3.1. Verification of Inversion Results Based on Numerical Modeling and In Situ Measurements
To verify the accuracy of the inverted flow field, we initially conducted a harmonic analysis of the inverted flow field dataset and derived the tidal ellipse for the M
2 constituent. The dependability of the MCC inverted flow field was evaluated by comparing the tidal ellipses of the inverted flow field and the model flow field and calculating the root-mean-square error of the long and short axes, and the inclination angle between the two. The comparative results reveal that the M
2 tidal ellipses of the inverted flow field and the model flow field exhibit substantial consistency in the sea area south of the Yangtze River Estuary (
Figure 4a). However, the consistency is comparatively poor in the Subei Shoal area. The calculated errors reveal that the average error of the long axis is 0.0335 m/s, the average error of the short axis is 0.0276 m/s, and the average error of the inclination angle is 6.89°. Moreover, among the data points with good consistency, the accuracy in the shelf area is superior to that in the nearshore area, which might be attributable to the fact that the vertical sediment settling motion of nearshore water is more vigorous, and the variation in TSS is not dominated by the horizontal flow. We performed a spectrum analysis for one month’s worth of tidal hourly elevation data, and the results (
Figure 4b) showed two distinct peaks with frequencies of 2.2401 × 10
−5 Hz and 2.3148 × 10
−5 Hz, with a magnitude of 366.8782 and 102.1478, corresponding to periods of 12.4 h and 12.0 h. We consider that these two tidal components are M
2 and S
2, which are the main components of tidal currents in the East China Sea and have a significant impact on the distribution of surface flow fields.
In addition, we compared the mean flow velocity of the inverted flow field in the coastal area on 14 February 2017, using the observed tidal elevation changes measured at a local tide gauge station. The findings reveal that the two exhibit good synchronicity in their changes (
Figure 4c). At the moment when the tidal elevation change was most drastic, the surface flow velocity also reached its maximum value. All of the above validation outcomes imply that the MCC algorithm in this study successfully inverts the ECS surface flow field, with good accuracy at different time scales, and can effectively represent the true flow conditions of the ECS surface water.
3.2. Pattern and Seasonal Variations of the Surface Residual Currents in the East China Sea
The comparative validation results demonstrate the reliability of the inverted flow field, thereby enabling analysis of circulation patterns and temporal variations in the ECS. To obtain the circulation patterns of the ECS, it is first necessary to remove the tidal current component from the inverted flow field. Therefore, we subtract the concurrently predicted tidal current data from TPXO8 in the inverted flow field, and the remaining residual current field represents the circulation of the ECS. By following the aforementioned methodology, we acquired the residual flow data pertaining to the ECS during the period ranging from 2013 to 2019. Afterward, by averaging the obtained data, we were able to derive the climatological (
Figure 5a) and seasonal (
Figure 6a) circulation distribution maps of the ECS.
Figure 5a depicts the climatic flow field within the ECS, while
Figure 5b portrays the spatial distribution of the number of velocity vectors that were utilized to compute the average. We can see that the quantity of velocity vectors is greater in the coastal regions of Zhejiang and Fujian provinces, as well as in the offshore zones situated in the western parts of Taiwan. This phenomenon can be chiefly ascribed to the heightened turbidity levels that prevail within these offshore areas, thereby rendering it more convenient for the algorithm to detect and track the motions of water masses. The sea surface flow field’s climatic state divulges that the ECS’s surface circulation is mainly composed of three prevailing current systems. The first current system, situated at the southeastern continental slope of the ECS, is the Kuroshio Current (KC), boasting the widest flow width and highest flow velocity [
21]. It courses northwards along the eastern coast of Taiwan, enters the ECS at the northeastern part of Taiwan Island, and then takes a northeastward direction along the continental slope toward Japan. The KC’s average velocity is approximately 0.86 m/s, and its maximum flow velocity can reach up to 1.42 m/s. As a robust western boundary current of the North Pacific, the KC exercises a significant influence on the circulation system of the ECS continental shelf area. The second oceanic current system is the Taiwan Warm Current (TWC), which flows northwards from the Taiwan Strait into the ECS throughout the year [
22]. The TWC primarily moves northwards along the 50–100 m isobaths, with high-turbidity coastal waters on its left and low-turbidity continental shelf water masses on its right. Upon reaching 27°N, the TWC bifurcates into two branches, with one branch moving northwards along the coast and the other turning northeastwards to merge into the KC. The TWC significantly impacts the circulation structure of the nearshore regions of the western ECS. The third current system is the Changjiang Diluted Water (CDW), which carries a substantial amount of sediment and flows into the ECS from the Changjiang River estuary. It courses northeastward initially, then turns southeastward around 124°, and wields a crucial influence on the ecological and dynamic systems near the Changjiang River estuary [
23].
Figure 6 and
Figure 7, respectively, showcase the seasonal variation in the inverted ECS surface current field and the seasonal distribution of the inverted sea current vector quantity. It is noticeable that the inverted sea current vector quantity from 2013 to 2019 exhibits significant seasonal differences: the summer season has the highest number of effective sea current vectors (
Figure 7b) at 84,248; the winter season has the least at 32,765 (
Figure 7d); while the spring and autumn seasons have 55,825 and 52,248 effective sea current vectors (
Figure 7a,c), respectively. This indicates that besides the gradient magnitude of the selected tracer, cloud occlusion is also a major factor affecting the inversion performance of the MCC algorithm [
24].
The seasonal distribution map of the reversed flow field exhibits pronounced seasonal fluctuations in the circulation system of the ECS (
Figure 6). The intensity, or flow velocity of the Kuroshio Current (KC) exhibits a summer maximum and a winter minimum. During the summer, the average flow velocity is approximately 0.9 m/s, with maximum velocities exceeding 1 m/s. In contrast, the average flow velocity during the winter is approximately 0.6 m/s. Additionally, during the summer, the main axis of the KC moves away from the continental slope of the ECS (
Figure 6b), while during the winter, the surface seawater is invaded by a large volume of water from the ECS shelf under the influence of the northward monsoon (
Figure 6d). The spring and autumn seasons show transitional states between the winter and summer conditions (
Figure 6a,c). Similar to previous studies, the Taiwan Warm Current (TWC) flows northward throughout the year between the 50–100 m isobaths, but its northward intensity varies seasonally. During the summer, the TWC bifurcates around 28.5°N, with one branch turning eastward to merge with the KC, while the main branch continues northward to the mouth of the Changjiang River beyond 30°N (
Figure 6b). In the winter, the TWC bifurcates around 27.5°N, with one branch merging with the KC, while the main branch continues northward to the eastern offshore of Zhoushan beyond 29°N before turning eastward (
Figure 6d). The Changjiang’s freshwater discharge converges with the continental shelf water mass behind its estuary and creates the Changjiang Diluted Water (CDW), which subsequently flows northeastwards toward Jeju Island. During the spring and summer seasons, the CDW veers southeastwards beyond the 124°E, streaming toward the adjacent waters of the KC (
Figure 6a,b). In autumn, the CDW maintains its eastward movement (
Figure 6c), while in winter, it is swayed by the south branch of the Yellow Sea circulation system and the northward Yellow Sea warm current east of 125°E, compelling it to deflect northeastwards (
Figure 6d). It is worth noting that during the winter, a branch of the CDW flows along the coastline southwards beyond its estuary (
Figure 6d), forming the Zhe-Min Coastal Current, which moves in a southerly direction due to the influence of the northerly monsoon (although it may be difficult to observe using satellite imagery because of its proximity to the coastline). In addition, satellite observations have successfully identified the seasonal characteristics of the Zhejiang Coastal Current (ZJCC): during autumn, it flows southward under the influence of the north wind (
Figure 6c), while in summer, it flows northward due to the impact of the south wind (
Figure 6b).
3.3. Diurnal Variability and Mechanisms of TSS in the Zhejiang Coastal Front
The Zhejiang Coastal Front (ZJCF), situated along the southeastern coast of China, serves as an intermediary zone between two oceanic currents—the nearshore Zhejiang Coastal Current (ZJCC), characterized by its low salinity and high turbidity, and the offshore Taiwan Warm Current (TWC), known for its high salinity and low turbidity. Due to the barrier effect of the oceanfront, highly turbid nearshore water masses are restricted in their ability to be transported across the front, resulting in their accumulation primarily along the coastal areas. As coastal ocean development continues to grow, increasing attention is being paid to marine environmental issues. The distribution and variation in suspended sediments and pollutants in nearshore waters have become new critical concerns in both physical oceanographic research and marine engineering projects. The transport, deposition, and resuspension of suspended sediments under the influence of ocean dynamics, such as ocean currents, waves, and monsoons, can cause changes in the topography of the seafloor. This, in turn, affects shoreline evolution, channel dredging, and the stability of underwater slopes. Moreover, fine particles of suspended sediment possess a strong adsorption capacity and often act as carriers of heavy metals, nitrogen, phosphorus, and other pollutants, influencing the accumulation and movement of contaminants in nearshore waters. Therefore, the distribution and variation in suspended sediments are an indispensable and integral component of sedimentation, material flux, and the physical environment of coastal waters. In this research, we initially chose a series of eight consecutive GOCI TSS satellite images captured on an unclouded day on 3 August 2013. Utilizing the 70 m isobath as the outermost boundary, we computed and scrutinized the intraday high-frequency changes in the spatially averaged TSS within the Zhejiang Coastal Front. Additionally, we conducted a diagnostic analysis of the underlying mechanisms of TSS variations based on the flow fields concurrently retrieved during the same period.
The results indicated that there is a diurnal variation in the spatial mean of surface TSS within the frontal zone along the coast of Zhejiang within the 70 m isobath (
Figure 8h): an increasing trend from 8:00 to 12:00, followed by a decreasing trend from 12:00 to 15:00. The concurrently inverted surface current field is well correlated with the observed tidal elevation from the local tide gauge stations (
Figure 8i), and it exhibits a primarily rotating tidal current pattern over time: from 8:00 to 12:00 (
Figure 8a–d), the surface current field moves offshore, then, from 12:00 to 14:00 (
Figure 8e–f), the current field changes to a northward alongshore flow, and from 14:00 to 15:00 (
Figure 8g), the current field further shifts to an onshore direction. Indeed, as evident from
Figure 8h, the alterations in TSS demonstrate a strong negative correlation with the alterations in tidal elevation, with a correlation coefficient of R = −0.83. This suggests that the alterations in TSS within the frontal zone along the coast of Zhejiang are predominantly influenced by the local tidal currents.
The variations in surface TSS are usually accompanied by both vertical and horizontal dynamical processes. To elucidate the main physical processes governing variations in surface TSS in the front zone along the Zhejiang coast, we conducted a diagnostic analysis based on the advection-diffusion equation for TSS [
25], which is given below:
where
is the TSS concentration;
is the surface horizontal flow velocity components
and
;
,
, and
represent anomalies of
,
, and
;
is the diffusion flux vector in the horizontal direction;
is the horizontal advection term;
is the horizontal diffusion term; and
represents all of the other terms related to the vertical, namely the vertical convection term, the vertical diffusion term, and the settling term.
The diagnostic results (
Figure 9) indicated that the magnitude for the hourly variations in TSS was 10
−5, while the magnitude for the advection term was 10
−6, the magnitude for the horizontal diffusion term was 10
−7, and the magnitude for the vertical processes was 10
−5. Thus, it can be inferred that the vertical processes, which involve vertical convective and diffusive transport of water masses, sediment settling, and resuspension, play a crucial role throughout the entire process of TSS changes. This is also evident from the degree of conformity between the two lines as they vary with time (The blue line and the purple line in
Figure 9). In comparison to the contribution of vertical processes, the contribution of the diffusion term is almost negligible, while the contribution of the advection term is of the same order as that of the vertical processes only after 12:00.
Figure 10 displays the spatial distribution of the absolute values of hourly variations in TSS, horizontal advection term, horizontal diffusion term, and vertical term from 10:00–11:00 and 11:00–12:00. It can be observed that the hourly variations in TSS exhibit a stable band-shaped distribution parallel to the coastline, with decreasing intensity offshore, indicating that water depth influences TSS changes. The distribution of horizontal advection and vertical terms is similar to that of hourly variations in TSS, and the intensity of the vertical term is greater than that of the horizontal advection term. The horizontal diffusion term is of the smallest magnitude and almost negligible. Thus, it is evident that vertical dynamic processes dominate the variations in surface TSS in the local area, with some contribution from horizontal advection, while the contribution from horizontal diffusion can be ignored.
The above analysis indicates that the vertical processes of water masses are the primary cause of the surface TSS variations along the Zhejiang Coastal Front. Previous research has suggested that short-term TSS variations near the East China Sea are dominated by bottom shear stress and tidal mixing [
26,
27]. As previously mentioned, the vertical term includes convective terms, vertical diffusion, sedimentation, and resuspension of sediments. In order to investigate how convective terms, i.e., tidal-induced mixing, affect TSS variations, we introduced the Simpson–Hunter index [
28], which represents the stability of the seawater stratification:
, where H is the water depth, and
is the surface horizontal velocity. A larger k value indicates greater seawater stratification stability and weaker vertical convection.
Figure 11 reveals that during the early to mid-ebb tide period (8:00–12:00), there is a negative correlation between the TSS changes and k, with an increase in the TSS variation rate corresponding to a decrease in k. However, after 12:00, during the slack tide period, the TSS variations appear to be unrelated to tidal mixing. At this point, it is speculated that the vertical movement is mainly due to sedimentation and the resuspension of sediments. Furthermore, during the slow water flow rate of the slack tide stage, the slow movement of the water body implies a longer response time for sedimentation, which generally lags by approximately 2 h.