| Summary: | 碩士 === 國立臺灣大學 === 海洋研究所 === 103 === We use a 3D hydrodynamic model (ROMS) to investigate the structure and mass distribution of hyperpycnal flows in the Gaoping submarine canyon. We focus on the sensitivity to sediment settling velocity (ws = 0 to 1 mm/s). To simulate hyperpycnal flow events, suspended sediment concentration of 60 g/L and river discharge of 3000 m3/s is specified at the Gaoping river mouth. The numerical model is first validated by comparing model-derived sea-level, vertical profiles of velocity and temperature against observations. Basic flow characteristics is reasonably reproduced by the model.
From our numerical experiments with ws of 0, 0.1, and 0.5mm/s, hyperpycnal discharge plunges when entering the coastal ocean. The suspended sediment is concentrated near the bottom (within 1 m) and is transported offshore following the Canyon topography. These features are consistent with hyperpycnal flow. However, when ws exceeds 1 mm/s such that the sediment advective distance (D=U* Ts, U is average outflow speed, Ts is the settling time) becomes shorter than the distance between river mouth and canyon mouth, most of the hyperpcnal discharge deposits on continental shelf before reaching the steep canyon head.
We use a simple mass balance model to illustrate the importance of resuspension/erosion processes in maintaining hyperpycnal flow in the canyon. The simple model is based on a balance of mass input from the river and mass removal due to deposition. Without considering resuspendsion processes, the simple model significantly underpredicts the suspended mass in the canyon, suggesting that appreciable amount of sediment is resuspended by the hyperpycnal flow itself to maintain its gravitational force.
We also quantify the offshore penetration of the hyperpycnal flow. The bottom-intensified velocity structure (i.e. nose-shaped) obtained from prior laboratory experiments is used as a reference. The penetration distance is defined as the location beyond which the hyperpycnal flow deviates from the nose-shaped structure. The estimated penetration distance is consistent with the center of mass. The positions where hyperpycnal flow loses the nose-shape structure and where divergence of mass flux elevates appear to lock in with topographic features (such as canyon bending, slope changes). These results suggest that the structure and the fate of hyperpycnal flow in the Gaoping cayon is greatly influenced by the canyon topography.
|
|---|
