of Tokyo. Measurement using a smartphone-mounted lidar and high-density level surveying at intervals of 10 cm in the cross-sectional direction and 20 cm in the longitudinal direction were conducted to obtain the 3D topographic data necessary for the hydraulic simulation. In addition, the distribution of flow velocity and depth was observed to validate the model and calculate the discharge.
The topographic data and discharge obtained from the observations were input to a hydraulic model to simulate a planar two-dimensional flow field. As for the topographic data, since lidar cannot accurately capture the topography below the water surface, replacing the topographic data with manual survey data was performed in areas with a water surface. Manning's roughness coefficient, a parameter of frictional resistance, was estimated by calibrating the model to minimize the error from the observed water level when the model was given the discharge calculated from the observed velocity distribution as the boundary condition at the upstream end. The RMSE between the water level calculated with the estimated roughness coefficient and the observed value was 4.2 cm.
The relationship between water level and discharge at any point in the subject section was then examined by inputting time-varying discharge into the model using the estimated values of the roughness coefficients. It was observed that the water level for the same discharge was higher when the discharge was decreasing than when it was increasing, which was more pronounced at low water. The water level fluctuation in response to discharge change in a stream occurs at the interval of a step-pool, and during low water, the water stays in the pool and the discharge at the step is dominant, but during flooding, the influence of the step-pool decreases, suggesting that a one-to-one relationship between water level and discharge is established.
Although further validation with observations is needed, the relationship between water level and discharge in the stream area could be estimated. The hydraulic characteristics revealed in this study are expected to provide important insights into the selection of water level measurement locations and the conversion method from water level to discharge for stream discharge monitoring.
The method for constructing a rating curve is to reproduce the flow of water in a stream area using a hydraulic model, and we attempted to demonstrate this method in the upstream area of the Aono River in the Aono Research Forest of the University of Tokyo. Measurement using a smartphone-mounted lidar and high-density level surveying at intervals of 10 cm in the cross-sectional direction and 20 cm in the longitudinal direction were conducted to obtain the 3D topographic data necessary for the hydraulic simulation. In addition, the distribution of flow velocity and depth was observed to validate the model and calculate the discharge.
The topographic data and discharge obtained from the observations were input to a hydraulic model to simulate a planar two-dimensional flow field. As for the topographic data, since lidar cannot accurately capture the topography below the water surface, replacing the topographic data with manual survey data was performed in areas with a water surface. Manning's roughness coefficient, a parameter of frictional resistance, was estimated by calibrating the model to minimize the error from the observed water level when the model was given the discharge calculated from the observed velocity distribution as the boundary condition at the upstream end. The RMSE between the water level calculated with the estimated roughness coefficient and the observed value was 4.2 cm.
The relationship between water level and discharge at any point in the subject section was then examined by inputting time-varying discharge into the model using the estimated values of the roughness coefficients. It was observed that the water level for the same discharge was higher when the discharge was decreasing than when it was increasing, which was more pronounced at low water. The water level fluctuation in response to discharge change in a stream occurs at the interval of a step-pool, and during low water, the water stays in the pool and the discharge at the step is dominant, but during flooding, the influence of the step-pool decreases, suggesting that a one-to-one relationship between water level and discharge is established.
Although further validation with observations is needed, the relationship between water level and discharge in the stream area could be estimated. The hydraulic characteristics revealed in this study are expected to provide important insights into the selection of water level measurement locations and the conversion method from water level to discharge for stream discharge monitoring.