An observational experiment of coastal acoustic tomography to map the structure of tidal currents in Sanmen Bay, China
Zhu Ze-Nan,Zhu Xiao-Hua,Zhang Chuan-Zheng,Fan Xiao-Peng,Liao Guang-Hong,Xuan Ji-Liang,Long Yu,Ma Yun-Long,Zhao Rui-Xiang,He Zhi-Guo,Zhang Tao,Zhang Xiang-Ming
DOI: https://doi.org/10.6038/cjg20150524
2015-01-01
Abstract:Synchronous tidal current data are usually difficult to collect in regions disturbed by fishery activity, such as Sanmen Bay, China. Coastal Acoustic Tomography (CAT) is a promising technique to solve this problem, with which we measure tidal data can be recorded continuously even in busy coastal regions. To observe the tidal currents in Sanmen Bay using CAT, we set up seven CAT stations in the bay in September, 2009 (6th to 9th). The transducer was suspended at a depth of 3 similar to 5 m. Acoustic signals (5 kHz), modulated by the 10th order M sequence were transmitted synchronously every 3 min. Meanwhile, the transducer received transmission signals from the six other stations to obtain the reciprocal travel time of each station pair. At night, an Acoustic Doppler Current Profiler (ADCP) was deployed to measure the current velocity and so verify the CAT results. In addition, conductivity-temperature-depth (CTD) casts were performed to generate the sound-speed profile. Inverse analysis was performed to the travel-time differences between each station pair, to obtain the horizontal distribution of the tidal currents. The inverse solution was obtained by the tapered least squares method, and the weighting factor (appearing in the inverse analysis) was determined using the L-curve method. The travel-time differences of the 21 station pairs show a typical semidiurnal oscillation. However, some of the data are missing due to an instrument failure. This failure also affects the subsequent inverse analysis. Missing data are interpolated using values predicted by harmonic analysis. The mean root mean square difference (RMSD) between the raw data and the predicted values is 0.49 ms, which indicates that the predicted values are very close to the raw data. The time series of the velocity fields (obtained by the inverse analysis) shows that the tidal currents flow mainly along the channel, i.e., toward northwest during the flood tide and southeast during the ebb tide. The maximum northwestward and southeastward current velocities are 1.03 m . s(-1) and 1.09 m . s(-1), respectively. The harmonic analysis of the tidal currents shows that the semidiurnal tidal constituent (M2) is dominant. The semi major axes of its tidal ellipses are directed southeast-northwest. The residual currents flow mainly northwestward, with an average velocity of 0.05 m . s(-1) The time series of the volume transport across the peripheral transects of the observed region during 5 periods of M2 shows that the mean inflow volume transport across all transects is 4735 m(3) . s(-1). This inflow volume transport is approximately equal to the outflow of 4371 m(3) . s(-1) The current velocities obtained by the inverse method agree with those obtained using the ADCP. The RMSD between the two measurements of the eastward and northward velocity components are 0.18 m . s(-1) and 0.16 m . s(-1), respectively. Thus, using CAT, we measured the continuous distribution of tidal currents over an area of about 58 km(2) in Sanmen Bay. The tidal and residual current structures we obtained supported previous numerical modeling results. Furthermore, our data agreed better with ADCP measurements than that data from previous CAT experiments, because we used the harmonic analysis method to interpolate the missing travel-time difference data. Our study suggests that CAT is a powerful tool to map tidal current structures. In addition, CAT could play an important role in the environmental monitoring of Chinese coastal regions.