Quantitative Assessment of Local Siltation Dynamics in Multi-Anabranching River System: Case Studies of Representative Port in the Lower Yangtze River and Engineering Interventions

Abstract

The Ma’anshan section of the lower Yangtze River features a complex multi-anabranching system, where the river divides into several branches around mid-channel sandbars, with distinct point bars alternately developing along both banks. Within this morphologically active system, Zhengpu Harbor suffered severe operational disruptions by accelerated siltation at its approach channel, primarily due to its vulnerable location downstream of the expanding Niutun River point-bar on the left bank. To systematically diagnose the mechanisms of siltation, this study integrates multi-method investigations: decadal-scale morphodynamic analysis using long-term bathymetric surveys, numerical modeling to quantify engineering impacts on flow dynamics, and multiple linear regression analysis for the contributions of key influencing factors. The result identifies three primary drivers of siltation, collectively responsible for 70% of the sediment accumulation, including the rightward shift of the thalweg in the Ma’anshan left branch, reduced flow diversion of the left Branch of Central bar, and the expansion of the Niutun River point bar. River engineering structures, such as bridges, contribute approximately 12%, while changes in upstream flow-sediment supply account for approximately 18%. To mitigate siltation at Zhengpu Harbor’s approach channel, this study proposes targeted engineering interventions to enhance local hydrodynamic conditions. The spur dikes were designed to enhance the morphological stabilization of the Central bar head to regulate flow distribution. A diversion channel could also be excavated at the tail of the Niutun River shoal, and emergency dredging was recommended at the harbor front. Numerical modeling indicates that these measures will increase flow velocity by over 0.1 m/s at the harbor front, mitigating the siltation situation. The study concludes that the proposed engineering measures can reduce annual siltation by approximately 30% under normal-year hydrological conditions, demonstrating their feasibility in mitigating siltation trends in multi-anabranching river systems. This research provides a reference for addressing siltation issues in harbors within complex anabranching river systems.

1. Introduction

Anabranching river systems, widely distributed globally in alluvial plains, exhibit a phenomenon where main channel shifting triggers chain reactions among upstream and downstream shoals. When the main flow direction changes, it directly scours convex bank shoals, causing their recession. Simultaneously, the flow distribution ratio in adjacent branches adjusts, gradually weakening the previously dominant flow dynamics. This alteration further impacts the flow conditions around marginal shoals. Such systems commonly face challenges of sediment transport and siltation, which threaten navigation, flood control, and ecological health [1].

Studies have shown that multi-anabranching patterns in rivers such as the Columbia River [2] and those on the Qinghai-Tibet Plateau [3] primarily result from upstream sediment overloading or island dynamics, emphasizing system disequilibrium under natural conditions. Banasik (2021) [4] further confirms that the sediment supply-demand imbalance serves as a core driver of anabranching evolution. Most prior studies on harbor siltation in braided rivers have relied on qualitative methods. For example, Guo (2023) [3] analyzed anabranching channels on the Qinghai-Tibet Plateau and identified island dynamics as crucial for morphologic stability. However, their research relied on qualitative observations, thus failing to quantify the relative contributions of hydrologic and geomorphic factors to siltation. Existing studies on harbor siltation in branching reaches mainly focus on qualitative analysis [5,6,7], lacking quantitative methods to assess contributing factors. Moreover, case studies on targeted engineering measures for multi-causal siltation are scarce. The innovation of this study lies in the quantitative analysis of siltation driving factors through a multiple regression model.

The middle and lower Yangtze River predominantly exhibits branching patterns, with numerous mid-channel bars and shoals along its banks [8,9]. The operational safety of harbors and other infrastructure downstream of shoals is closely linked to the evolution of bars, branches, and shoal-channel dynamics. The Ma’anshan reach, spanning approximately 44 km, is a representative multi-branching section. Upstream of Dongliang Mountain lies the Chenjiazhou branching segment of the Wuyu reach; from Dongliang Mountain to the head of Xiaohuangzhou is the Jiangxinzhou branching segment; and downstream to the head of Xinshengzhou is the Xiaohuangzhou branching segment. The Dongliang–Xiliang Mountain node has limited control, resulting in strong interactions between upstream and downstream branches and significant temporal variations in diversion ratios [10,11,12].

Numerous studies have examined the riverbed evolution, influencing factors, and remediation measures in the Ma’anshan reach using hydrological and topographic data. Hao Jieyu (2020) [13] analyzed historical changes in the left and right branches, including shoal-channel dynamics and flow-sediment diversion ratios. Pan Qingshen (2011) [14] summarized the reach’s evolutionary characteristics and remediation experiences. Zhengpu Harbor, located on the north bank of the left branch in Ma’anshan, Anhui, is a national first-class open harbor and a key deep-water hub on the Yangtze River, situated in the core area of the Ma’anshan Zhengpu New District Modern Industrial Park [15]. However, since its Phase I operation in 2014, severe siltation has occurred at the harbor front, with cumulative deposition exceeding 5 m by 2021, reducing berthing width and causing multiple grounding incidents.

Since its Phase I operation in 2014, Zhengpu Harbor has experienced severe siltation at the harbor front, with cumulative deposition exceeding 5 m by 2021. This has resulted in the narrowing of berthing and maneuvering spaces and insufficient navigable water depth, rendering it unsafe for vessels exceeding 5000 DWT to berth and triggering multiple grounding incidents. Such conditions have increased port operation-maintenance costs and logistics delay risks. Ecologically, the sedimentation threatens ecological balance by altering aquatic habitats and nutrient cycling within the multi-anabranching river system. Modeling is indispensable for unraveling complex sediment dynamics in braided systems [16], in which field observations alone cannot capture transient flow-sediment interactions. This study first adopts field measurement data analysis to systematically identify key influencing factors affecting local siltation dynamics in multi-anabranching river systems. Subsequently, a multiple linear regression method is applied to quantify the contribution rates of each driving factor, while a two-dimensional hydrodynamic model is used to simulate complex water-sediment interactions under different scenarios, providing a scientific basis for siltation engineering interventions.

2. Data Sources and Research Methods

2.1. Data Sources

The study collected measured topographic, hydrological, and engineering data from the Ma’anshan reach of the Yangtze River between 1999 and 2022. Specifically, underwater topographic survey data from multiple years (1999, 2002, 2012, 2016, 2020, and 2022) were obtained, primarily sourced from the Yangtze Estuary Hydrological Bureau of the Changjiang Water Resources Commission and the Yangtze River Waterway Bureau. The topographic data were uniformly referenced to the 1954 Beijing Coordinate System and the 1985 National Elevation Datum, and a digital elevation model was established using Surfer 8 software [17].

The study collected daily discharge and sediment concentration data from the Datong Station for the period 2002–2022, sourced from the Hydrology Bureau of the Changjiang Water Resources Commission, Ministry of Water Resources. The daily discharge and sediment concentration data were statistically averaged on a monthly basis to derive the average flood season discharge and sediment concentration. Additionally, the study compiled records of major water-related engineering projects implemented in the Ma’anshan reach between 2002 and 2022, covering river regulation, navigation channel improvements, and cross-river bridges.

2.2. Research Methods

2.2.1. Multiple Regression Analysis

Changes in the river channel near the harbor area are influenced by multiple factors, including upstream water and sediment conditions, upstream river regime, local river morphology, and water-related engineering projects. To quantitatively analyze the relative contributions of these factors to long-term riverbed erosion and deposition, this study employed multiple linear regression analysis [18,19,20,21].

The regression analysis in this study adheres to classical linear regression assumptions, positing a linear relationship between siltation degree (characterized by channel volume below 6 m at Zhengpu Harbor front) and independent variables categorized into flow-sediment conditions, upstream river regime changes, local channel dynamics, and engineering impacts. We further assume that the error terms of individual observations are independent of each other. Multiple linear regression analyzes the influence of independent variables on a dependent variable. Assuming the dependent variable is $y$ and the independent variables are ${\chi }_1$, ${\chi }_2$, ${\chi }_3$, …, ${\chi }_k$, the model is expressed as:

$$y={\beta }_0+{\beta }_1{\chi }_1+{\beta }_2{\chi }_2+{\beta }_k{\chi }_k+\epsilon$$

(1)

where $\epsilon$~N(0, σ²) is the random error, and ${\beta }_0,{\beta }_1,{\beta }_2,\dots {\beta }_k$ represents the regression coefficients. The estimated regression coefficients were obtained using the least squares method, yielding the following equation:

$$\widehat{y}=\widehat{{\beta }_0}+\widehat{{\beta }_1}{\chi }_1+\widehat{{\beta }_2}{\chi }_2+\widehat{{\beta }_k}{\chi }_k$$

(2)

The adjusted coefficient of determination ${\overline{R}}^2$ was used to evaluate the goodness of fit of the regression equation. A value closer to 1 indicates that the equation better explains the dependent variable.

Further, the variables were standardized, and a standardized regression equation was derived. The absolute values of the standardized regression coefficients (β′) were compared to determine the relative influence of each factor.

The standardized coefficients (β′) were further processed to calculate the weight of each independent variable in the regression equation based on the cumulative sum of squares:

$${\partial }_k={\displaystyle \frac{{{\beta }_k^{\prime }}^2}{\sum {{\beta }_k^{\prime }}^2}}$$

(3)

This weight serves as a reference for assessing the contribution of each influencing factor to long-term riverbed erosion and deposition.

2.2.2. Numerical Model Calculation

As the core modeling tool in this study, Delft3D 2020.2 is primarily used to quantitatively simulate the hydrodynamic processes and sediment transport mechanisms in multi-anabranching river systems. By constructing a two-dimensional hydrodynamic model, it characterizes the velocity field changes under different engineering intervention scenarios, validates the effect of measures such as spur dikes and diversion channels in enhancing the flow velocity in front of the harbor, and is suitable for calculating flow and sediment dynamic processes in inland rivers and estuaries [22].

The governing equations of the plane two-dimensional flow-tide model are the Reynolds-averaged Navier-Stokes (RANS) equations for incompressible fluids under the shallow water assumption, considering the effects of the Coriolis force caused by the Earth’s rotation, frictional resistance, etc. The governing equations are as follows:

$${\displaystyle \frac{\partial u}{\partial t}}+u{\displaystyle \frac{\partial u}{\partial x}}+v{\displaystyle \frac{\partial u}{\partial y}}+g{\displaystyle \frac{\partial \zeta }{\partial x}}+{\displaystyle \frac{g{n}^{2}u\sqrt{u^2+v^2}}{h^{4/3}}}-\mu ({\displaystyle \frac{\partial u^2}{\partial x^2}}+{\displaystyle \frac{\partial u^2}{\partial y^2}})-fv=F_x$$

(4)

$${\displaystyle \frac{\partial v}{\partial t}}+u{\displaystyle \frac{\partial v}{\partial x}}+v{\displaystyle \frac{\partial v}{\partial y}}+g{\displaystyle \frac{\partial \zeta }{\partial y}}+{\displaystyle \frac{g{n}^{2}v\sqrt{u^2+v^2}}{h^{4/3}}}-\mu ({\displaystyle \frac{\partial v^2}{\partial x^2}}+{\displaystyle \frac{\partial v^2}{\partial y^2}})+fu=F_y$$

(5)

$${\displaystyle \frac{\partial h}{\partial t}}+{\displaystyle \frac{\partial \left(hu\right)}{\partial x}}+{\displaystyle \frac{\partial \left(hv\right)}{\partial y}}=q$$

(6)

$u$, $v$ represent the depth-averaged velocities in the $x$ and $y$ directions (m/s); $h$ is the water depth (m); $\zeta$ is the water level relative to sea level (m); $n$ is the Manning coefficient (s/m^1/3); $u$ is the eddy viscosity coefficient (m²/s); $f$ is the Coriolis force-related coefficient (1/s); $F_x$ and $F_y$ are the momentum source terms in the $x$ and $y$ directions (m/s²); $q$ is the areal runoff rate (m/s).

For the plane two-dimensional suspended sediment model, the calculation of suspended sediment movement mainly solves the depth-averaged advection-diffusion equation:

$${\displaystyle \frac{\partial h{s}_i}{\partial t}}+{\displaystyle \frac{\partial hu{s}_i}{\partial x}}+{\displaystyle \frac{\partial hv{s}_i}{\partial y}}-{\displaystyle \frac{\partial }{\partial x}}\left(h{\epsilon }_h{\displaystyle \frac{\partial s_i}{\partial x}}\right)-{\displaystyle \frac{\partial }{\partial y}}\left(h{\epsilon }_h{\displaystyle \frac{\partial s_i}{\partial y}}\right)=F_i$$

(7)

$s_i$ represents the vertical average suspended sediment concentration of the $i$-th sediment component (kg/m³); ${\epsilon }_h$ denotes the horizontal turbulent diffusion coefficient of sediment (m²/s); $F_i$ indicates the source-sink term of the $i$-th sediment component, representing the sediment flux exchange between the water body and bed surface (kg/m²/s).

The calculation of bedload transport is based on the Van Rijn formula for non-cohesive sediment:

$$\left|S_{bi}\right|=0.006w_{si}{D}_{50,i}{\displaystyle \frac{u(u-u_{cri}{)}^{1.4}}{\left[\right(s_i-1)g{D}_{50,i}{]}^{1.2}}}$$

(8)

$\left|S_{bi}\right|$ denotes the unit-width bedload transport rate of the $i$-th non-cohesive sediment fraction (m²/s); $D_{50,i}$ represents the median grain size of the $i$-th non-cohesive sediment fraction (m); $u_{cri}$ is the critical incipient velocity of the $i$-th non-cohesive sediment fraction (m/s).

Considering factors such as boundary conditions, hydrological data, and the scope of engineering impacts, the model domain extended from Yijishan in the Wuyu reach upstream to the Nanjing Dashengguan Yangtze River Bridge downstream. The spatial step size of the model was set at Δs = 3–300 m, with approximately 106,127 grid nodes and 107,260 computational cells. The engineering area was refined with a minimum spacing of 3 m to accurately reflect the influence of engineering measures [21]. A diagram of the computational grid for the study reach is shown in Figure 1. The model bathymetry was based on measured underwater topography.

The model was validated using hydrological measurements of water levels and cross-sectional velocity distributions from March 2023.

A comparison between calculated and measured water levels along the reach is presented in Figure 2, showing good agreement. The model effectively captured the comprehensive resistance effects of the river channel, with water level errors within 0.05 m. Figure 3 compares the calculated and measured velocity distributions at various cross-sections, demonstrating close alignment. The validation results confirm that the model accurately represents the hydrodynamic characteristics of the study reach.

This model is used to verify the scouring and silting of the riverbed among different terrains in the Anhui Ma’anshan section of the Yangtze River. In this study, the measured terrain data from October 2020 to March 2023 were adopted for repeated model verification.

Figure 4 shows the plane distribution maps of measured and calculated scouring-silting in the engineering reach from October 2020 to March 2023. It can be seen that the trends of calculated and measured scouring-silting distributions are generally consistent, and the scouring-silting volumes are roughly equivalent. This indicates the rationality of model parameter selection, and the numerical calculation basically reflects the scouring-silting law of the beach and channel in the engineering reach, with errors generally within 20%.

3. Analysis of Siltation Status in Typical Harbor Areas

The Zhengpu Harbor Area is located on the left bank of the left branch of the Jiangxinzhou channel in the Ma’anshan section of the Yangtze River, as shown in Figure 5. The harbor has one 5000-ton-class container berth and two 5000-ton-class general cargo berths. The first phase of the Zhengpu Harbor project commenced operations in December 2014, with the goal of establishing it as a Yangtze River Delta River-Sea Intermodal Transharboration Hub and a Regional and National Logistics Hub [5]. The harbor is immediately upstream of the Niutun River Point Bar, downstream of the left branch of the Central Bar, and faces the Shanghejiazhou Bar and Jiangxinzhou across the river. The interaction between local sandbars and branch channels in the harbor area significantly affects the water depth conditions at the harbor’s front.

3.1. Recent Overall Sedimentation Conditions

3.1.1. Interannual Erosion-Deposition Status of Zhengpu Harbor

Since the completion and operation of Zhengpu Harbor, the average sedimentation thickness from 2016 to 2021 was 0.5 m. The most significant sedimentation occurred between 2019 and 2020, with an annual sedimentation thickness of 1.5 m. The average sedimentation rate from 2016 to 2021 was 10.6%, with a maximum annual sedimentation rate of 39.7%, indicating intense front-channel siltation.

Figure 6 presents statistics on the channel volumes below 0 m, 3 m, and 5 m elevations at the front of Zhengpu Harbor across different years. Here, the 0 m elevation represents the low beach, and the 5 m elevation represents the high beach. The results show that from 2002 to 2024, the channel volume exhibited an accelerated declining trend, indicating overall bed aggradation. The sedimentation rate was faster from 2002 to 2015 and slowed after 2015. From 2002 to 2024, the channel volume below the 0 m elevation decreased by 427 million m³, below the 3 m elevation by 479 million m³, and below the 5 m elevation by 502 million m³. This demonstrates that, below the 5 m elevation, sedimentation over the years has primarily concentrated below the 0 m low beach (accounting for 85% of the total sedimentation), but sedimentation also occurred on the 0 m~5 m high beach.

3.1.2. Progradation of Niutun River Point Bar near the Harbor Area

The Niutun River point bar is located upstream of Zhengpu Harbor. In 2023, the 0 m contour line of the point bar was only 1655 m from the harbor shoreline. The point bar can be divided into a high beach (exposed during mid-low water levels) and a submerged harborion (covered during low water levels). The high beach consists of two parts: a nearshore high beach and an offshore high beach adjacent to the flood embankment. The offshore high beach forms an exposed sandbar during mid-high water levels, separated from the left bank by a narrow channel. The downstream progradation of the Niutun River point bar mainly occurred after 2000. Based on the distance between Reference Point 3 and the Ma’anshan Yangtze River Highway Bridge in Figure 1, Figure 3 statistically analyzes the downstream migration of the −5 m and −10 m contour lines of the Niutun River point bar from 2002 to 2022 (upstream direction is positive).

For the −5 m contour line at the tail of the point bar, continuous downstream migration occurred from 2000 to 2022. From 2000 to 2020, the average migration rate was 200 m~500 m per year, with faster migration around 2010 due to the implementation of point bar stabilization works (Figure 5). The migration rate subsequently decreased, and from 2020 to 2022, cumulative migration was approximately 150 m. The −10 m contour line also migrated downstream at a rate of 100 m~300 m per year. Since Zhengpu Harbor is located downstream of the Niutun River point bar, the deterioration of water depth conditions at the harbor front is directly related to the continuous progradation of the point bar after 2000.

3.2. Local Sedimentation Conditions at Zhengpu Harbor Within a Year

Local dredging was conducted at Zhengpu Harbor in 2014, and additional dredging was performed in January-February 2022 due to siltation at the harbor front. For intra-annual variations, Figure 7 shows local erosion-deposition changes at Zhengpu Harbor from May to October 2022, red represents the sedimentation amount, and blue represents the erosion amount. After dredging, the nearshore channel and harbor front generally experienced siltation of 1~5 m, and the navigation channel still faced insufficient water depth. Thus, both interannual and intra-annual analyses indicate that the harbor front and the river reach from the Ma’anshan Yangtze River Highway Bridge to the inlet of the left branch of the central bar remain in an aggradational environment. This sedimentation trend will continue to reduce the harbor’s water depth.

4. Study on the Causes of Siltation at the Harbor Front

The evolution of braided river reaches exhibits interconnected upstream-downstream dynamics [23]. Therefore, when analyzing the causes of siltation changes at the front of a braided-channel harbor area, it is necessary to consider both upstream river regime changes (flow and sediment conditions) and local bar-channel evolution. Additionally, river engineering structures, such as bank protection works, spur dikes, and bridge construction within the reach, can alter hydrodynamic conditions at the harbor front. Based on the analysis of sedimentation at Zhengpu Harbor, this study focuses on upstream river regime changes, local channel adjustments, and responses to river engineering structures to reveal the primary causes of siltation at Zhengpu Harbor.

4.1. Impact of Upstream River Regime Changes on Siltation in the Harbor Area

The confluence pattern at the two outlets of the Chenjiazhou bifurcation directly affects the hydrodynamic conditions at the entrance of the Ma’anshan river section. By 1998, the maximum flow distribution ratio of the left branch of Chenjiazhou once reached 35%. After 1998, the flow distribution ratio gradually decreased, and it is currently maintained between 10% and 20%. Figure 8 compares the distance changes between the downstream confluence point 1 of Chenjiazhou and the reference cross-section z1 (east-west Liangshan) under different years, as well as the relationship with the flow distribution ratio of the left branch of Chenjiazhou. As the flow distribution ratio changes, the position of the thalweg at the entrance of the Ma’anshan river section has also shifted. From 1988 to 2008, the reduced flow distribution ratio of the left branch of Chenjiazhou weakened the blocking effect of the left branch outflow, further enhancing the mainstream deflection by Dongliang Mountain. After exiting Dongliang Mountain, the mainstream shifted leftward, scouring the upper section of the Niutun River Point Bar. After 2008, the increased flow in the left branch of Chenjiazhou strengthened the blocking effect of its outflow, causing the mainstream exiting Dongliang Mountain to shift rightward and impact the head of Pengxingzhou (Central Bar). Consequently, the Niutun River Point Bar accreted. With the implementation of upstream waterway regulation projects, the flow distribution changes between the left and right branches stabilized, and upstream influences on downstream areas diminished. Overall, as the flow distribution of the Chenjiazhou bifurcation shifted to “left strong, right weak,” the confluence point at the entrance of the Ma’anshan river section moved downstream continuously. The mainstream in the left branch of Jiangxinzhou shifted rightward, weakening the hydrodynamic forces along the left edge of the Niutun River Point Bar, leading to long-term accretion of the point bar. This is unfavorable for maintaining the water depth conditions at the front of Zhengpu Harbor.

4.2. Impact of Local Sandbar-Channel Evolution on Harbor Siltation

(1)

Changes in the Nearshore Thalweg along the left edge of Jiangxinzhou and downstream Migration of Zhengpu Harbor’s confluence point

In recent years, the flow distribution ratio of the right branch of Jiangxinzhou has been 10–14%, while the left branch ratio is 86–90%. After the impoundment of the Three Gorges Reservoir, the riverbed in the upper Jiangxinzhou section of the Ma’anshan River experienced overall erosion. Changes in the deep channel of the upper left branch of Jiangxinzhou affected the evolution of the Niutun River Point Bar and downstream sandbar-channel patterns. The main dynamic axis of the left branch of Jiangxinzhou shifted rightward. Under major flood events, the head of Jiangxinzhou eroded and retreated, causing the scouring point along the left edge of Jiangxinzhou to migrate downstream. The deep channel shifted downstream along the left edge of Jiangxinzhou, providing additional space for the accretion of the Niutun River Point Bar.

Using the 1999 thalweg position as a baseline,

Table 1. Changes in Displacement Distance of the Nearshore Thalweg along the Left Edge of Jiangxinzhou and Downstream Migration of Zhengpu Harbor’s Confluence Point.

Year	Displacement Distance of Nearshore Thalweg at Cross-Section z2 (m) Distance	from Confluence Point 2 of Zhengpu Harbor Bifurcation to Cross-Section z3 (m)
1999	0	/
2002	122.6	0
2005	184.5	311.3
2010	388.7	2353.8
2016	386.3	2763.0
2020	404.3	3172.2
2023	383.2	3572.4

quantifies the displacement distance of the nearshore thalweg along the left edge of Jiangxinzhou at reference cross-section z2 after 1999. From 1999 to 2010, the nearshore thalweg along the left edge of Jiangxinzhou shifted rightward rapidly, averaging 35.3 m per year. After the implementation of bank protection works at the head and left edge of Jiangxinzhou from 2010 to 2020 (see Figure 4), the rightward shift slowed to less than 5 m per year. Using the 2002 thalweg position as a baseline, the downstream migration distance of confluence point 2 to cross-section z3 was calculated. From 2002 to 2005, the thalweg migrated downstream at an average rate of 103 m per year. From 2005 to 2010, the migration accelerated to approximately 400 m per year. After 2010, due to regulatory projects, the migration slowed to approximately 100 m per year.

(2)

Changes in Flow Distribution Ratio of the Branch Channel Between Xiahejiazhou and the Central Bar

With adjustments in upstream and downstream river regimes and erosion of the deep channel in the left branch of Jiangxinzhou, the main dynamic axis transitioning from the scouring point along the left edge of Jiangxinzhou to the front of Zhengpu Harbor shifted rightward. The head of the Central Bar continued to erode and retreat, and the entire Central Bar migrated downstream. By 2016, the 0 m isobath reference point 3 at the head of the Central Bar had migrated approximately 1138 m downstream from cross-section z3. The left branch of the Central Bar generally experienced siltation, while the right branch developed and the left branch declined. Figure 9 shows that in 2006, the left branch of the Central Bar accounted for approximately 75% of the flow. The flow distribution ratio of the left branch gradually decreased, reaching a minimum of 49.6% in 2021 (a reduction of ~20%). Meanwhile, the flow distribution ratio of the right branch increased overall, peaking at 34.6% in 2022. The weakened hydrodynamic forces downstream of the Taiyang River Estuary created favorable conditions for the accretion of the Niutun River Point Bar. After erosion and retreat of the Central Bar’s head, the reduced flow distribution ratio of the left branch and development of the right branch caused the mainstream in the left branch of Ma’anshan to shift rightward. This further weakened hydrodynamic forces near the Taiyang River Estuary, exacerbating the downstream accretion of the Niutun River Point Bar and siltation in the Zhengpu Harbor area.

4.3. Impact of Surrounding River Engineering Structures on Harbor Siltation

The river engineering structures around the harbor area mainly include river regulation projects, waterway improvement projects, and bridge projects. Specifically, the river regulation projects include Phase I and Phase II regulation projects of the Ma’anshan river section. The waterway improvement projects include the Phase I waterway improvement project of the Jiangxinzhou-Wujiang River, the waterway improvement project of the Jiangxinzhou River, and the Phase II waterway improvement project of the Jiangxinzhou-Wujiang River (under construction). The cross-river passage projects include Ma’anshan Yangtze River Highway Bridge and Ma’anshan Yangtze River Road-Rail Bridge (under construction), as shown in Figure 5.

Based on the two-dimensional hydrodynamic numerical model of the Ma’anshan section of the Yangtze River established by the research institute, the flow velocity changes under the influence of different projects were calculated, as shown in Figure 10, blue represents an increase in flow velocity, and red represents a decrease in flow velocity. Taking the bankfull discharge as the calculation condition, the calculation results show that after the implementation of Phase I of the waterway improvement project of the Jiangxinzhou-Wujiang River, the overall flow velocity of the point bar decreased under the influence of the protection belt of the Niutun River Point Bar, with a reduction range generally of 0.02–0.15 m/s. The Ma’anshan Yangtze River Highway Bridge has three bridge piers in the left and right branches of Jiangxinzhou, respectively. The flow velocity in the sheltered areas upstream and downstream of the bridge piers decreased, while the flow velocity between bridge spans increased. After the project implementation, the flow velocity near Zhengpu Harbor slightly decreased by approximately 0.01 m/s. After the implementation of the waterway improvement project of the Jiangxinzhou River, the flow velocity at the head of Central Bar slightly decreased under the influence of the bank protection project, with a reduction range generally of 0.02–0.05 m/s. The flow velocity near Zhengpu Harbor slightly increased, but the amplitude was small, generally within 0.02 m/s. Phase II of the waterway improvement project of the Jiangxinzhou-Wujiang River heightened the original 3 bed protection belts at Shanghejiazhou, newly built 2 bed protection belts to protect the head of the Central Bar of Jiangxinzhou, and implemented bank protection for the high beach of Xiahejiazhou. After the implementation of the Phase II regulation project of the Ma’anshan river section, the flow velocity slightly decreased under the influence of the bank protection project downstream of the bridge along the left edge of Jiangxinzhou, with a reduction range generally of 0.02–0.05 m/s. After the project implementation, the flow velocity near Zhengpu Harbor slightly increased, but the amplitude was small, generally within 0.02 m/s. The Ma’anshan Yangtze River Road-Rail Bridge has 3 bridge piers on Niutun River Point Bar and 2 main piers in the left and right branches of Jiangxinzhou, respectively. The flow velocity near Zhengpu Harbor slightly decreased, but the amplitude was small, generally approximately 0.01 m/s.

From the calculation results, the main impact of the constructed river engineering structures in the area is still concentrated locally. After the implementation of the Shanghejiazhou protection project and the Central Bar head protection project, the flow velocity near the harbor front increased.

5. Multivariate Regression Quantitative Analysis of Dominant Siltation Factors

According to the multivariate regression analysis method, a multivariate regression relationship was established between the siltation degree at the front of Zhengpu Harbor and influencing factors [24]. The siltation degree at the front can be characterized by the channel volume between Ma’anshan Bridge and Taiyang River Estuary. The influencing factors are further divided into four categories: flow and sediment conditions, upstream river regime changes, local river regime changes, and impacts of key river engineering structures [25], as shown in

Table 2. Main variables for multivariate regression analysis of siltation degree at the Zhengpu Harbor front.

Variable Category	Specific Features	Variable Value	Variable Code
Dependent Variable	Siltation degree at Zhengpu Harbor front	The channel volume is below 6 m at Zhengpu Harbor front (108 m3)	Y1
Independent Variables	Flow and sediment conditions	The average flood season (July–August) discharge at Datong Station (m3/s)	X1
		The channel volume of Ma’anshan reach (108 m3)	X2
	Upstream river regime changes	The flow distribution ratio change in the Chenjiazhou left branch (%)	X3
		The retreat distance of the Chenjiazhou right margin (−5 m isobath) (m)	X4
	Local river regime and flow distribution changes	The distance between the left branch −20 m deep channel, and the Jiangxinzhou Bar embankment (m)	X5
		The distance between the left branch −20 m deep channel and the bridge (upstream as positive) (m)	X6
		The distance between the main dynamic axis and the left margin (m)	X7
		The distance between −5 m isobath of accretion/downstream shift of Niutun River Point Bar and the bridge (m)	X8
		The flow distribution ratio of the channel between Xiahejiazhou Bar and Jiangxinzhou Bar (%)	X9
		The flow distribution ratio of Jiangxinzhou left branch (%)	X10
	Major river engineering structures	Phase I Jiangxinzhou-Wujiang reach waterway improvement project (0/1)	X11
		Ma’anshan Yangtze River Highway Bridge (0/1)	X12
		Jiangxinzhou reach waterway improvement project, (0/1)	X13
		Phase II Ma’anshan reach regulation project (0/1)	X14
		Phase II of the Jiangxinzhou-Wujiang reach waterway improvement project (under construction) (0/1)	X15
		Ma’anshan Yangtze River Road-Rail Bridge (under construction) (0/1)	X16

.

Flow and sediment conditions were rooted in fundamental theories of sediment transport and fluvial dynamics. For flow and sediment conditions, they are characterized by the average flood season (July–August) discharge at Datong Station and the channel volume of the Ma’anshan reach. According to the sediment transport theory, an increase in discharge enhances sediment-carrying capacity; thus, the flood-season discharge is used to characterize the impact of upstream water-sediment conditions on local siltation. The channel aggradation-erosion equilibrium theory employs channel volume as an indicator of long-term fluvial bed evolution, where a decreasing volume signifies intensified siltation. The water-sediment coupling principle emphasizes the interactive effects of discharge and channel morphology on sediment dynamics, justifying the combined use of discharge and channel volume to analyze hydrodynamic forces and siltation processes.

For upstream river regime changes, they are mainly characterized by two indicators: the flow distribution ratio change in the Chenjiazhou left branch and the retreat distance of the Chenjiazhou right margin (−5 m contour line distance from right margin). The selection of upstream river regime changes and local river regime and flow distribution changes variables is rooted in the characteristics of multi-anabranching rivers, as the total volume and process of upstream inflow directly determine the discharge distribution ratio of each branch. Meanwhile, changes in the main flow direction of upstream and downstream inflows directly lead to the deflection of the thalweg axis.

Local river regime and flow distribution changes are subdivided into three aspects: First, the local hydrodynamic environment is specifically characterized by three indicators: the distance between the left branch, the −20-m deep channel, and the Jiangxinzhou bar embankment; the distance between the left branch, the −20-m deep channel, and the bridge (upstream direction is positive); and the distance between the main dynamic axis and the left margin.

Second, the effect of the Niutun River point bar is characterized by the distance between −5 m isobath of accretion/downstream shift of the Niutun River Point Bar and the bridge (upstream direction is positive).

Third, the influence of the local flow distribution pattern, characterized by two indicators: the flow distribution ratio of the channel between Xiahejiazhou Bar and Jiangxinzhou Bar, and the flow distribution ratio of Jiangxinzhou’s left branch.

Major river engineering structure variables are selected based on their clear impact on flow structure, which influences sediment transport by altering river boundary conditions. The impacts of major river engineering structures specifically include six main projects: Phase I of the Jiangxinzhou-Wujiang reach waterway improvement project, the Ma’anshan Yangtze River Highway Bridge, the Jiangxinzhou reach waterway improvement project, Phase II of the Ma’anshan reach regulation project, Phase II of the Jiangxinzhou-Wujiang reach waterway improvement project (under construction), and the Ma’anshan Yangtze River Road-Rail Bridge (under construction). The project impact variables can be represented by 0/1 variables [26].

Based on SPSS Stat 29 calculations [27], a multivariate regression equation was finally established with the channel volume of the siltation-sensitive area at Zhengpu Harbor front as the dependent variable and other factors as independent variables. The regression coefficients for each variable are presented in

Table 3. Multivariate regression analysis results of siltation degree at Zhengpu Harbor front.

Model	Unstandardized Coefficients		Standardized Coefficients	t	Significance
	β	Standard Error	β′
Constant	−1.347	6.253		−0.215	0.840
X1	0.000	0.000	0.083	1.694	0.166
X2	0.187	0.532	0.186	0.353	0.742
X3	−0.005	0.009	−0.161	−0.610	0.575
X4	0.001	0.001	0.177	0.779	0.480
X5	0.001	0.000	0.238	1.807	0.145
X6	0.000	0.000	0.136	0.227	0.831
X7	−0.001	0.002	−0.469	−0.640	0.557
X8	0.000	0.000	0.504	2.230	0.090
X9	−0.011	0.009	−0.491	−1.184	0.302
X10	0.003	0.005	0.059	0.549	0.612
X11	−0.083	0.035	−0.255	−2.397	0.075
X12	−0.011	0.025	−0.033	−0.422	0.695
X13	0.003	0.018	0.010	0.189	0.860
X14	−0.086	0.086	−0.186	−1.005	0.372
X15	0.070	0.052	0.126	1.340	0.251
X16	0.066	0.027	0.087	2.433	0.072

.

From the multivariate regression calculation results, for the siltation at Zhengpu Harbor front, the most influential factors are local hydrodynamic environment changes, with the order of influence being distance between the main dynamic axis and the left margin (X7), distance between the left branch −20 m deep channel and the Jiangxinzhou Bar embankment (X5), and distance between the left branch −20 m deep channel and the bridge (upstream positive) (X6).

The effect of the key point bar, Niutun River point bar progradation (X8), is also significant, with slightly lower influence than local hydrodynamic environment changes. Changes in local flow distribution pattern, are important influencing factors, mainly the increased flow distribution ratio between Xiahejiazhou Bar and Jiangxinzhou Bar (X9).

Upstream river regime changes are important influencing factors, including increased flow distribution ratio of Chenjiazhou left branch (X3) and retreat of Chenjiazhou right margin (X4). Among major river engineering structures, those with greater impacts are Phase I of the Jiangxinzhou-Wujiang reach waterway improvement project (0~1) (X11), the Ma’anshan Yangtze River Highway Bridge (0~1) (X12), and the Ma’anshan Yangtze River Road-Rail Bridge (under construction) (0~1) (X16).

Under new flow-sediment conditions, overall channel erosion (X2) and upstream flow changes (X1) also have some influence. The standardized coefficients β’ were further processed to obtain the weight of each independent variable in the regression equation influence. The calculation results show that for siltation at Zhengpu Harbor front: Overall flow and sediment condition changes account for 4.1%; upstream river regime changes account for 14.2%; local river regime and flow distribution changes account for a relatively large proportion, including: Local hydrodynamic environment changes: 28.5%, the Niutun River point bar progradation: 18.9%, local flow distribution pattern changes: 22.1%; Impacts of nearby major river engineering structures account for 12.2%.

In summary, the study found that the rightward deflection of the main dynamic axis in the Ma’anshan left branch, the decreased flow distribution ratio of the Central Bar left branch, and progradation of the Niutun River point bar are key causes of siltation at Zhengpu Harbor, accounting for nearly 70%. The implementation of river engineering structures such as bridges has a certain influence, accounting for approximately 12%. Changes in upstream flow-sediment conditions and upstream branch flow distribution account for approximately 18%.

Comparative analyses with the Mekong and Amazon basins reveal the universal mechanisms in anabranching river systems: regardless of tidal-dominated (Mekong) or fluvial-dominated (Amazon) regimes, migration of mid-channel sandbars and hydrodynamic variations are core drivers of siltation [28,29]. Channel morphology changes exacerbate local siltation by altering flow velocity and sediment transport pathways. Notably, both basins demonstrate that sediment dynamics arise from the interplay of natural hydrosedimentary processes and anthropogenic modifications, highlighting the need for integrated management of sandbar stability and hydrodynamic optimization to mitigate siltation. This cross-basin consistency confirms that mid-channel sandbar migration and human-induced flow modifications collectively dictate sediment accumulation patterns in anabranching rivers.

6. Engineering Interventions for Siltation Issues at the Harbor Front

Based on the dominant factors of harbor front siltation proposed in this study, corresponding engineering management measures are designed. The effectiveness of various engineering measures is analyzed under bankfull discharge conditions.

To address the issue of weakened hydrodynamic conditions at the harbor front caused by the progradation of the Niutun River point bar, diagonal dredging can be conducted at the tail of the Niutun River point bar. This will enhance the hydrodynamic forces at the harbor front through channel diversion via the flanking channel, combined with dredging of the harbor basin and approach channel to improve water depth conditions. Numerical model calculations show that under bankfull discharge conditions, the flow velocity at the downstream harbor front generally increases by 0.01–0.05 m/s due to dredging and diversion (Figure 11a). In areas with significant local dredging volumes, the unit-width discharge increases due to deeper water depths, but flow velocity decreases. Based on past experience, subsequent siltation will occur.

To address the retreat of Central Bar’s head and the reduced flow distribution ratio of the left branch, spur dike stabilization works can be implemented at the head of the Central Bar in the Jiangxinzhou waterway. Under bankfull discharge conditions, post-implementation flow velocity decreases at the bar head, stabilizing it (Figure 11c). The flow velocity at Zhengpu Harbor’s front slightly increases by ~0.02 m/s.

Additionally, modifying the Shanghejiazhou spur dikes can further increase the flow distribution ratio of Central Bar’s left branch and improve hydrodynamic conditions at the harbor front. One new spur dike will be added upstream of the existing Shanghejiazhou spur dikes, while the original two upstream spur dikes will be elevated and extended. Under bankfull discharge conditions, pre- and post-implementation velocity changes are shown in Figure 11d. The figure shows that after adjusting the Shanghejiazhou spur dikes (with dike head elevation at −5 m), flow velocity decreases significantly in the shielded area downstream of the dikes. The velocity reduction in the right inlet of Central Bar’s right branch ranges from 0.05 m/s to 0.15 m/s, while the velocity increase in Central Bar’s left branch ranges from 0.01 m/s to 0.04 m/s.

Considering the combined effects of all schemes, the flow velocity at Zhengpu Harbor’s front increases by over 0.1 m/s after implementation.

To validate the direct impact of the project on sedimentation issues in the study area, ten measuring points were selected to statistically analyze terrain elevation changes under different conditions. The precise locations of the ten measuring points are shown in Figure 5, including three harbor basin points and seven approach channel points. Based on the statistical elevation changes, a line chart (Figure 12) was created in which black arrows indicate the variation between post-dredging terrain elevation and terrain elevation after siltation stabilization, defined as D. Red arrows denote the variation between post-dredging terrain elevation and terrain elevation after project implementation, defined as D₁: the elevation difference from project implementation to siltation stabilization minus A. The siltation reduction efficiency is calculated as D₁ divided by D multiplied by 100 percent, thereby providing an intuitive demonstration of the project’s siltation reduction effect. It should be noted that all terrain elevation data in the chart are based on normal-year computational conditions.

According to the above calculation method, the results show that the average siltation reduction efficiency ratio of the first three harbor basin measuring points is approximately 32%, which slowly increases with the increase in distance. In the approach channels of measuring points 4–10, the efficiency generally shows an upward trend, with a Maximum ratio of approximately 47.6%. However, the siltation reduction effect is not obvious at measuring points 9 and 10.

7. Conclusions

(1)

This study takes Zhengpu Harbor in the Ma’anshan reach as a typical harbor, analyzing the response relationship between harbor front siltation and upstream river regime, local bar-channel evolution, and nearby major river engineering structures in the Yangtze River’s braided reach. Regarding the causes of siltation: as the mainstream at the confluence outlet of upstream Chenjiazhou deflects rightward, the impact point along the left margin of Jiangxinzhou Bar migrates downstream, and the continuous retreat of the left margin provides space and hydrodynamic conditions for the downstream advancement of the Niutun River point bar. On the other hand, after the mainstream transitions from the impact point along Jiangxinzhou Bar’s left margin to near Taiyang River Estuary, as Central Bar’s head continues to erode and retreat, the area in front of Taiyang River Estuary becomes detached from the mainstream zone, the hydrodynamic forces in Central Bar’s left branch decrease, and the hydrodynamic forces along Zhengpu Harbor’s front weaken significantly. Consequently, siltation at Zhengpu Harbor’s front is inevitable. Under current river regime conditions, siltation at Zhengpu Harbor’s front will persist.

(2)

Through multivariate regression analysis, this study quantitatively quantifies the impacts of factors, including flow-sediment conditions, upstream river regime changes, local channel adjustments, and key river engineering structures. In terms of weight the rightward deflection of the main dynamic axis in the Ma’anshan left branch, the reduced flow distribution ratio of Central Bar’s left branch, and the progradation of the Niutun River point bar are the key causes of harbor siltation, accounting for nearly 70%. Changes in upstream flow-sediment conditions and upstream branch flow distribution account for approximately 18%. The implementation of water-related projects such as bridges has a certain influence, accounting for approximately 12%.

(3)

In response to the identified key factors contributing to the siltation at Zhengpu Harbor, this study has proposed a series of targeted engineering interventions. These include stabilizing the heads of evolving shoals to optimize the flow distribution of local branches, excavating a diversion channel at the tail of the Niutun River shoal to enhance flow dynamics, and conducting emergency dredging at the harbor front. Numerical modeling results demonstrate that implementing these measures can increase the flow velocity at the harbor front by more than 0.1 m/s, effectively alleviating the siltation problem. The study concludes that the proposed engineering measures can reduce annual siltation by approximately 30% under normal-year hydrological conditions, demonstrating their feasibility in mitigating siltation trends in multi-anabranching river systems. Overall, the findings of this research offer valuable references for tackling siltation issues in harbors situated within complex anabranching river systems, providing practical solutions and theoretical support for similar engineering challenges.