Sediment Distribution, Depositional Trends, and Their Impact on the Operational Longevity of the Jiahezi Reservoir

Abstract

The Jiahezi Reservoir is a typical plain reservoir constructed on a sediment-laden river in the mountainous region of Xinjiang. After 58 years of operation, it faces critical challenges—particularly insufficiently characterized sediment size distribution and siltation behavior—that continue to undermine its long-term operational performance. This study investigates particle-size distribution and depositional trends through integrated field sampling and two-dimensional (2D) numerical sediment modeling, using the Jiahezi Reservoir as a case study. Results show that the median particle size (d₅₀) is 0.027 mm, with 76.48% of the particles ranging from 0.005 to 0.25 mm, indicating predominantly fine-grained sediment. Particle size gradually decreases both downstream and laterally from the channel. Hydrodynamic sorting produces distinct spatial distribution patterns, with fine particles occupying more extensive areas than coarse ones. The sedimentation patterns within the reservoir area can be broadly categorized into four distinct zones. By analyzing sediment transport mechanisms and depositional characteristics, this study establishes a foundation for developing targeted strategies to enhance the reservoir’s operational longevity.

1. Introduction

Sand sedimentation in reservoirs [1,2] poses a significant challenge to water conservancy worldwide, as it disrupts riverbed evolution, impairs reservoir functionality, and shortens operational lifespan—significantly compromising sustainable operation of the hydraulic infrastructure [3,4]. Reservoirs built on sediment-laden rivers have long faced sedimentation problems [5], making sedimentation a critical concern requiring urgent attention from international, academic, and engineering communities.

These challenges are particularly acute in regions such as Xinjiang, where geomorphological conditions intensify sediment-related problems. The region’s landscape is characterized by “three mountains and two basins,” resulting in high sediment concentrations in many mountainous rivers [6]. In addition, sediment management was often neglected during the early stages of reservoir construction, leading to poorly understood particle size distribution and depositional patterns [7]. These factors contribute to accelerated reservoir infilling and reduced storage capacity, ultimately shortening reservoir service life [8,9]. Addressing these knowledge gaps is essential for developing targeted strategies to extend the reservoirs’ in-service duration.

Understanding sediment characteristics—particularly particle size distribution—is crucial to managing deposition in rivers and reservoirs [10,11]. To that end, numerous studies have been conducted to analyze sediment characteristics in various river systems, highlighting spatial variations in particle size. For example, Xu et al. [12] used a laser particle size analyzer to determine that sediment in the Songnen River becomes coarser downstream. In addition, Xiao et al. [13] reported that suspended sediment in the Yongjiang River estuary and nearby waters consists mainly of silt and clay, with low sand content, while Hou et al. [14] found that in the Xiuliugou River, median sediment size decreases from hilly to desert to plain regions. However, detailed sediment data remain lacking for many rivers and reservoirs [15,16].

Both domestic and international researchers have employed a range of methods—including prototype data analysis, physical models, and numerical simulations—to investigate reservoir sedimentation [17,18,19,20,21,22,23]. With advancements in computing power and sediment transport theory, numerical modeling has become increasingly sophisticated [24]. For instance, Yang et al. [25] employed a one-dimensional (1D) model for unsteady flow and non-uniform sediment to demonstrate that inflow sediment load and flood-season water levels are key factors influencing sedimentation in the Three Gorges Reservoir. In an earlier study, Guo et al. [26] used the MIKE 21 sediment module to simulate sediment behavior within a reservoir, and found that particle size affects transport and deposition processes, and that downstream water levels are influenced by flood peaks and sediment concentration. Jia et al. [27] observed a downstream fining trend in the Xiaolangdi Dam area, with finer particles settling near the dam. More recently, Guo et al. [28] examined sediment deposition patterns and influencing factors in the Three Gorges Reservoir using field data from 2003 to 2022. Similarly, Zulfan et al. [29] used the MIKE 21 FM 2D model to assess sedimentation in Indonesia’s Pandan Duri Reservoir. Notably, compared to 1D models, 2D sediment models offer higher computational efficiency and spatial accuracy [30].

In general, most Chinese studies have focused on sedimentation in reservoirs along major river systems, such as the Yangtze and Yellow Rivers. In contrast, research on reservoirs affected by mountain torrents in Xinjiang—particularly using 2D modeling—is limited. This gap underscores the need for focused investigation of reservoirs such as Jiahezi.

The Jiahezi Reservoir is a typical plain reservoir influenced by sediment-laden mountainous rivers in Xinjiang. As with many similar hydraulic infrastructures in the region, the sediment particle size distribution and depositional characteristics remain poorly understood. As such, this research aims to investigate sedimentation processes in the Jiahezi reservoir by combining field sampling with two-dimensional numerical modeling. Specifically, the study seeks to clarify the spatial differentiation characteristics of sediment particle size within the Jiahezi reservoir area, elucidate the transport and distribution patterns of different particle size composition and systematically uncover seasonal sedimentation patterns by comparatively analyzing the spatial distribution and intensity differences in reservoir sedimentation under different hydrological and operational conditions. Investigating sedimentation in Xinjiang’s plain reservoirs has important practical implications for extending the service life of reservoirs in cold-arid regions and enhancing water resource security in China’s northwest border areas.

2. Study Area

2.1. River and Reservoir Characteristics

The Manas River stems from Glacier No. 43 in the Eren Habirga Mountains, located in the central part of the northern Tianshan range. Flowing northward into Manas Lake, it is a typical sediment-laden mountainous river with a total length of ~450 km. The upper reaches are characterized by rapid currents and numerous gorges, which transition into open and flat downstream plains, with well-developed meanders. Runoff is highly seasonal with ~80.11% of the annual runoff occurring during the summer months, from June to September. The primary sources of runoff include snow and glacial meltwater from the high mountains, as well as precipitation in the upper reaches. The river flows through a mountain-oasis-desert system.

The Jiahezi Reservoir is located at the northern foot of the Tianshan Mountains and the southern margin of the Junggar Basin. Specifically, it lies between the Shilidun and Xiaqiaozi sections of the Manas River, functioning as a river-blocking reservoir along the river’s middle reaches. The reservoir is situated within a geomorphic unit with ground elevation ranging from 380 to 400 m above sea level.

Jiahezi is classified as a large (Type II) reservoir and serves as a water conservancy hub, with flood control and irrigation as its primary and secondary purposes, respectively. The reservoir’s total storage capacity is 101.4 million m³. Its normal water level is 399.8 m, corresponding to an active storage volume of 82 million m³. The relative positions of the Manas River and the Jiahezi Reservoir area are shown in Figure 1.

2.2. Hydrological Characteristics

Flow data was obtained from Shihezi City’s water service center. The Jiahezi Reservoir’s inflow pattern exhibits significant seasonal variability, primarily influenced by the regulatory operation of its water intake sluice. The intra-annual distribution of inflow is uneven, with the highest monthly average recorded in July 2024, reaching 45.78 m³/s. Inflow is largely concentrated in the summer and autumn months—specifically July through October—which together account for ~81.7% of the total annual average inflow. This seasonal concentration was even more pronounced in 2023, when these four months contributed 88.4% of the year’s total inflow. A more detailed breakdown is available in Figure 2.

3. Sampling Strategy and Methodology

Given that scour and deposition within the river channel are more sensitive to sediment-laden flow variations than other reservoir zones, the analysis focused on sediment distribution patterns along the channel alignment and in the longitudinal excavation direction. The following key control points were selected along the flow path: point 1—intake sluice; point 2—flood discharge sluice; points 3 and 4—river bend within the reservoir; and point 5—geometric center between the east and west outlet sluices.

This sampling was conducted in collaboration with the Xinjiang Geological Survey and Design Institute, a systematic grid-based sampling design was employed to collect silt and sand samples within the Jiahezi Reservoir, following the approach outlined by Wang and Qi [31]. The spacing of regional grid points and the vertical sampling interval, was based on the specifications of the Code for Engineering Geological Investigation of Water Resources and Hydropower (GB 50487-2008) [32]. Sampling points were established on a 300~500 m grid across the reservoir area. At each point, vertical stratified sampling was conducted using the test pit method. During the excavation of the pit’s longitudinal section, groundwater was encountered at a depth of 5 m. Based on this finding, the final excavation depth of the pit was determined to be 5 m [32], the Luoyang drill is used for borehole sampling, with each borehole having a depth of 5 m [33]. A mixed soil sample is collected at every 1 m interval. For check dams where crops are cultivated, the plow layer must be removed prior to drilling and sampling. These control and sampling points’ spatial distribution is shown in Figure 3, while the expanded visual shows the five layers that were sampled at each point.

All sediment samples were analyzed for particle size distribution using a Beckman Coulter LS 13 320 laser particle size analyzer (Miami, FL, USA). Prior to initiating the sediment particle size analysis, the parameters of the laser diffraction particle size analyzer—including pump speed, ultrasonic intensity, ultrasonic duration, and obscuration—were calibrated. The accuracy of the instrument was verified and calibrated using standard reference particles. For sample weighing, approximately 0.5 g of the specimen was measured using a precision pan balance [34]. Three consecutive readings were taken and confirmed consistent before proceeding. The sample was then introduced into the laser particle size analyzer, with the obscuration adjusted to approximately 20% (not exceeding 30%). Subsequently, the instrument performed ultrasonic dispersion of the soil sample, followed by laser scattering measurement for 30 s. The soil suspension was automatically drained via the drainage control system, and the system was rinsed 2–3 times. The above procedure was repeated before each sample measurement. Each sample was measured in triplicate, with the mean value recorded as the final result. The acquired sediment particle size distribution data were used to define key input parameters and boundary conditions for the sediment transport module in the subsequent hydrodynamic and sediment numerical modeling part of the study. To support subsequent sediment transport modeling, median particle sizes for each depth layer were calculated using a particle size-weighted average across all reservoir sampling points. The results are shown in Figure 3.

4. Hydrodynamic and Sediment Transport Model

4.1. Model Principles

The 2D hydrodynamic model employed in this study is derived from the three-dimensional (3D), incompressible, Reynolds-averaged Navier–Stokes equations. It is formulated under the Boussinesq approximation and the hydrostatic pressure assumption. The governing equations include the continuity equation, horizontal momentum equations, and sediment transport equations, as detailed below.

The local continuity equation is expressed as:

$${\displaystyle \frac{\mathsf{\partial }h\overline{u}}{\mathsf{\partial }x}}+{\displaystyle \frac{\mathsf{\partial }h\overline{v}}{\mathsf{\partial }y}}+{\displaystyle \frac{\mathsf{\partial }h}{\mathsf{\partial }t}}=hS$$

(1)

The two horizontal momentum equations in the x and y directions are given by:

$$\begin{array}{l}{\displaystyle \frac{\mathsf{\partial }h\overline{u}}{\mathsf{\partial }t}}+{\displaystyle \frac{\mathsf{\partial }h{\overline{u}}^2}{\mathsf{\partial }x}}+{\displaystyle \frac{\mathsf{\partial }h\overline{u}\overline{v}}{\mathsf{\partial }y}}=f\overline{v}h-gh{\displaystyle \frac{\mathsf{\partial }\eta }{\mathsf{\partial }x}}-{\displaystyle \frac{h}{{\rho }_0}}{\displaystyle \frac{\mathsf{\partial }{p}_a}{\mathsf{\partial }x}}-{\displaystyle \frac{g{h}^2}{2{\rho }_0}}{\displaystyle \frac{\mathsf{\partial }\rho }{\mathsf{\partial }x}}+{\displaystyle \frac{{\tau }_{sx}}{{\rho }_0}}-{\displaystyle \frac{{\tau }_{bx}}{{\rho }_0}}-{\displaystyle \frac{1}{{\rho }_0}}({\displaystyle \frac{\mathsf{\partial }{s}_{xx}}{\mathsf{\partial }x}}+{\displaystyle \frac{\mathsf{\partial }{s}_{xy}}{\mathsf{\partial }y}})\\ +{\displaystyle \frac{\mathsf{\partial }(h{T}_{xx})}{\mathsf{\partial }x}}+{\displaystyle \frac{\mathsf{\partial }(h{T}_{xy})}{\mathsf{\partial }y}}+h{u}_{s}S\end{array}$$

(2)

$$\begin{array}{l}{\displaystyle \frac{\mathsf{\partial }h\overline{v}}{\mathsf{\partial }t}}+{\displaystyle \frac{\mathsf{\partial }h{\overline{v}}^2}{\mathsf{\partial }y}}+{\displaystyle \frac{\mathsf{\partial }h\overline{u}\overline{v}}{\mathsf{\partial }x}}=-f\overline{u}h-gh{\displaystyle \frac{\mathsf{\partial }\eta }{\mathsf{\partial }y}}-{\displaystyle \frac{h}{p_0}}{\displaystyle \frac{\mathsf{\partial }{p}_a}{\mathsf{\partial }x}}-{\displaystyle \frac{{gh}^2}{2{\rho }_0}}{\displaystyle \frac{\mathsf{\partial }\rho }{\mathsf{\partial }y}}+{\displaystyle \frac{{\tau }_{sy}}{{\rho }_0}}-{\displaystyle \frac{{\tau }_{by}}{{\rho }_0}}\\ -{\displaystyle \frac{1}{{\rho }_0}}({\displaystyle \frac{\mathsf{\partial }{s}_{yx}}{\mathsf{\partial }x}}+{\displaystyle \frac{\mathsf{\partial }{s}_{yy}}{\mathsf{\partial }y}})+{\displaystyle \frac{\mathsf{\partial }(h{T}_{xy})}{\mathsf{\partial }x}}+{\displaystyle \frac{\mathsf{\partial }(h{T}_{yy})}{\mathsf{\partial }y}}+h{v}_{s}S\end{array}$$

(3)

where t is time; x, y are the Cartesian spatial coordinates; η is the water level; d is the static water depth; h is the total water depth (h = η + d); $\overline{u}$, $\overline{v}$ represent the depth-averaged velocity components in the x and y directions, respectively; f is the Coriolis force coefficient, calculated as F = 2w sin φ, where w is the Earth’s angular velocity and φ is the local latitude; g is the gravitational acceleration; ρ is the density of turbid water; ρ₀ is the density of water; Pa is the local atmospheric pressure; S_xx, S_xy, S_yx, and S_yy are radiation stress components; τ_sx and τ_sy are wind shear stresses on the water surface, τ bx and τ by are bed shear stresses; T_xx, T_xy, T_yx, and T_yy are horizontal viscous stress components; S is a general source term; and u_s and v_s represent the x and y components of velocity due to point or distributed sources.

The sediment transport model is governed by the advection–diffusion equation for vertically averaged sediment concentration:

$${\displaystyle \frac{\mathsf{\partial }\overline{c}}{\mathsf{\partial }t}}+u{\displaystyle \frac{\mathsf{\partial }\overline{c}}{\mathsf{\partial }x}}+\upsilon {\displaystyle \frac{\mathsf{\partial }\overline{c}}{\mathsf{\partial }y}}={\displaystyle \frac{1}{h}}{\displaystyle \frac{\mathsf{\partial }}{\mathsf{\partial }x}}(h{D}_x{\displaystyle \frac{\mathsf{\partial }\overline{c}}{\mathsf{\partial }x}})+{\displaystyle \frac{1}{h}}{\displaystyle \frac{\mathsf{\partial }}{\mathsf{\partial }x}}(h{D}_x{\displaystyle \frac{\mathsf{\partial }\overline{c}}{\mathsf{\partial }x}})+Q_L{C}_L{\displaystyle \frac{1}{h}}-s$$

(4)

where c is the depth-averaged sediment concentration (kg/m³); x, y are spatial coordinates; h is the water depth (m); Q_L is the horizontal source term flow rate per unit area (m/s); C_L is the source sediment concentration (kg/m³); u, v are the depth-averaged flow velocities in the x and y directions (m/s), respectively; Dx, Dy are the turbulent diffusion coefficients in the x and y directions (m²/s), respectively; and S is the bed exchange (erosion/deposition) source term (kg/(m³·s).

Sediment deposition is modeled based on the settling velocities of particles, which vary with grain size:

$$\omega ={\displaystyle \frac{M\upsilon }{Nd}}{\left[\sqrt{{\displaystyle \frac{1}{4}}+{({\displaystyle \frac{4N}{3M^2}}{\displaystyle \frac{\Delta g}{{\upsilon }^2}}{d}^3)}^{1/n}}-{\displaystyle \frac{1}{2}}\right]}^n$$

(5)

where ω is the settling velocity of sediments particle, clear water or low-turbidity water. Δ = (γs − γ)/γ is the sediment’s relative submerged density, where γs is the sediment particle density and γ is the fluid density; g is the gravitational acceleration, d is the particle diameter; and M, N, and n are empirical or calibration parameters depending on sediment characteristics and flow conditions.

The erosion and deposition module is as follows:

$${\varphi }_b=F_e-F_d$$

(6)

where ${\varphi }_b$ is the exchange of sediment between the water bottom and the water column. F_e is the deposition rate. F_d is the erosion rate. (F_d = ω × c_b, ω is the settling velocity of sediments, c_b is the concentration of suspended sediment near the bed surface [35])

$${\displaystyle \frac{c_b}{c}}=0.4{\left({\displaystyle \frac{d^2}{D_{sg}}}\right)}^{d-64}+1.64$$

(7)

where Dsg is the geometric mean particle size of suspended sediment, d is the diameter of sediment. c represents the sediment concentration in water.

$$F_e=\omega P{E}_s$$

(8)

where E_s is the sediment entrainment coefficient [36] (E_s = 1.3 × 10⁻⁷Z⁵/1 + 4.3 × 10⁻⁷Z), p is the Volume fraction of bottom sediments [37].

$$Z={\alpha }_1{\displaystyle \frac{\sqrt{c_D}\left|u\right|}{\omega }}{R}_j^{{\alpha }_2}$$

(9)

where R_j is the sand grain Reynolds number. ${\alpha }_1$, ${\alpha }_2$ representative Coefficient [38].

The friction module is as follows [39]:

$$\tau (u)=\rho u_{*}^2(1+r_w)$$

(10)

where u_∗ is the shear velocity of the bed layer [40].

$$u_{*}^2={\displaystyle \frac{g{\mu }_f^2{u}^2}{h^{1/3}}}$$

(11)

μ_f is the Manning coefficient. r_w is the upper interface-to-bed resistance ratio.

4.2. Model Construction

A hydrodynamic model was constructed based on the 2021 measured topographic map, that focused on a typical plain reservoir built on a mountainous sediment-laden river in Xinjiang. The simulation domain encompasses the entire Jiahezi Reservoir area. A triangular mesh was used for spatial discretization. The density of the computational mesh has a direct impact on convergence stability and the calculation results. A numerical simulation of the Jiahezi Reservoir was conducted with three distinct mesh sizes, containing approximately 40.64 thousand, 150.6 thousand, and 403.9 thousand elements, respectively. A consistent approach to mesh generation and simulation was employed for all operating conditions. To balance model accuracy and computational efficiency, grid refinement was applied at key hydraulic locations, including the intake sluice, flood discharge sluice, east and west water discharge sluices, and main channel. In these key areas, the grid spacing was refined to 15 m, 7.5 m, and 5 m. The results indicate that with a mesh count of 40.64 thousand, the simulation results deviated significantly from the experimental model values. However, when the mesh count was increased to 150.6 thousand, the simulated results showed good agreement with the experimental data. Further increasing the mesh count to 403.9 thousand elements yielded results essentially identical to those obtained with the 150.6 thousand mesh. Therefore, a mesh count of 150.6 thousand is sufficient to satisfy the requirement for grid independence. The numerical simulation validation results are shown in Figure 4 and intake sluice total bed thickness changes is shown in Figure 5.

In the end, the grid spacing was refined to 7.5 m and the triangular elements were constructed with a side length of 5 m. The model consisted of 75,580 nodes and 150,601 triangular elements. The specific grid refinement layout is shown in Figure 6, while Figure 7 presents the imported simulation software interface—a 2D diagram of sediment deposition in the simulated reservoir area.

4.3. Boundary Conditions

The parameter settings for the hydrodynamic module are summarized in

Table 1. Key Parameter Settings for the Hydrodynamic Module.

Critical Parameter Adjustments for Hydrodynamic Simulations
Dry–Wet Boundary	Water Depth (m)
Drying Depth	0.005
Flooding Depth	0.05
Wetting Depth	0.1 Coefficients
Eddy Viscosity Coefficient	0.28
Manning Coefficient	32 (m−3/s)

.

In the hydrodynamic model, boundary conditions were configured to reflect the actual operational conditions of the Jiahezi Reservoir. Jiahezi Reservoir is primarily designed for flood control and irrigation, and does not serve power generation purposes. Consequently, its dispatch mode does not require frequent discharge regulation, resulting in notably stable outflow rates within monthly cycles as a key characteristic of its release pattern. Specifically, the intake sluice was defined as the inflow boundary, with water and sediment inputs specified by measured flow rate, sediment particle size, and concentration. The flood discharge sluice, east outlet sluice, and west outlet sluice were designated as outflow boundaries. All other reservoir boundaries were treated as solid-wall boundaries with no-flux conditions.

The sediment transport module was configured based on sedimentation conditions. Field sampling indicates that particles <0.5 mm account for 99% of the sediment reaching Jiahezi Reservoir. In order to better align with practical needs and given its satisfactory application outcomes, previous discussions on this aspect have been relatively limited. This paper specifically provides supplementary research on this issue. Under the same operating conditions, this study set the maximum particle size to 0.5 mm, 0.8 mm, and 1.0 mm, respectively, to compare the sedimentation conditions in the reservoir. The positions of No. 1, No. 3, and No. 8 are shown in Figure 8.

According to the results shown in Figure 9, under the condition of constant working conditions, the maximum sedimentation error caused by particle sizes of 0.5 mm and 1 mm is only 10%.

The settling velocity for sediment within this size range was calculated and used as input for the sediment transport simulation.

Table 2. Key Parameter Settings for the Sediment Module. Note—Five distinct sediment classes were defined, along with two layers—soft and hard—to simulate erosion and deposition processes in the reservoir.

Critical Parameter Adjustments for the Sediment Module
Sediment Particle Density	2650 (kg/m3)
Sediment Critical Shear Stress	0.07 (N/m2)
Sediment Classes	5
Number of Layers	2

summarizes the key parameter settings for the sediment module.

4.4. Model Validation

Model validation was performed using water level and flow velocity verification and topographic comparisons. Simulated water levels and flow velocity on the 5th day of each month from 2021 to 2022 were compared with corresponding measurements to assess accuracy. The comparison showed that both the simulated and observed water levels exhibited similar fluctuations—rising, then falling, and rising again. Overall, the simulated values closely matched the measured data, with a maximum error of 1.05% in April 2022. The result is shown in Figure 10.

Flow velocity verification compares the monitored flow velocity cross-section in the reservoir inflow area of the hydrologic station with the simulated cross-Section 0 the location of cross-section 0 is shown in Figure 8 the two sections are located 20 m apart along the longitudinal direction (y-axis) and 10 m apart along the transverse direction (x-axis), with the channel between them being straight. The two cross-sections exhibit morphology, as shown in Figure 11. Therefore, the flow velocity measured at the monitoring section can be taken as a representative value for cross-section 0. The velocity values for section 0 were extracted from the simulation results, the maximum relative error for flow velocity is merely 8.8%, within the acceptable error margin and the two sets of data show close agreement, as illustrated in Figure 4.

Topographic values measured at all sampling points in August 2022 also aligned well with the simulation, and the patterns of scour and siltation were largely consistent. Sampling locations are shown in Figure 8. These results demonstrated strong agreement, confirming that the simulation achieved the required accuracy for the model. Refer to Figure 5 and

Table 3. Simulated and measured topographic and morphological changes.

No.	Sampling Point Elevation (m)			Siltation Thickness (m)
	Original Data (2021)	Measured Value (2022)	Simulated Value (2022)	Simulated Value (2022)	Measured Value (2022)
1	398.30	395.582	395.48	−2.82	−2.718
2	398.63	395.372	396.406	−2.224	−3.258
3	398.18	403.325	402.19	4.01	5.145
4	397.30	398.200	397.70	0.4	0.9
5	396.66	396.205	395.13	−1.53	−0.455
6	396.36	396.980	397.05	0.69	0.62
7	395.82	396.757	396.93	1.11	0.937
8	395.43	394.73	395.01	−0.42	−0.7
9	396.71	397.85	397.05	0.34	1.14
10	394.88	396.44	397.03	2.15	1.56

for a visual summary of the validation data.

4.5. Data Coverage

The hydrological and sediment datasets used in this study covered the period from 2021 to 2024, which included both typical and high-flow years within the Manas River Basin. Although the absence of a longer observational record limited the ability to assess interannual variability, the selected period captured the representative hydrological and sediment transport regimes that governed deposition in the Jiahezi Reservoir. The inclusion of dry-season (2021) and wet-season (2023) sediment measurements allowed the model to reproduce the contrasting conditions associated with baseflow and flood inflows. Previous studies have shown that for data-scarce reservoirs, calibration based on representative hydrological years can provide reliable estimates of spatial deposition patterns and sediment gradation when inflow events dominate sediment dynamics [41]. Therefore, the present simulations remained valid for assessing short- to medium-term depositional behavior and operational impacts under current hydrological conditions. Future studies that integrate longer-term or extreme-event data would further refine projections of cumulative sedimentation and reservoir longevity. While direct hydrodynamic measurements such as ADCP velocity profiles or in-lake suspended sediment sampling were not feasible during field campaigns, model validation was performed using long-term water-level records and terrain elevation data. The sediment cores integrate approximately 50 years of deposition, capturing cumulative inflow, transport, and sedimentation processes, and thus provide a robust indirect check on model performance. Such core-based validation is widely accepted in reservoir sedimentation studies when continuous in-lake measurements are unavailable [18]. This approach allows reliable simulation of spatial sedimentation patterns and long-term trends, while acknowledging that short-term hydrodynamic fluctuations and extreme events are less constrained.

5. Reservoir Simulation Operating Conditions

As shown in Figure 2, the measured reservoir inflow from 2021 to 2024 was used to define representative operating conditions for simulation. This analysis identifies the period from September to October 2023 as a key representative hydrological condition during the high-flow season. Its representativeness is primarily demonstrated in two aspects: first, the extreme flow magnitude, with monthly average discharge approximately four times the annual mean; second, the pronounced flow variability, as it is the period with consecutive monthly runoff records within the year during which the inter-monthly flow difference reaches its peak. The period from February to March 2021 serves as a representative low-flow condition for the following reasons: it coincides with the initial stage of seasonal snowmelt, during which runoff begins to form but remains limited in volume, accurately reflecting the characteristics of low discharge and snowmelt-dependent water supply typical of dry seasons. Moreover, its flow process exhibits relatively stable and low variability, effectively representing the hydrological stability of low-flow periods.

Two distinct hydrological periods were identified:

(1)

Wet Season (September–October 2023)—This period, which is marked by the transition from flood to normal flow conditions, is characterized by sharp inflow fluctuations and is considered highly representative of wet-season dynamics. The corresponding sediment concentration was 1.56 kg/m³, with an initial water level was 398.68 m. The sluice openings at the west and east outlets were 0.95 m and 0.27 m in September, and 0.14 m and 0 m in October, respectively.

(2)

Dry Season (February–March 2021)—This interval exhibits consistently low inflow rates, closely reflecting actual dry season reservoir operations. The measured sediment concentration was 0.035 kg/m³, with an initial water level of 400.02 m. The sluice openings at the west and east outlets were 0.17 m and 0 m in February, and 0.15 m and 0 m in March, respectively.

Both scenarios were simulated using the hydrodynamic and sediment transport model detailed above. The results are presented and discussed below.

6. Results and Analysis

6.1. Sampling Results and Analysis

6.1.1. Particle Size Distribution Along the Flow Direction and Depth Profile

Sediment samples collected from the Jiahezi Reservoir were classified according to the Standard for Engineering Classification of Soil (GB/T 50145-2007) [42]. The sediments were grouped into five categories based on particle size: clay, silt, fine sand, medium sand, and coarse sand. A stratified weighted-averaging approach was used to classify the sediment components in all five profile layers at each sampling point along the river channel (Sites 1–5). The results are shown in Figure 12. Although the particle size data were obtained from a single field campaign in 2021, the sediment core represents approximately 50 years of cumulative deposition in the Jiahezi Reservoir, encompassing multiple wet, dry, and average years. Sediment core analysis inherently integrates temporal variability, capturing long-term trends in sediment composition rather than reflecting only year-to-year fluctuations. This approach has been widely applied in reservoir sediment studies to infer historical changes in sedimentation and particle size trends [43,44]. Therefore, the sampled core provides a robust basis for evaluating long-term sedimentation patterns and assessing reservoir sedimentation risk.

Analysis of the sampling results indicates that sediment within the Jiahezi Reservoir is dominated by fine sand and silt, with an overall trend toward finer grain sizes. The average particle size decreases with increasing depth. In each vertical profile layer, silt consistently exhibits the highest content, with values at 1, 2, 3, 4, and 5 m depths recorded as 47.35%, 48.21%, 46.5%, 49.08%, and 50.6%, respectively. This indicates a pattern of initial increase, decrease, and subsequent increase in silt content with depth. Silt remains the dominant fraction across all sampled layers.

The fine sand content exhibits a different pattern. It increases from 23.81% at 1 m to a peak of 35.45% at 2 m, then decreases to 33.01% at 3 m, 26% at 4 m, and 22.4% at 5 m. At the 5 m depth, both silt (50.6%) and clay (16.9%) contents reach their maximum values, indicating that this layer contains the finest sediment in the profile.

To simplify the analysis of coarser sediment transport, coarse sand (0.5 < d ≤ 2 mm) and medium sand (0.25 < d ≤ 0.5 mm) were combined into a single category termed “coarse and medium sand.” These coarser fractions account for 16.85% at 1 m, decrease sharply to 7.61% at 2 m and 7.99% at 3 m, then slightly rise to 8.68% at 4 m and 10.1% at 5 m, demonstrating a trend of initial decrease followed by a gradual increase.

Together, silt and fine sand make up 76.48% of the total sediment, confirming the predominance of finer particles in the reservoir deposits. However, because coarse and medium sand are prioritized during dredging, their distribution provides critical information for dredging and sediment management. To quantify this, a parameter M was defined as the sum of coarse and medium sand contents across all depth layers at each sampling point.

Specifically, for each site, sediment contents were normalized to 100% per layer. Parameters A, B, C, D, and E represented the particle assemblages at the 1, 2, 3, 4, and 5 m depths, respectively, while values 1 and 2 denoted the content of coarse sand and medium sand, respectively. Thus, the total content of coarse and medium sand at a given site was calculated as:

M = A1 + B1 + C1 + D1 + E1 + A2 + B2 + C2 + D2 + E2.

As shown in Figure 13 the calculated M values for sampling points 1–5 are 1.73, 0.93, 0.04, 0.004, and 0, respectively, which represents the stretch from the intake sluice to the east/west outlet sluices. These results reveal a clear decreasing trend in coarse and medium sand along the flow direction. Point 1, located nearest the headgate, contains the highest concentration, approximately double that of point 2. From point 3 onward, the coarse fraction rapidly diminishes and is essentially absent in the sample from the sluice. This distribution suggests that coarse and medium sand particles are predominately deposited near the headgate, where flow velocities are highest. It is thus inferred that as the flow proceeds downstream through the reservoir, hydrodynamic energy decreases due to increasing resistance along the thalweg, leading to progressive hydraulic sorting. Due to their larger mass and higher settling velocities, coarser particles settle earlier in the transport process, resulting upstream deposition and near-total depletion further into the reservoir.

6.1.2. Particle Size Distribution Throughout the Jiahezi Reservoir

Particle size analysis across all sampling sites within the Jiahezi Reservoir shows that deposited sediments are primarily distributed within the 0.00982–0.234 mm range, accounting for 91.66% of the total mass. The bulk composition consists mainly of silt and fine sand. The characteristic particle diameters were determined to be as follows: d₁₀ = 0.003 mm, d₃₀ = 0.013 mm, d₅₀ = 0.027 mm_, and d₆₀ = 0.031 mm. The coefficient of uniformity (Cu), calculated as Cu = d₆₀/d₁₀ = 10.33, indicates a wide particle size distribution (Cu > 5), reflecting substantial variation between coarse and fine particles. The coefficient of curvature (Cc), calculated as Cc = (d₃₀)²/(d₁₀ × d₆₀) = 1.872, falls within the range of 1 and 3, suggesting a smooth and continuous gradation curve. The particle size distribution curve for sediments in the Jiahezi Reservoir is depicted in Figure 14.

The particle sizes at all reservoir sampling points were categorized according to the classification standards in GB/T 50145-2007 and compared to samples obtained directly from the river.

6.2. Flow Velocity Field Variations

6.2.1. During the Wet Season

Based on the flow velocity variations observed within the reservoir (Figure 15), six representative locations between the intake sluice and the east outlet sluice were selected to characterize spatial trends in the reservoir’s flow velocity field. The analysis is summarized as follows:

The region extending from the intake sluice to the flood discharge sluice exhibits the highest streamline density and count, the maximum flow velocity is 1.28 m/s. Velocities consistently remain high, averaging above 0.5 m/s.

Area ①—Flood Discharge Sluice: the average flow velocity rapidly drops to below 0.2 m/s. Compared to the section from the intake sluice to the flood discharge sluice, the average flow velocity rate has decreased by 60%. Area ②—Mid Channel Section: channel narrowing at this location increases streamline density and flow velocity. Velocities at this point consistently exceed 0.1 m/s. Area ③—Reservoir Bending Zone: flow is fastest in the center of the channel, where the cross-section narrows, and slower along the flanks where the channel widens. The flow velocity in the middle of the river is 0.041 m/s. The flow velocity on the left bank is 0.035 m/s, whereas on the right bank it is 0.022 m/s. Area ④—West outlet Sluice: this zone lies at a lower elevation than Area ③. flow velocity the flow velocity is 0.56 m/s. Area ⑤—East outlet Sluice: reaching 0.29 m/s. Area ⑥—East outlet Sluice Backwater Zone: flow velocities were predominantly (85%) in the range of 0.09–0.003 m/s, with very few measurements below 0.003 m/s. In terms of streamline distribution, the mid-section appeared sparse, while the front and reservoir tail sections showed greater concentration, the highest being in the front of area.

6.2.2. During the Dry Season

Based on the flow velocity variations observed within the reservoir (Figure 16), Consistent with the high-flow period, six typical cross-sections were selected to analyze the flow velocity during the low-water period.

The pattern of flow lines and velocity is consistent with that during the high-flow period. The peak flow velocity reach 0.52 m/s, which is 40.6% of the flow velocity during the high-flow period.

Area ①—Flood Discharge Sluice: the flow velocity downstream of the flood discharge sluice is reduced by 50% compared to the upstream velocity, rapidly decreasing to below 0.15 m/s. Area ②—Mid Channel Section: channel narrowing at this location increases streamline density and flow velocity. Velocities at this point consistently exceed 0.07 m/s. Area ③—Reservoir Bending Zone: the flow velocity in the middle of the river is 0.03 m/s. The flow velocity on the left bank is 0.035 m/s, whereas on the right bank it is 0.012 m/s. Area ④—West outlet Sluice: the flow velocity is 0.17 m/s. Area ⑤—East outlet Sluice: reaching 0.065 m/s. Area ⑥—East outlet Sluice Backwater Zone: flow velocities were predominantly (73%) in the range of 0.06–0.001 m/s.

6.3. Scour and Deposition Characteristics in Reservoir Channels

To visualize sediment transport dynamics under high-flow conditions, this study analyzed longitudinal variations in bed elevation by substituting simulated elevations with mean cross-sectional elevations. The longitudinal profile extends from the upstream face of the intake sluice to the east sluice along the primary flow direction, as shown in Figure 8.

During the operation of the reservoir, the scouring and sedimentation trends in the reservoir’s river channel exhibit a high degree of consistency between the wet and dry seasons. Specifically, at extraction point 1, the degree of scouring during the dry season is only half that of the rainy season. Data from extraction points 3 to 10 reveal that the trend of sedimentation gradually decreases with increasing location. Under wet season conditions, the sedimentation depth at point 5 is only 13% of that at point 1, and this proportion slightly increases to 14% under dry season conditions. It is noteworthy that point 2 has the most severe sedimentation depth, while point 8 exhibits the least sedimentation depth. The results are shown in Figure 17.

6.4. Spatial Distribution of Sediment Size Fractions

6.4.1. During the Wet Season

As shown in Figure 18, sediment concentration patterns vary significantly across particle size classes during high-flow periods. Particles 0.5 and 0.0252 mm in diameter exhibit peak concentrations primarily in front of the flood discharge sluice and at the east outlet sluice locations. In contrast, the finest particles, with diameters approximating 0.001729 mm, show maximum concentrations in the backwater zone near the east outlet sluice.

Particle sizes 1 and 2 (D₅₀ = 0.50 mm and 0.364 mm) are mainly concentrated upstream of the intake sluice and near the east outlet sluice within the main channel. Their concentrations increase along the channel toward the flood-discharge sluice, reaching 0.012 kg/m³ for particle size 1 and 0.021 kg/m³ for particle size 2. All particle sizes 1–4 attain peak values at the junction of the east outlet sluice and its backwater zone (0.140, 0.195, 0.132, and 0.116 kg/m³, respectively).

Particle size 3 (D₅₀ = 0.142 mm) exhibits the highest concentration near the flood-discharge sluice (0.068 kg/m³), decreasing radially outward; it first appears in the backwater zone of the east outlet sluice.

Particle size 4 (D₅₀ = 0.0252 mm) shows markedly higher concentrations within the main channel than the three coarser sizes, at 2–3 times that of size 3.

Particle size 5 (D₅₀ = 0.00173 mm) can be transported to the dam-face tailwater by flow; its high-concentration zone is biased toward the backwater area, reaching a maximum of 0.396 kg/m³ at the right end of that region.

The total suspended sediment concentration (TSSC) equals the sum of all five particle-size concentrations. In the backwater area at the reservoir tail, TSSC increases primarily due to size 5, which accounts for more than 70% of the load; near the intake sluice, coarser particles (sizes 1–3) dominate, comprising over 50%.

Approximately 80% of particle sizes 1 and 2 (D₅₀ = 0.50 mm and 0.364 mm) are concentrated within 1000 m along the river channel from the intake sluice, while size 5 particles (D₅₀ = 0.00173 mm) can be transported in large quantities to the reservoir tail area, with a travel distance exceeding 7000 m.

6.4.2. During the Dry Season

As shown in Figure 19. Particles of sizes 1–4 (D₅₀ = 0.50 mm, 0.364 mm, 0.142 mm, and 0.00173 mm) are mainly concentrated in the main channel upstream of the inlet sluice. Their concentrations increase along the channel toward the flood-discharge sluice, reaching 0.016 kg/m³ for size 1, 0.0195 kg/m³ for size 2, 0.021 kg/m³ for size 3, and 0.056 kg/m³ for size 4 near the east outlet sluice.

Particle size 4 (D₅₀ = 0.0252 mm) Compared to the first three particle sizes, the sediment concentration in the river channel increased significantly. In the channel connecting the intake sluice and the flood sluice, the sediment concentration (0.0326 kg/m³) is approximately 5 times that of particle size class 3.

Particle size 5 (D₅₀ = 0.00173 mm) unlike the first four particle size classes, particle size class 5 reached a peak concentration of 0.16 kg/m³ near the east outlet sluice. It was also the only class distributed within the east sluice’s backwater area, where its concentration measured 0.14 kg/m³.

The total suspended sediment concentration (TSSC). The sediment concentration near the flood outlet sluice reached its peak (0.184 kg/m³), measuring 1.5 to 2 times higher than that in the upstream connecting channel (the section between the intake and flood outlet sluice). The sediment concentration in front of the east outlet sluice is 0.16 kg/m³, while that in the backwater zone of the east outlet sluice measures 0.14 kg/m³.

6.5. Scouring and Silting Variations in the Reservoir Area

6.5.1. During the Wet Season

Under high-flow conditions, sediment deposition had extended from the inlet sluice to the east outlet sluice by the 17th hour of reservoir operation, representing a 56% reduction in the time required for deposition to reach the same area compared to dry season conditions. At this stage, the scouring depth in front of the inlet sluice reached 0.4 m, while the sedimentation thickness in front of the flood discharge sluice was 0.3 m. Sediment deposition of 0.12 m occurred at the west outlet sluice, while a deposition of 0.03 m occurred at the east outlet sluice. As the time progressed, when it reached 85 h under the wet season condition. Sediment deposition had not yet occurred in the backwater area of the east outlet sluice. Sediment deposition has occurred in the backwater area of the east outlet sluice, with the thickness ranging from 0.3 m to 0.009 m. Additionally, there are two localized areas of scouring. Within the red box, the peak scouring depth on the left side is 0.20 m, while on the right side it is 0.03 m. The area of the left scouring zone is approximately four times that of the right side. During this phase, the scouring near the intake gate reaches 0.67 m, and the deposition in front of the flood discharge reaches 0.63 m. Sediment deposition of 0.14 m occurred at the west outlet sluice, while a deposition of 0.15 m occurred at the east outlet sluice. By the 150th hour, the sediment deposition range in the backwater area of the east outlet sluice had expanded compared to the earlier operational stage, reaching 1.2 times the original extent. Concurrently, the peak deposition thickness in this area increased to 0.26 m. Sediment deposition of 0.2 m occurred at the west outlet sluice, while a deposition of 0.31 m occurred at the east outlet sluice.

Time-varying diagrams illustrating sedimentation dynamics during the high-flow period are presented in Figure 20.

6.5.2. During the Dry Season

The simulation period during the dry season matched that of the wet season. However, the prevailing hydrological conditions had a pronounced impact on sediment transport dynamics. Compared to the wet season, inflow discharge at the intake sluice were significantly reduced by a factor of more than 10, leading to weakened hydrodynamic forces and a corresponding decline in sediment transport capacity. Under these conditions, a significant reduction in scouring effects is demonstrated by the scour depth around the intake and discharge flood sluices dropping to about half its wet-season value and by an approximately one-third decrease in the total erosion area across the reservoir. Throughout the entire reservoir, sedimentation depth typically ranges from one-half to one-third of the values during the flood season. Measured depths at key structures—the flood sluice, west outlet sluice, and east outlet sluice—are 0.5 m, 0.13 m, and 0.28 m, respectively, indicating substantial sediment deposition and are significantly elevated compared to other zones, where depths remain below 0.1 m. Furthermore, the sedimentation area in the east sluice backwater region constitutes approximately half of its dry-season extent. While the scouring-silting equilibrium zone expanded significantly, compared to the wet season, the zone of sediment equilibrium has migrated. Refer to the white area in the river channel within the red box above Figure 21c.

In the river bend section, sediment transport exhibits pronounced lateral differentiation: the convex bank experiences almost no deposition, with an average thickness of approximately 0.05 m; meanwhile, localized scouring occurs along the concave bank, with a maximum scour depth of up to 0.35 m. Within the inner part of the channel, both scouring and deposition processes coexist, though deposition remains the dominant mechanism overall. Near the east outlet sluice, there exists a sediment transport equilibrium zone. The frontal part of this area is subject to scouring, with a depth of approximately 0.1 m, while corresponding sedimentation occurs in the rear part, with a thickness of about 0.09 m. Overall, sediment transport within this zone exhibits dynamic equilibrium, indicating a state of scour and deposition stability. The backwater area at the reservoir tail exhibited almost no sediment accumulation and deposition thickness is concentrated between 0.003 m and 0.0001 m, decreasing with increasing distance.

Due to low sediment concentrations and weaker hydrodynamic conditions, sediments were unable to reach the reservoir tail, resulting in substantially reduced extent and sedimentation thickness. The average sediment accumulation was ~ 10 cm—markedly less than during the wet season— demonstrating the critical influence of hydrological conditions on sediment transport behavior.

In summary, sediment transport during the dry season was characterized by an expanded scouring-silting equilibrium zone, reduced deposition extent and thickness. Temporal variations in sedimentation during the dry season are illustrated in Figure 21.

7. Discussion

7.1. Hydrodynamic Sorting of Sediments from River Channel to Reservoir Analysis

Results show that sediments within the reservoir are generally finer than those sampled directly from the river channel. This conclusion is consistent with the findings of Hagstrom, C. A. et al. [45]. Classification and comparison details are provided in

Table 4. Comparative Analysis of Sediment Grain Sizes in the River Channel (N) and Reservoir (M).

Sampling Points	Coarse Sand (%) (0.5 < d ≤ 2 mm)	Med. Sand (%) (0.25 < d ≤ 0.5 mm)	Fine Sand (%) (0.075 < d ≤ 0.25 mm)	Silt (%) (0.005 < d ≤ 0.075 mm)	Clay (%) (d ≤ 0.005 mm)
(N)	1.6	6.9	19.3	57.9	14.2
(M)	1.6	8.7	28.1	48.3	13.3

, where “M” represents all the river channel sample points and “N” all the reservoir sample points. The coarse sand content for both M and N is 1.6%. The medium sand content is 8.7% for M, higher than 6.9% for N; while the fine sand content is 28.1% for M, compared to 19.3% for N. These values indicate that sediments at the river channel sampling points (M) are generally coarser than those in the reservoir area (N).

This difference is attributable to hydrodynamic processes during sediment transport [46]. As sediment-laden water flows from the river channel into the reservoir interior, flow resistance increases, resulting in a gradual decrease in velocity and sediment-carrying capacity. Once the flow velocity drops below the settling velocity of larger particles, these particles settle [47,48]. This mechanism explains the observed coarser grain sizes at river sampling points (M) relative to the finer sediments farther within the reservoir (N).

The identical proportion of coarse sand content in both M and N is likely due to exceptionally high-energy hydrodynamic conditions present during flood periods. This conclusion is consistent with the findings of Brouwer, R.L. et al. [49]. Under such conditions, frictional resistance along the flow path is minimal, rendering the blocking and decelerating effects on sediment negligible [50]. Consequently, coarse sediment is more uniformly distributed throughout the reservoir, showing limited differential deposition and reflecting a relatively homogeneous sedimentation pattern.

7.2. Mechanisms of Hydrodynamic Change in the Reservoir Area

The spatial variation in reservoir flow velocity observed in this study is primarily governed by two types of hydrodynamic mechanisms: first, the increase in flow velocity caused by energy input and localized topographic concentration effects; second, the decrease in flow velocity resulting from energy dissipation due to boundary resistance, frictional losses along the flow path, and cross-sectional expansion. The specific driving mechanisms for each region are as follows:

• Primary Factors Leading to Increased Flow Velocity

The increase in flow velocity mainly stems from the efficient conversion of gravitational potential energy and the spatial concentration of flow lines. Acceleration via Gravitational Potential Energy Conversion: in region ④—the west outlet sluice, a sudden topographic drop concentrates and converts the gravitational potential energy of the flow into kinetic energy, representing the most significant acceleration mechanism within the entire reservoir area. Localized Kinetic Energy Gain: region ⑤—the east outlet sluice also benefits from this type of potential energy conversion. However, due to the smaller flow-cross section of the sluice, the acceleration effect is relatively limited.

2.

Primary Factors Leading to Decreased Flow Velocity

The attenuation of flow velocity is dominated by multiple energy dissipation pathways during the flow process, which can be summarized as follows: Sudden Boundary Dissipation: in region ①—the area in front of the flood discharge sluice, the direct impact of flow against the closed gate causes significant localized head loss due to abrupt flow direction changes and intense turbulence [51]. Sustained Frictional Losses Along the Flow Path: in region ②—the Mid-Channel Curved Reach, transverse circulation intensifies turbulence and bed friction, continuously dissipating longitudinal kinetic energy [52]. In region ⑥—the Backwater Area of the east outlet sluice, the longest flow path results in the greatest cumulative frictional resistance [53], making it the region with the lowest average flow velocity. Energy Dispersion Due to Cross-Sectional Expansion: in region ③—Channel Widening Zone Area, a significant expansion of the channel cross-sectional area (approximately 40%) leads to a decrease in kinetic energy per unit volume of water, consequently reducing the flow velocity [54].

7.3. Erosion-Deposition Zones in the Reservoir

This study uses the complete hydrological event process as the fundamental time scale to classify the sedimentation and erosion dynamics in the reservoir area. The defined regional types are not static but undergo dynamic transformation in response to variations in hydrological conditions. The specific classification criteria based on sedimentation and erosion characteristics are as follows: the scour-sedimentation transition zone (Following the hydrological event, the riverbed in this area exhibited a dynamic alternation between erosion and deposition), scouring-silting equilibrium zone (Following the hydrological event, the riverbed in this area exhibited neither erosion nor deposition, with its elevation remaining constant [49]), scouring zone [55] (Following the hydrological event, this area exhibited riverbed degradation), and deposition zone [4] (Following the hydrological event, this area exhibited riverbed uplift).

Scour-sedimentation Transition Zones. As sediment-laden water enters, deposition occurs when the sediment concentration exceeds the flow’s carrying capacity, causing the riverbed to rise. After the deposited layer exceeds a critical thickness, altered flow conditions (e.g., increased velocity) may cause the sediment-carrying capacity to exceed the sediment load, triggering erosion and downstream sediment transport. Thus, these areas exhibit alternating phases of erosion and deposition.

Scour-Silting Equilibrium Zones (In this paper, a region is defined as a sediment equilibrium zone if its elevation remains unchanged from the initial to the final stage throughout the entire process of a given working condition.)—Located between deposition and scouring zones, these areas maintain a dynamic balance between erosion and deposition, where net accumulation is negligible. This conclusion is consistent with the findings of Sedlácek, J. et al. [56]. This balance results from either (a sediment from upstream passing through without settling) or (a state in which erosion from the riverbed matches incoming deposition, yielding no topographic change.) These zones are characterized by morphological stability.

Scouring Zones—Bed scouring dominates the main channel between the intake sluice and the flood discharge sluice, excepting the area immediately adjacent to the sluice. Scouring intensifies closer to the intake sluice due to stronger hydrodynamic forces, including higher flow velocity and sediment-carrying capacity, leading to severe scouring in the upstream main channel, with scouring depths increasing over time [57,58].

Deposition Zones—In addition, “Scaly” sediment deposition patterns appear in the outer channel sections, especially in low-lying depressions identified in pre-operation topographic maps, which act as natural sediment traps. Moreover, changes in flow direction at flow diversion junctions induce a rapid decline in hydrodynamic forces, subsequently reducing flow velocity and sediment-carrying capacity, and promoting extensive localized deposition.

8. Conclusions

This study investigated sediment particle size distribution in the Jiahezi Reservoir through field sampling and analysis of existing hydrological data. A 2D sediment-water model was applied to simulate sediment transport under two representative operational scenarios—wet season and dry season. The key findings are as follows:

Sediment Particle Size and Distribution—Sediment particle sizes generally decrease along both the river’s flow direction and the longitudinal excavation axis. The sediment is predominantly fine-grained, with a median particle size (d₅₀) of 0.027 mm. Particles with diameters ranging from 0.005 to 0.25 mm account for 76.48% of the total mass, constituting the primary component of reservoir sedimentation.

Influence of Hydrological and Topographical Conditions—Hydrodynamic and topographic conditions play a critical role in sediment transport and deposition patterns. The hydrodynamic sorting effect leads to a broader spatial distribution for finer particles. During the wet season, sediment particles with a grain size of 1–3 reach their concentration peak at the east outlet sluice. In the dry season, this peak shifts to the flood discharge sluice, with no distribution observed in front of the east outlet sluice. Meanwhile, the overall suspended sediment concentration in the backwater area of the east discharge sluice during the wet season is approximately 3–4 times that of the dry season.

Sediment Transport Under Different Operating Conditions—The sedimentation patterns within the reservoir area are generally similar during both the wet and dry seasons, and can be broadly categorized into four distinct zones: the scour-sedimentation transition zone, scouring-silting equilibrium zone, scouring zone, and deposition zone. The intensity of both scouring and sedimentation observed during the dry season is generally about 1/3 to 1/2 of that during the wet season. Sedimentation at the reservoir tail is significantly reduced compared to that near the flood discharge sluice area. During the dry season, a scouring-silting equilibrium zone exists in the connecting channel between the east outlet sluice and its backwater area.