Transport Mechanism of Suspended Sediments and Migration Trends of Sediments in the Central Hangzhou Bay

: Based on the 2013 ﬁeld survey data of hydrology, suspended sediments and bottom sediments in the Central Hangzhou Bay, this paper explores the dynamic mechanism of suspended sediments in Hangzhou Bay by employing material ﬂux decomposition. Meanwhile, the migration trends of bed sediments are also investigated by analyzing grain size trends. The results show that during an ebb or ﬂood tide, the hydrograph of suspended sediment concentration of Hangzhou Bay is dominated by an M shape (bimodal), which is attributed primarily to the generation of a soft mud layer and a separate ﬂuid mud layer. Laterally, the distribution of suspended sediment concentration is high in the south and low in the north. From a macroscopic perspective, the net sediment transport in the study area displays a “north-landward and south-seaward” trend, presenting a “C”-shaped transport mode. That is, the sediments are transported from the bay mouth to the bay head on the north side and from the bay head to the bay mouth on the south side. The sediment transports by advection and tidal pumping are predominant, while the sediment transport by vertical circulation makes little contribution to the total sediment transport. Moreover, the sediment transport in the center of the reach area is dominated by advection, whereas that near both sides of the banks is controlled by tidal pumping. The asymmetry of the tides, i.e., ﬂood-dominance in the north and ebb-dominance in the south, is the primary cause of the dynamic mechanism for the overall “C”-shaped transport mode in Hangzhou Bay. Additionally, coupled with the narrow-head wide-mouth geomorphology, Hangzhou Bay remains evolving by south shore silting and north shore scouring.


Introduction
Bay areas, as a land-to-ocean connection and transition zone, have always been a key area of great ecological and socio-economic significance due to their rich containment of natural resources, including: reclamation development, ports and waterways, tide and wave energy, aquaculture, seawater desalination, offshore oil, natural gas, mineral resources as well as tourism resources [1][2][3][4]. In spite of the unique ecosystem, the environment of bay areas is also extremely fragile, as they are significantly affected by a range of factors, including: the runoff [5,6], sediment supply [7,8], loss of wetland [9] and sea-level rise [10]. At present, with the growing development activities in the bay areas, environmental issues have become a hot topic in the oceanography field. Additionally, sediment grains are the carriers of organic matters and nutrients, exerting a vital role in the migration and circulation of matter [11]. Therefore, studies related to suspended sediment transport and sediments migration have gained increasing attention [12][13][14][15].

Study Area
Hangzhou Bay, located in the middle of China's coastal area, is famous for its large tidal range. It is adjacent to the Changjiang River Estuary in the north, and adjacent to the Qiantang River in the west. Morphologically, Hangzhou Bay is a trumpet-shaped bay with an east-west strike ( Figure 1): its width increases continuously from the Ganpu region (the inner bay head) to the bay mouth reaching 98.5 km. Moreover, the water area of the bay is approximately 5000 km 2 [21]. The Qiantang River in the west is the largest river flowing into the bay, whose mean annual water discharge is approximately 952 m 3 /s and mean annual suspended sediment discharge is 6 million t/a [19]. The topography of Hangzhou Bay is relatively flat, and the mean water depth is generally 8-10 m. It is characterized by irregular semidiurnal shallow tides. The tidal range is 3-4 m at the bay mouth, increasing gradually towards the bay head to 4-6 m. The tidal currents are stronger within the bay, with a velocity up to 3 m/s. Meanwhile, the wind wave is small, with a mean annual height ranging between 0.2 m and 0.5 m. The sediments in the bay are composed primarily of clayey silt with the median grain size between 20 µm and 40 µm, and the suspended sediment concentration could reach 3.3-4.4 kg/m 3 [28]. Owing to the abundant sources of suspended sediments and the high sediment content, the topography is eroded and silted severely by the strong tidal currents [17,20].
Water 2020, 12, x FOR PEER REVIEW 3 of 16 increasing gradually towards the bay head to 4-6 m. The tidal currents are stronger within the bay, with a velocity up to 3 m/s. Meanwhile, the wind wave is small, with a mean annual height ranging between 0.2 m and 0.5 m. The sediments in the bay are composed primarily of clayey silt with the median grain size between 20 and 40 μm, and the suspended sediment concentration could reach 3.3-4.4 kg/m 3 [28]. Owing to the abundant sources of suspended sediments and the high sediment content, the topography is eroded and silted severely by the strong tidal currents [17,20].

Field Observations
During three separate periods in 2013: (1) from 8:00 a.m. on 31 January to 9:00 a.m. on 1 February (spring tide), (2) from 9:00 a.m. on 4 February to 10:00 a.m. on 5 February (intermediate tide), and (3) from 13:00 p.m. on 8 February to 14:00 p.m. on 9 February (neap tide), three stations (H1, H2 and H3 in Figure 1b) near the eastside of the Hangzhou Bay bridge were selected for simultaneous in-situ hydrological observations. The observations of each station were made for a continuous period of 26 h, ensuring that two complete tidal cycles were included. The mean water depth of the three stations H1, H2 and H3 were 14.4 m, 12.5 m and 8.9 m, respectively. During the observations, the flow direction and velocity were collected via acoustic Doppler current profilers (ADCP, 1 MHz) installed on the surveying vessels. The transducer draught was 1 m. The measurement interval was set to 10 min, and the vertical sampling bin size was 0.5 m. Regarding the collection of suspended sediment, horizontal water samplers were employed to collect three-layer water samples at 0.2-, 0.6-and 0.8-h (h represents the water depth). Additionally, in May of the same year, thirty-three bottom sediment samples were collected in this survey area using a clamshell dredge, with a sampling interval set around 5000 m (Figure 1).

Sample Processing
The water samples were weighed in the laboratory after suction filtration (membrane pore size 0.45 μm) and drying, so as to obtain the data on sediment contents in water bodies. Meanwhile, the bottom sediment samples were removed of organic matters by adding excess hydrogen peroxide and removed of carbonates by adding dilute hydrochloric acid. After complete reaction, distilled water was added and it was left to stand for 24 h, followed by the discarding of the supernatant. Subsequently, the residue was added with 4% sodium hexametaphosphate solution and dispersed by ultrasonication. Afterwards, the samples were poured into the Mastersizer 2000 laser diffraction particle size analyzer (Malvern Instruments, measuring range 0.02-2000 μm) for analysis. Finally, the

Field Observations
During three separate periods in 2013: (1) from 8:00 a.m. on 31 January to 9:00 a.m. on 1 February (spring tide), (2) from 9:00 a.m. on 4 February to 10:00 a.m. on 5 February (intermediate tide), and (3) from 13:00 p.m. on 8 February to 14:00 p.m. on 9 February (neap tide), three stations (H1, H2 and H3 in Figure 1b) near the eastside of the Hangzhou Bay bridge were selected for simultaneous in-situ hydrological observations. The observations of each station were made for a continuous period of 26 h, ensuring that two complete tidal cycles were included. The mean water depth of the three stations H1, H2 and H3 were 14.4 m, 12.5 m and 8.9 m, respectively. During the observations, the flow direction and velocity were collected via acoustic Doppler current profilers (ADCP, 1 MHz) installed on the surveying vessels. The transducer draught was 1 m. The measurement interval was set to 10 min, and the vertical sampling bin size was 0.5 m. Regarding the collection of suspended sediment, horizontal water samplers were employed to collect three-layer water samples at 0.2-, 0.6-and 0.8-h (h represents the water depth). Additionally, in May of the same year, thirty-three bottom sediment samples were collected in this survey area using a clamshell dredge, with a sampling interval set around 5000 m ( Figure 1).

Sample Processing
The water samples were weighed in the laboratory after suction filtration (membrane pore size 0.45 µm) and drying, so as to obtain the data on sediment contents in water bodies. Meanwhile, the bottom sediment samples were removed of organic matters by adding excess hydrogen peroxide and removed of carbonates by adding dilute hydrochloric acid. After complete reaction, distilled water was added and it was left to stand for 24 h, followed by the discarding of the supernatant. Subsequently, the residue was added with 4% sodium hexametaphosphate solution and dispersed by ultrasonication. Afterwards, the samples were poured into the Mastersizer 2000 laser diffraction particle size analyzer (Malvern Instruments, measuring range 0.02-2000 µm) for analysis. Finally, the mean grain size (µ), sorting coefficient (δ) and skewness (S k ) of the bottom sediment samples were calculated by the moment formulae Mcmanus [29]: , where P i denotes the percentage of the grain size s i and n is the total number of grain size fractions.

Data Analysis Methods
The balance of the suspended sediment flux was analyzed using the method proposed by Dyer [30], in which the tidal mean sediment flux through a unit width cross section could be calculated with the following equation: where u = u(z, t) and c = c(z, t) are the current velocity and SSC (suspended sediment concentration) at the depth respectively. Moreover, " " denotes a tidal-averaged value; and " " denotes a depth-averaged value. At any depth, u = u + u and c = c + c , where u and c are the deviations of the observed values from the depth-averaged values. The tidal fluctuations are written as u = u + u t and c = c + c t , where u t and c t are the deviations from the depth-averaged values in a tidal cycle. Meanwhile, the tidal height is h = h + h t , where h t is the deviation of the tidal height from the mean depth. In this equation, T 1 is the flux due to the non-tidal drift, the Eulerian velocity; T 2 is the flux due to Stokes drift and T 1 + T 2 is called advection term. T 3 + T 4 + T 5 is the tidal pumping term which is caused by the phase differences between SSC and tidal currents, i.e., flood-ebb asymmetry in sediment resuspension and water stratification [31]. T 6 + T 7 is the vertical circulation term related to gravitational circulation and vertical profile of velocity and SSC [32,33].
In the present study, a method proposed by Gao and Collins [34] was employed to investigate the net transport trends of bottom sediments in the central Hangzhou Bay. The method was derived from sandy beds, but after nearly 30 years' application, it is verified to be applicable to silty beds [35][36][37]. In particular, Wang and Gao [37] successfully adopted the method in the investigation of sediment transport trends in Hangzhou Bay.
This semi-quantitative filtering-based method assumes that the grain size trends of bed material occur more frequently along the sediment transport direction than in other reverse directions. It extracts the grain size parameters of sediments quantitatively and converts them into the results of sediment transport trends. Previous research [38,39] has revealed that three grain size parameters µ, δ and S k would be feasible to identify the sediment transport trends. Hence, these three parameters were further calculated and analyzed herein.
For any two adjacent sampling points in the sampling sites, eight types of grain size trends could be obtained through the comparison among three grain size parameters µ, δ and S k . Verified by previous research in the marine sedimentary environments [34,39], two types of grain size trends on sediment migration direction were found to occur more frequently: (1) the sediments are better sorted in the migration direction, with finer and more negatively skewed grains, when abrasion occurs along the transport path; and (2) the sediments are better sorted in the migration direction, with coarser and more positively skewed grains, when bed erosion occurs along the transport path [38]. The grain size parameters of each sampling point were compared one by one with those of the immediately adjacent sampling points. If one of the two situations was satisfied between sampling point a and its adjacent sampling point b, then a unit vector pointing from the sampling point a towards the sampling point b would be defined, so as to determine whether the two sampling points were adjacent. This could be verified with the characteristic distance D cr (D cr usually represents the maximum sampling interval). If the distance between two sampling points was less than D cr , they would be considered adjacent.
Otherwise, they would be considered non-adjacent. For each sampling point, all its grain size trend vectors were found out and then added together, so as to derive its resultant vector. Finally, the resultant vectors were subjected to "smooth de-noising" treatment to reflect the net transport trends of bed sediments. The detailed computational method can be found in Gao and Collins [34].

Flow Velocity Characteristics
The flow velocity processes ( Figure 2) and statistical results (Table 1) (Table 1). This indicates that the spring-intermediate-neap variations of flow velocity exhibited a positive correlation with the tidal range changes. According to the inter-station comparison of the observations during the spring tide, the flow velocity was highest at the station H3, up to 3.25 m/s, followed by the station H2, while the station H1 showed the lowest velocity, with a mean velocity of 0.93 m/s. On the whole, laterally, the velocity displayed a decreased trend from the south (H3) to the north (H1).
Water 2020, 12, x FOR PEER REVIEW 5 of 16 reflect the net transport trends of bed sediments. The detailed computational method can be found in Gao and Collins [34].

Flow Velocity Characteristics
The flow velocity processes ( Figure 2) and statistical results (Table 1)

Temporal and Spatial Distribution of SSC
Similar to other bays, the process of suspended sediment concentration in Hangzhou Bay exhibited variations in the tidal cycle and spring-neap tide, as well as spatial differences ( Figure 2). Relatively high suspended sediment concentration was yielded: (1) the maximum depth-averaged sediment concentration over a tidal cycle was found at station H3, which was 4.1 kg/m 3 in the Spring

Temporal and Spatial Distribution of SSC
Similar to other bays, the process of suspended sediment concentration in Hangzhou Bay exhibited variations in the tidal cycle and spring-neap tide, as well as spatial differences ( Figure 2). Relatively high suspended sediment concentration was yielded: (1) the maximum depth-averaged sediment concentration over a tidal cycle was found at station H3, which was 4.1 kg/m 3 in the Spring tide; and (2) the highest sediment concentration was observed at the bottom layer of station H3, with a value of up to 12.6 kg/m 3 .

Intertidal Processes of SSC
In Figure 3, the variation processes of flow velocity, flow direction and sediment concentration within the tidal cycles are illustrated for the three stations in the central Hangzhou Bay. Clearly, during a flood or ebb tide process, the sediment concentration hydrographs mainly displayed a M (bimodal) shape, i.e., two crests appeared within a flood or ebb tide cycle. The first crest appeared within 1-2 h after the turn of tidal currents, immediately following which a trough appeared. Afterwards, the second crest appeared within 1-2 h after the maximum flow velocity. Moreover, the differences between the magnitude of two crests vary among the tides. During the spring and intermediate tides, the second crests were markedly larger than the first crests, whereas during the neap tide, this was no longer evident. Regarding the occurrence frequency of bimodality, a total of 25 dual crests were observed over 36 tidal cycles of the three stations during the measuring period.
Water 2020, 12, x FOR PEER REVIEW 6 of 16 tide; and (2) the highest sediment concentration was observed at the bottom layer of station H3, with a value of up to 12.6 kg/m 3 .

Intertidal Processes of SSC
In Figure 3, the variation processes of flow velocity, flow direction and sediment concentration within the tidal cycles are illustrated for the three stations in the central Hangzhou Bay. Clearly, during a flood or ebb tide process, the sediment concentration hydrographs mainly displayed a M (bimodal) shape, i.e., two crests appeared within a flood or ebb tide cycle. The first crest appeared within 1-2 h after the turn of tidal currents, immediately following which a trough appeared. Afterwards, the second crest appeared within 1-2 h after the maximum flow velocity. Moreover, the differences between the magnitude of two crests vary among the tides. During the spring and intermediate tides, the second crests were markedly larger than the first crests, whereas during the neap tide, this was no longer evident. Regarding the occurrence frequency of bimodality, a total of 25 dual crests were observed over 36 tidal cycles of the three stations during the measuring period.

Spring-Neap Processes and Horizontal Distribution of SSC
The mean depth-averaged suspended sediment concentration over spring, intermediate and neap tides, acquired from the three stations in the central Hangzhou Bay, are displayed in Figure 4. As is clear, the sediment concentration was between 1.8-4.1 kg/m 3 , 1.3-2.8 kg/m 3 and 0.6-1.5 kg/m 3 during spring tides, intermediate tides and neap tides, respectively. That is, the spring tide period exhibited the highest sediment concentration, followed by the intermediate tide period, while the neap tide period showed the lowest concentration. The ratio of mean sediment concentrations during the spring, intermediate and neap tides was 2.7:1.9:1, presenting an evident fortnightly tidal variability. According to a study conducted by Shen et al. [40], if the suspended sediment concentration at a site is correlated strongly with the tidal current velocity, this indicates that the concentration is attributed mainly to the scouring and resuspension of bottom sediments resulting from velocity increase. Otherwise, this indicates that the sediment originates primarily from the advection transport from other sea areas.

Spring-Neap Processes and Horizontal Distribution of SSC
The mean depth-averaged suspended sediment concentration over spring, intermediate and neap tides, acquired from the three stations in the central Hangzhou Bay, are displayed in Figure 4. As is clear, the sediment concentration was between 1.8-4.1 kg/m 3 , 1.3-2.8 kg/m 3 and 0.6-1.5 kg/m 3 during spring tides, intermediate tides and neap tides, respectively. That is, the spring tide period exhibited the highest sediment concentration, followed by the intermediate tide period, while the neap tide period showed the lowest concentration. The ratio of mean sediment concentrations during the spring, intermediate and neap tides was 2.7:1.9:1, presenting an evident fortnightly tidal variability. According to a study conducted by Shen et al. [40], if the suspended sediment concentration at a site is correlated strongly with the tidal current velocity, this indicates that the concentration is attributed mainly to the scouring and resuspension of bottom sediments resulting from velocity increase. Otherwise, this indicates that the sediment originates primarily from the advection transport from other sea areas.  The analysis of the correlation between tidal mean flow velocity and tidal mean sediment concentration over six tidal cycles at three stations revealed the presence of a significant correlation between the two ( Figure 5). For the three stations H1, H2, and H3, the linear correlation coefficients were 0.84, 0.82 and 0.90, respectively ( Figure 5). The correlation between mean flow velocity and mean sediment concentration was rather high for all the three stations, suggesting that the strength of tidal current was the primary factor responsible for the change of sediment concentration in the Central Hangzhou Bay.
The magnitude relationship of sediment concentration was compared between the three stations, finding that the sediment concentration distribution in the Central Hangzhou Bay was high in the south and low in the north laterally ( Figure 4).

Transport Mechanism of Suspended Sediments
Based on formula (1), the contributions of unit-width sediment transport intensity and major sediment transport factors to the total transport amount within the spring, intermediate and neap tide cycles were computed ( Table 2). The advection transport and tidal pumping transport were the major factors of suspended sediment transport in Hangzhou Bay, while the vertical circulation transport contributed little to the net sediment transport (Table 2). Additionally, the contribution of various factors of suspended sediment transport also varied among stations. For the stations H1 and H3, the tidal pumping transport had the greatest contribution, whereas for the station H2, advection transport was predominant ( Figure 6). Table 2. Suspended sediment transport rate and contribution of main transport terms. T 1 is the flux due to the non-tidal drift, the Eulerian velocity; T 2 is the flux due to Stokes drift and T 1 + T 2 is called advection term. T 3 + T 4 + T 5 is the tidal pumping term which is related to the sediments exchange between water column and stream bed under periodic tidal dynamics. T 6 + T 7 is the vertical circulation term that is related to gravitational circulation and vertical profile of velocity and The analysis of the correlation between tidal mean flow velocity and tidal mean sediment concentration over six tidal cycles at three stations revealed the presence of a significant correlation between the two ( Figure 5). For the three stations H1, H2, and H3, the linear correlation coefficients were 0.84, 0.82 and 0.90, respectively ( Figure 5). The correlation between mean flow velocity and mean sediment concentration was rather high for all the three stations, suggesting that the strength of tidal current was the primary factor responsible for the change of sediment concentration in the Central Hangzhou Bay.
The magnitude relationship of sediment concentration was compared between the three stations, finding that the sediment concentration distribution in the Central Hangzhou Bay was high in the south and low in the north laterally ( Figure 4). The analysis of the correlation between tidal mean flow velocity and tidal mean sediment concentration over six tidal cycles at three stations revealed the presence of a significant correlation between the two ( Figure 5). For the three stations H1, H2, and H3, the linear correlation coefficients were 0.84, 0.82 and 0.90, respectively ( Figure 5). The correlation between mean flow velocity and mean sediment concentration was rather high for all the three stations, suggesting that the strength of tidal current was the primary factor responsible for the change of sediment concentration in the Central Hangzhou Bay.
The magnitude relationship of sediment concentration was compared between the three stations, finding that the sediment concentration distribution in the Central Hangzhou Bay was high in the south and low in the north laterally ( Figure 4).

Transport Mechanism of Suspended Sediments
Based on formula (1), the contributions of unit-width sediment transport intensity and major sediment transport factors to the total transport amount within the spring, intermediate and neap tide cycles were computed ( Table 2). The advection transport and tidal pumping transport were the major factors of suspended sediment transport in Hangzhou Bay, while the vertical circulation transport contributed little to the net sediment transport (Table 2). Additionally, the contribution of various factors of suspended sediment transport also varied among stations. For the stations H1 and H3, the tidal pumping transport had the greatest contribution, whereas for the station H2, advection transport was predominant ( Figure 6). Table 2. Suspended sediment transport rate and contribution of main transport terms. T 1 is the flux due to the non-tidal drift, the Eulerian velocity; T 2 is the flux due to Stokes drift and T 1 + T 2 is called advection term. T 3 + T 4 + T 5 is the tidal pumping term which is related to the sediments exchange between water column and stream bed under periodic tidal dynamics. T 6 + T 7 is the vertical circulation term that is related to gravitational circulation and vertical profile of velocity and SSC [32,33]. The total net flux was computed by T1 + T2 + T3 + T4 + T5 + T6 + T7. The ratio indicates the proportion of each flux to the sum of all flux magnitudes, while the "-" only represents the landward direction of sediment transport.

Transport Mechanism of Suspended Sediments
Based on formula (1), the contributions of unit-width sediment transport intensity and major sediment transport factors to the total transport amount within the spring, intermediate and neap tide cycles were computed ( Table 2). The advection transport and tidal pumping transport were the major factors of suspended sediment transport in Hangzhou Bay, while the vertical circulation transport contributed little to the net sediment transport (Table 2). Additionally, the contribution of various factors of suspended sediment transport also varied among stations. For the stations H1 and H3, the tidal pumping transport had the greatest contribution, whereas for the station H2, advection transport was predominant ( Figure 6). Table 2. Suspended sediment transport rate and contribution of main transport terms. T 1 is the flux due to the non-tidal drift, the Eulerian velocity; T 2 is the flux due to Stokes drift and T 1 + T 2 is called advection term. T 3 + T 4 + T 5 is the tidal pumping term which is related to the sediments exchange between water column and stream bed under periodic tidal dynamics. T 6 + T 7 is the vertical circulation term that is related to gravitational circulation and vertical profile of velocity and SSC [32,33]. The total net flux was computed by T 1 + T 2 + T 3 + T 4 + T 5 + T 6 + T 7 . The ratio indicates the proportion of each flux to the sum of all flux magnitudes, while the "-" only represents the landward direction of sediment transport.

Station
Tidal Cycle  6. Components of total net suspended sediment fluxes at the three stations. T1 + T2 is the advection term, T3 + T4 + T5 is the tidal pumping term; and T6 + T7 is the vertical circulation term.

Advection Transport T 1 + T 2
The sediment transport by advection, reflecting the roles of Euler and Stokes residual currents in transporting the suspended sediments, was constituted by factors T1 and T2. Overall, the intensity of advection transport at the three stations increased gradually with the tidal cycle from neap to spring tides (Table 2). For example, at the station H1, the intensity of advection transport was 1.5 kg/(m·s) during spring tides, which was six times that of the neap tide intensity of 0.24 kg/(m·s). Moreover, the sediment transport by advection was positive at both H1 and H2, while was negative at H3, indicating an opposite direction of sediment transport.

Tidal Pumping Transport T 3 + T 4 + T 5
Within one tidal cycle, the variation of sediment concentration in the water body was attributed primarily to the periodic exchange of sediments between the water body and the bed. The sediment carrying capacity of tidal current changed with the strength of flood and ebb currents, leading to an asymmetric transport of bottom sediments within a tidal cycle during the exchange process ( Figure  3). Meanwhile, a certain phase difference was present in the changing processes of sediment concentration and flow velocity (Figure 3). During the high and low tide periods, the sediments exchanged between bed and water body could not be offset completely within a tidal cycle, thereby resulting in net sediment transport, called the tidal pumping transport, consisting of T3, T4 and T5.
Regarding the direction of tidal pumping transport (Table 2), it could be observed that the Figure 6. Components of total net suspended sediment fluxes at the three stations. T 1 + T 2 is the advection term, T 3 + T 4 + T 5 is the tidal pumping term; and T 6 + T 7 is the vertical circulation term.

Advection Transport T 1 + T 2
The sediment transport by advection, reflecting the roles of Euler and Stokes residual currents in transporting the suspended sediments, was constituted by factors T 1 and T 2 . Overall, the intensity of advection transport at the three stations increased gradually with the tidal cycle from neap to spring tides ( Table 2). For example, at the station H1, the intensity of advection transport was 1.5 kg/(m·s) during spring tides, which was six times that of the neap tide intensity of 0.24 kg/(m·s). Moreover, the sediment transport by advection was positive at both H1 and H2, while was negative at H3, indicating an opposite direction of sediment transport.

Tidal Pumping Transport T 3 + T 4 + T 5
Within one tidal cycle, the variation of sediment concentration in the water body was attributed primarily to the periodic exchange of sediments between the water body and the bed. The sediment carrying capacity of tidal current changed with the strength of flood and ebb currents, leading to an asymmetric transport of bottom sediments within a tidal cycle during the exchange process ( Figure 3). Meanwhile, a certain phase difference was present in the changing processes of sediment concentration and flow velocity (Figure 3). During the high and low tide periods, the sediments exchanged between bed and water body could not be offset completely within a tidal cycle, thereby resulting in net sediment transport, called the tidal pumping transport, consisting of T 3 , T 4 and T 5 .
Regarding the direction of tidal pumping transport (Table 2), it could be observed that the sediment transport direction was kept consistent at each tidal cycle for all three stations, which was always opposite to the direction of advection transport as well. Within a tidal cycle, the sediments were transported landward at the stations H1 and H2, whereas they were transported seaward at the station H3. The intensity of tidal pumping transport also presented an upward trend with the tidal cycle, indicating the continuously strengthened exchange of sediments between bed and water body in Hangzhou Bay with the increasing tidal range. The two nearshore stations H1 and H3 exhibited high intensity of tidal pumping transport. Particularly, the southern station H3 had a total transport intensity of 21.06 kg/(m·s) within a tidal cycle ( Table 2). This value was far greater than that at H2 in the central part of the research area, which was 1.91 kg/(m·s). This suggests that the sediment exchange between bed and water body is considerably more intense at the southern station H3 than at the central station H2.

Vertical Circulation Transport T 6 + T 7
Due to the vertical inconsistency in the direction and magnitude of residual currents and the uneven distribution of suspended sediment concentration vertically, the vertical circulation of sediment transport within tidal cycle was common. Given the shallow water depth, large tidal range and strong tidal current in Hangzhou Bay, the suspended sediments were well mixed vertically. Therefore, it can be concluded from Table 2 that compared to the advection and tidal pumping transport, the vertical circulation transport was smaller by 1-2 orders of magnitude, which made an extremely limited contribution to the net sediment transport.

Net Sediment Transport
Regarding the direction of unit width sediment transport within tidal cycles at the three stations, the northern station H1 transported sediments landward within tidal cycles, while the southern station H3 and the central station H2 transported sediments seaward ( Figure 6). In other words, on the north side of Hangzhou Bay, the suspended sediments flew into Hangzhou Bay from the outer sea. In contrast, the situation was entirely the opposite on the south side of Hangzhou Bay, where the sediments were transported from the bay to the outer sea. Regarding the amount of sediment transport at the three stations, H3 exhibited the highest intensity of sediment transport within tidal cycles, with a total sediment transport of 10.43 kg/(m·s) ( Table 2). For the station H1, its total sediment transport within tidal cycles was 2.92 kg/(m·s). Meanwhile, H2 showed the least total sediment transport within tidal cycles, with a value of 1.70 kg/(m·s) only ( Figure 6 and Table 2).

Contribution Rates
From the perspective of the major contribution of sediment transport factors to the total transport amount, the advection and tidal pumping transports were predominant in the Central Hangzhou Bay, while the vertical circulation transport made little contribution ( Figure 6 and Table 2). At the station H1, the contribution rate of the vertical circulation transport within tidal cycles was merely 1%. The advection and tidal pumping transports constituted the major factors of suspended sediment transport in the Central Hangzhou Bay. Nevertheless, the contribution rates of various suspended sediment transport factors varied among the stations.
At the central station H2, advection transport played a dominant role and was the primary contributing factor of sediment transport, followed by the tidal pumping transport ( Table 2). The contribution of advection transport to sediment transport within tidal cycles at the station H2 was 1.8 times that of the tidal pumping transport. In contrast, at the northern station H1 and the southern station H3, the tidal pumping transport prevailed over the advection transport to become the primary contributing factor of sediment transport. The contribution of tidal pumping transport to sediment transport within tidal cycles was 2.4 times and 2.2 times that of the advection transport, at the station H1 and H3 respectively, indicating that the sediment exchange between bed and water body at the nearshore stations H1 and H3 is considerably more intense than that at the central station H2.

Distribution of Grain Size Parameters µ, δ and S k of Bed Sediments
The mean grain size represents the characteristics of the bed sediments directly, which is able to reflect the strength of dynamic environment. The mean grain size µ varied between 4 and 8 Φ (Figure 7a). Larger grain sizes were observed at the southwest corner near the Sanbei Shoal, where the water depth was the shallowest, along with the northeast corner of the Jinshan Trough, where the water depth was the deepest in the reach area. The sorting coefficient δ varied between 1.4 and 2.2 ( Figure 7b). Similar to the distribution of µ, the value of δ was smaller near the banks, whereas it was larger in the middle. In contrast, the distribution of skewness S k displayed an opposite trend with that of µ. The largest S k value of over 2 was present at the north and southwest corners, while in the middle areas, the value was less than 1 (Figure 7c).
Water 2020, 12, x FOR PEER REVIEW 10 of 16 water depth was the deepest in the reach area. The sorting coefficient δ varied between 1.4 and 2.2 ( Figure 7b). Similar to the distribution of , the value of δ was smaller near the banks, whereas it was larger in the middle. In contrast, the distribution of skewness S displayed an opposite trend with that of . The largest S value of over 2 was present at the north and southwest corners, while in the middle areas, the value was less than 1 (Figure 7c).

Sediment Transport Trends
The Gao-Collins grain size trend analysis model was utilized to explore the sediment transport trends in the study area, where the characteristic distance was set at 5360 m (maximum sampling interval). The vectors for sediment transport trends in the Central Hangzhou Bay (Figure 8) were

Sediment Transport Trends
The Gao-Collins grain size trend analysis model was utilized to explore the sediment transport trends in the study area, where the characteristic distance was set at 5360 m (maximum sampling interval). The vectors for sediment transport trends in the Central Hangzhou Bay (Figure 8) were derived based on the results of Figure 7. The direction of the vectors represents the sediment transport direction, while the length of the vector represents the significance of grain size trend, rather than representing the magnitude of sediment transport rate.
Based on Figure 8, the surface sediments on the north side of Hangzhou Bay showed a transport trend from the outer sea to the bay head. The situation on the south side of the bay was the opposite, where a transport trend from the bay head to the outer sea was presented.

Generation of Soft Mud Layer in Hangzhou Bay
Structural stratification of bottom sediments is the primary cause of the occurrence of two crests of sediment concentration during a tidal cycle in Central Hangzhou Bay. During the turn of the tidal current, the flow velocity declines and the flow turbulence weakens, resulting in the settling of suspended sediments [41,42]. Given the sorting function of the tidal current, the coarse-grained sediments settle faster, forming a layer of somewhat consolidated soft mud (soft mud layer) on the seabed. In contrast, due to the flocculation, the fine-grained sediments with a high viscosity settle slowly. Consequently, they would be released on the surface of the soft mud layer, forming an unconsolidated viscous layer that is called the fluid mud layer. Compared to the soft mud layer, the grain size of the fluid mud layer is finer, the viscosity higher and the sediment bulk density smaller. Thus, the fluid mud layer is more easily initiated and resuspended. The fluid mud phenomenon has been found to be present at the estuaries of the Gironde [43], Humber [44] and Changjiang rivers [45,46]. Moreover, Shi [42] also observed the presence of fluid mud at Hangzhou Bay. Therefore, according to the variation of SSC within the tidal cycles (Figure 3), we imply that it is highly

Generation of Soft Mud Layer in Hangzhou Bay
Structural stratification of bottom sediments is the primary cause of the occurrence of two crests of sediment concentration during a tidal cycle in Central Hangzhou Bay. During the turn of the tidal current, the flow velocity declines and the flow turbulence weakens, resulting in the settling of suspended sediments [41,42]. Given the sorting function of the tidal current, the coarse-grained sediments settle faster, forming a layer of somewhat consolidated soft mud (soft mud layer) on the seabed. In contrast, due to the flocculation, the fine-grained sediments with a high viscosity settle slowly. Consequently, they would be released on the surface of the soft mud layer, forming an unconsolidated viscous layer that is called the fluid mud layer. Compared to the soft mud layer, the grain size of the fluid mud layer is finer, the viscosity higher and the sediment bulk density smaller. Thus, the fluid mud layer is more easily initiated and resuspended. The fluid mud phenomenon has been found to be present at the estuaries of the Gironde [43], Humber [44] and Changjiang rivers [45,46]. Moreover, Shi [42] also observed the presence of fluid mud at Hangzhou Bay. Therefore, according to the variation of SSC within the tidal cycles (Figure 3), we imply that it is highly connected to the generation of the fluid mud.
At the initial stage of flood or ebb tide following a turn of tidal current, the flow velocity begins to increase, and the erosive action of water body acts initially on the fluid mud layer (Figure 3). Due to high viscosity and unconsolidated characteristics, the fluid mud layer can be initiated and resuspended at low flow velocities. Thus, the erosion and resuspension of fluid mud layer within 1 to 2 h of the initial stage of flood or ebb tide form the first crest of SSC over the tidal cycle process. After the rapid erosion of the fluid mud layer, the soft mud layer below is exposed. Compared to the fluid mud layer, the erosion resistance of the soft mud layer increases abruptly, leading to the decrease of erosion rate, i.e., limiting the increase of SSC. Accordingly, the flow velocity increases, while the SSC decreases, reaching the trough 1 h later. Afterwards, with the further increase of flow velocity, the erosion rate of the soft mud layer increases, reaching the maximum at the moment of maximum flood or ebb flow. However, the second crest of SSC generally exits 1-2 h later, rather than at the maximum flow velocity. This is because although the flow is decreasing, the bed keeps being eroded due to the high flow velocity, which is called phase shift between the flow velocity and the bottom shear stress [47,48]. Therefore, the erosion and resuspension process of the soft mud layer is the primary contributing factor to the M-shaped hydrograph of the SSC.
Additionally, the second crest is found to be markedly larger than the first crest during the spring and intermediate tides, suggesting greater erosion and resuspension of the soft mud layer than the fluid mud layer. In contrast, during the neap tide, no significant difference is found between the first and second crests, suggesting approximative erosion and resuspension between the soft mud layer and the fluid mud layer. Xie et al. [13] revealed that due to the coastal embankment and sediment accumulation, the hydrograph of tides in the inner bay is changed; for example, the flood dominance is increased. Thus, the changed tides would further affect the suspended sediment transport, since it was found to be highly related to the erosion and deposition of the fluid mud layer. Due to data limitations, in this paper, it is difficult to quantify the effect of changed tides on the sediment transport attributed to coastal embankment, although it is confirmed to be a significant factor on sediment transport and morphology evolution. Hence, further work needs to be carried out on this, in order to provide guidance for decision-making process of major engineering projects in Hangzhou Bay.

Sediment Transport Mode in Hangzhou Bay
Through analyzing the suspended sediment transport flux and sediment grain size trend, it is found that Hangzhou Bay presents a "north-landward and south-seaward" trend, a "C"-shaped sediment transport mode macroscopically. To be specific, the sediments are transported from the bay mouth to the bay head on the north side of the bay and from the bay head to the bay mouth on the south side. The sediment transport by tidal pumping at H1 and H3 is the main contributing factor of sediment transport. This is attributed primarily to the "funnel-shaped" geomorphology of Hangzhou Bay, as well as the large tidal range and shallow water depth (the spring tidal range at station H3 is even greater than the mean water depth). Moreover, Hangzhou Bay has a semidiurnal tidal feature, with frequently alternating high and low tides and strong currents (the maximum mean vertical velocity is observed at the station H3 in this study, with a value of 3.25 m/s). Moreover, the grain size composition of sediments is also rather fine. Thus, the bottom sediments can be easily initiated and resuspended in the water body at the time of rapid flow. Even if the sediments fall to the bed during slack current, they will be uplifted by the following tidal currents without being compacted, thereby resulting in repeated transport of sediments.

Relationship between the Sediment Transport Mode and the Overall Erosion and Deposition in Hangzhou Bay
According to the study of Chen et al. [19], Hangzhou Bay has historically exhibited an evolution pattern of a scoured, receding north shore, and a silted south shore. The current work considers that the "C"-shaped sediment transport mode of Hangzhou Bay decides its evolution pattern, i.e., south shore silting and north shore scouring.
The north shore of Hangzhou Bay closely adjoins the Changjiang River Estuary. Through the comparative analysis of heavy mineral contents, Cao et al. [49] and Liu et al. [20] found that the sediments in Hangzhou Bay came mainly from the Changjiang River Estuary. During the southward diffusion, the runoff from the Changjiang River into the sea entered Hangzhou Bay initially along its north shore with the flood tide process of the bay. At this point, during the advancement of the water body with low sediment concentration into the bay along the north shore, the water flow was concentrated due to the narrowing of the bay mouth. The flow velocity increased and the sediment carrying capacity of the water body was enhanced during flood tides. Moreover, the bottom sediments initiated quickly, forming a water body with high sediment concentration, and transported towards the bay head resulting in scouring on the north bay shore. In ebb tide periods, during the recession of water body with high sediment concentration at the bay head towards the outer sea, when the ebb tide water with strong sediment carrying capacity and high sediment concentration flew through the Andong Shoal on the south bay shore, the entrance of the bay widened suddenly and the water depth shallowed due to the presence of the Andong Shoal. The ebb current was diffused and the sediment carrying capacity decreased. Thus, the suspended sediments were deposited, resulting in the continuous siltation on the south shore of Hangzhou Bay. In summary, the "C"-shaped transport mode and the narrow-head wide-mouth geomorphology determine the evolution pattern of silted south shore and scoured north shore for Hangzhou Bay.

Mechanism of Hydrodynamics on Controlling Sediment Transport in Hangzhou Bay
It is universally acknowledged that in bays, the shear forces resulting from tidal currents and waves act on the bed surface to suspend and settle sediments, which is the primary dynamic condition leading to the exchange between bed material and suspended sediments [50]. A study carried out by Nidzieko and Ralston [51] stated that in estuary bays without significant runoff impact, the asymmetry of flood and ebb currents determines the net transport direction of masses, which is an important dynamic mechanism for net mass transport. In the present study, the statistical method-based "skewness" (γ) proposed by Nidzieko and Ralston [51] is used to analyze the asymmetry of tidal currents in Hangzhou Bay, which can be computed as follows: where the m-th origin moment is defined as: In the formula, u i denotes the flow velocity at the i-th moment; and N denotes the total number of flow velocity samples. We set the ebb current as positive, and the flood current as negative. Thus, if γ > 0, it indicates that the ebb current is dominant, and if γ < 0, it indicates that the flood current is dominant.
The skewness values of flow velocity during the spring, intermediate and neap tides are calculated according to the formula (2) for the stations H1, H2 and H3 in the Central Hangzhou Bay, as shown in Figure 9. Clearly, the skewness values of the station H1 were negative, indicating that the north side of Hangzhou Bay is dominated by the flood current. For the station H2, the flow velocity is positively skewed during the spring and intermediate tides, whereas it is negatively skewed during the neap tide, without showing a unified predominant direction of tidal currents. Nevertheless, the absolute values of skewness are small, suggesting a mild degree of asymmetry in the flood and ebb currents at H2. As for the station H3, the skewness values are positive, indicating that the south side of Hangzhou Bay is dominated by ebb current. This finding is consistent with the result of the "C"-shaped transport mode derived above in this study. Hence, this paper argues that the asymmetry of flood and ebb currents on the south versus north sides of Hangzhou Bay is the main dynamic mechanism leading to the overall "C"-shaped transport mode of sediments in the bay. current is dominant.
The skewness values of flow velocity during the spring, intermediate and neap tides are calculated according to the formula (2) for the stations H1, H2 and H3 in the Central Hangzhou Bay, as shown in Figure 9. Clearly, the skewness values of the station H1 were negative, indicating that the north side of Hangzhou Bay is dominated by the flood current. For the station H2, the flow velocity is positively skewed during the spring and intermediate tides, whereas it is negatively skewed during the neap tide, without showing a unified predominant direction of tidal currents. Nevertheless, the absolute values of skewness are small, suggesting a mild degree of asymmetry in the flood and ebb currents at H2. As for the station H3, the skewness values are positive, indicating that the south side of Hangzhou Bay is dominated by ebb current. This finding is consistent with the result of the "C"-shaped transport mode derived above in this study. Hence, this paper argues that the asymmetry of flood and ebb currents on the south versus north sides of Hangzhou Bay is the main dynamic mechanism leading to the overall "C"-shaped transport mode of sediments in the bay.

Conclusions
Based on the 2013 field survey data of hydrology, suspended and bottom sediments in the Central Hangzhou Bay, this study explored the dynamic mechanism of suspended sediments, as well as the migration trends of bed sediments.
The M-shaped hydrograph of suspended sediment concentration primarily attributes to the generation of both soft mud and fluid mud layers. Meanwhile, the erosion and resuspension of the fluid mud layer is the first-order control on the second peak of the M shape. Xie et al. [13] proved that the coastal embankment and sediment accumulation changed the characteristics of tides in Hangzhou Bay. Therefore, the changed tides would affect the shape of the suspended sediment concentration hydrograph, and thereby the suspended sediment transport and morphology evolution. Hence, further work needs to be completed on this, in order to provide guidance for decision-making processes of major engineering projects in Hangzhou Bay.
Hangzhou Bay presents a "north-landward and south-seaward" trend, a "C"-shaped sediment transport mode, due to the unbalance of flood and ebb currents laterally-the flood current prevalence on the north and the ebb current prevalence on the south. Coupled with the narrow-head, wide-mouth geomorphology, Hangzhou Bay remains evolving by south shore silting and north shore scouring.

Conflicts of Interest:
The authors declare no conflict of interest.