| Summary: | 碩士 === 國立中興大學 === 水土保持學系所 === 104 === Due to climate change, Taiwan had frequent rainfall that made lots of flood and debris flow happened. Bedload transport caused severe disaster like river channel vicissitude and riverbed evolution that had tremendous effects of downstream resident’s lives and property.
In order to investigate the frequency range of seismic signal induced by bedload transport and water flow, we conducted a controlled dam breach experiment at Landao Creek, Huisun Forest, Taiwan. We utilized the advantage of bedload trap which can collect different sediment particles, and installed bedload trap at the downstream of dam. We collected sediment after dam break to analyze its grain size distribution. Simultaneously, we cooperated with portable seismometers near the river to receive the immediately continuous seismic signal, and analyzed the correlation between bedload transport and seismic signal. We also conducted comparison and confirmation via image, water depth, 3D LiDAR and other in situ data. At last, we used sediment impact model to convert bedload flux of each experiment. This study mainly divided into two parts. First we used two different position’s seismometers (near dam and downstream of dam) to discriminate frequency range of seismic signal induced by bedload transport and water flow, and investigated the relationship between bedload transport and water flow. Secondly, we used sediment grain size inside bedload trap, water depth and in situ data to estimate the possible error range of bedload flux when conducted model calculation.
The results demonstrated that the main frequency range of three times experiments is at 15-55 Hz, and the seismic signal bandwidth together with hydrology parameter were mainly appeared clockwise hysteresis phenomenon. In addition, when bedload trap is closer to the dam, we can collect relatively big size sediment transport. Thus, seismometer can receive relatively low frequency noise signal as it near to the bedload trap, and V-shape will shift to the lower frequency direction. The high frequency signal received by every station will obviously decrease as the distance increase. Therefore, when seismometer far away from river, it received weaker high frequency signal. In the end, when we used model to convert bedload flux, we found out that as we have good control of bedload median diameter (D50), water depth (H), and attenuate parameter (Qs), the model has certain reliability estimating median diameter (D ̂50) and Power Spectral Density curve. As downstream open water has stronger hydrology condition, the bedload flux estimated by model will become larger, conversely, smaller.
|
|---|
