Rivers are conduit that transport terrestrial matters to the ocean. Studying how the composition and transport mechanism of terrestrial matters influenced by tide, wave, wind and discharge is an important issue in the disparate of river-sea systems. The aim of this study is to understand the temporal and spatial variability of the grain size distribution of a river plume, using EOF (Empirical Orthogonal Function) to analyze the correlation between environmental processes and suspended sediments. We used CTD and LISST-100 to collect upper-colum profile data at Gaoping River mouth in five (R/V Ocean Research III) cruises from 2005 to 2009. We also deployed an instrumented tetrapods and a moored buoy at the inner Gaoping River mouth and the inner continental shelf off Gaoping River mouth for two days, respectively. In all the upper-column data, the volume concentration increased in surface column when the salinity decreased. The variability of volume concentration are dominant in grain sizes between 10-250 £gm. We used EOF to analyz the time series to investigate the correlations among the volume concentration of 5 grain-size groups (<3, 3-10, 10-63, 63-153, and 153-250 £gm), salinity, water temperature, and alongshore and across-shore winds. The first eigenmode explains about 50 % of the total correlations. This mode describes the dominant influence of the river runoff that affected all the grain-size classes within the plume. This mode suggests that the grain-sizes between 3-250 £gm are of terrestrial origin (low salinity, high water temperature) exported during the ebbing tide. The second eigenmode accounts for about 20 % of the total correlations. This mode describes the dominant influence of the wind. When under the upwelling-favorable across-shore wind, the upwelling brings low temperature and clay-sized suspended sediment from the submarine canyon to the surface. The results suggest that the size-classes greater than 3 £gm are terrestrial suspended sediment and transported by the river plume, the size-classes finer than 3 £gm are mostly from the submarine canyon by the upwelling-favorable winds.

Kao-ping submarine canyon is straightly connected with Kao-ping River. Thus, it is influenced by the river¡¦s discharge. In order to investigate this seasonal difference in the canyon, we conducted several researches, which were based on temporal and spacial observations of the hydrological and dynamical characteristic of the canyon. We also studied the transport of suspended sediment particles in the canyon according to the distribution of SSC (suspended sediment concentration). In our research, we made along-canyon profile observations. We collected data by using CTD, ADCP and Transmissometer on R/V Ocean Researches Vessel ¢». The data include flood season cruises (CR536, CR248, CR552, CR634 and CR639), and dry season cruises (CR572, CR598, CR608). Based on the temporal and spacial observations, in June and July 1999 we found a cold pool at the location where canyon topography changed sharply. We also found the density disturbance and high concentration at the same location. When the tidal current was landward in the deep canyon, we can observes upwelling in the head region of the canyon. Conversely, we can observe downwelling while the tidal current was seaward. Moreover, we observed a strong vertical flow in the dry season, which can make resuspended easily. Besides, we deployed an instrument covey, including sediment traps, Recording Current Meter (RCM), wave gauge (S700) and Laser In-Suit Scattering and Transmissiometry (LISST-100), at the location where we previously observed SSC localizes high for one month (from June to July 2000) to investigate the temporal relations among temperature, salinity, flow, and suspended particles. From this time series observation we found a coincidence between cold temperature signal and the northward flow. It showed that the cold water from deep sea was transported to the head region of the canyon by a ¡¥Pump¡¦ process. We also observed the suspended particles at the top (195m), middle (245m), and the bottom (285m) of the ocean. According to their different sizes, we found that the upper depth has more sands which were from the continental shelf, and the middle and lower depth have more silt and clay. Subsequently, we used Empirical Orthogonal Function (EOF) analysis to explain the relationship among hydrological and flow factors of the canyon. We concluded that two major modes to explain the observed relationship: Submarine canyon seasonal effect, and dynamic stability.