Reconstructing Nanwang hub operations on Ming-Qing Grand Canal with HEC-RAS

Abstract

The Nanwang Water Diversion Hub, a core hydraulic heritage in the Ming-Qing Beijing-Hangzhou Grand Canal’s Huitong River section, was vital for water regulation and navigation. Using HEC-RAS, we constructed a numerical restoration model to simulate its impact on local hydraulics, water management, and navigation capacity. Simulations show the Hub crucially regulated water levels, improved navigation, and reduced riverbed erosion. By elevating levels, stabilizing flow, and mitigating hydraulic jumps, it significantly enhanced navigation safety. Precise energy distribution management boosted the canal’s diversion, storage, and drainage capacities, enabling sustainable water supply. Matching an optimal energy gradient effectively reduced local shear stress, inhibited erosion/sedimentation, strengthened channel stability, and prolonged canal lifespan. This study quantitatively reveals ancient composite hydraulic system mechanisms and provides a scientific basis for digital conservation, functional reconstruction, and sustainable utilization of this World Heritage canal. The methodology is applicable to other ancient canal heritage conservation.

Introduction

As the world’s longest artificial canal, the Beijing-Hangzhou Grand Canal was not only a crucial water transport artery in ancient China but also a linear cultural heritage of immense historical, cultural, and technological value globally¹. Constructed and expanded over multiple dynasties since the 5th century BC, the canal reached its zenith during the Ming and Qing dynasties (1368–1912). Serving as a north-south transportation artery, it fostered economic, cultural, and social prosperity along its banks. The canal’s hydraulic engineering was key to ensuring water resource management and maintaining the smooth flow of waterborne transport (grain tribute). The Nanwang Water Diversion Hub is located at the highest point of the canal, in a watershed area with significant elevation differences in the riverbed. The water engineering facilities are highly concentrated and technically complex, making it one of the most challenging and representative engineering projects along the canal. In contrast to typical international canal projects such as the Canal de MIDI in France, the Grand Union Canal in the UK, which are primarily constructed in flat terrain regions. Their hydraulic structures are primarily designed with navigation as the core objective, featuring simplified water resource allocation and structural layout, primarily relying on continuous locks to regulate water levels, with few complex hydraulic structures integrated into elevation-based control hubs. The Nanwang Water Diversion Hub not only plays a crucial role in controlling water flow and stabilizing water levels but also provides reliable support for agricultural irrigation and water resource management in the surrounding areas, demonstrating higher functional integration and hydraulic facility integration. This combination of technical and functional characteristics gives it unique value within the global linear water heritage system^2,3.

With advancements in hydraulic engineering technology and growing emphasis on canal heritage conservation^4,5, international scholars are increasingly focusing on canal hydraulic facilities and their impact on Hydraulic Characteristics of River^6,7. Internationally, extensive research achievements exist on canal systems⁸. For instance, France’s Canal du Midi^9,10 and Wales’s Pontcysyllte Aqueduct^11,12, hailed as “impossible engineering”, have consistently received high attention from international scholars regarding their technology and site conservation. Additionally, significant research has accumulated on other major canal hydraulic heritages, such as Canada’s Rideau Canal^13,14, Spain’s Castilla Canal^15,16, and the United States’ Erie Canal^17,18, primarily concerning water resource management, canal heritage operation and maintenance, along the canal biodiversity, and tourism development^19,20,21. Most of these studies focus on the spatial patterns along canals, interactions between canals and ecosystems, and adaptive reuse of heritage, illustrating the close connection between canals and contemporary socio-economics^22,23.

In contrast, the hydraulic facilities of the Beijing-Hangzhou Grand Canal form the core component of its hydraulic heritage. Some domestic Chinese scholars have approached this from a macro perspective^24,25. For example, scholars have used Gephi to explore the correlation network within the overall hydraulic facility system of the Ming-Qing Beijing-Hangzhou Grand Canal²⁶, employing structural equation modeling to investigate the relationship between the spatial distribution of hydraulic facilities along the canal and natural factors²⁷. Additionally, some scholars have analyzed the spatial evolution and social impact of Ming-Qing canal hydraulic facilities through global spatial autocorrelation^28,29.

Although extensive research on canals has been conducted both domestically and internationally at the macro level^30,31, scholars have utilized methods such as Gephi network analysis and spatial autocorrelation to explore the spatial relationships and functional hierarchies of hydraulic engineering facilities. However, such analyses primarily focus on the macro-level impacts of hydraulic engineering facilities after their construction^32,33. making it difficult to reveal the operational mechanisms of individual hydraulic engineering projects³⁴. Especially when addressing core issues such as navigation safety and sediment transport, these methods cannot quantify key hydrodynamic parameters such as Froude number and shear stress. This limitation is particularly evident in the study of the Nanwang Water Diversion Hub. The operation of the Nanwang Water Diversion Hub relies on complex hydraulic engineering synergistic mechanisms, and traditional macro-level methods struggle to elucidate its specific operational mechanisms.

To address this research gap, this study utilized the HEC-RAS (Hydraulic Engineering and River Analysis System) to restore and reconstruct the Nanwang Water Diversion Hub model. From a micro perspective, by comparing and simulating the hydrological conditions before and after the reconstruction of individual hydraulic structures, the study systematically explored their operational mechanisms and their impact on river hydraulic characteristics and the overall operational capacity of the canal system. This study not only reveals the critical role of restored hydraulic structures in regulating water flow, ensuring canal transportation, and managing water resources but also provides reliable technical support and quantitative evidence for digital and virtual display, thereby laying the foundation for the long-term protection and sustainable utilization of cultural heritage.

Methods

The core objective of this study is to conduct a comprehensive analysis of the operational mechanisms of the Nanwang Water Diversion Hub following its restorative reconstruction, thereby elucidating its role in water flow regulation, river management, and the canal transportation system. This analysis aims to uncover the hydraulic control logic embedded in ancient hydraulic engineering and establish quantitative evaluation criteria for the protection of canal heritage. This study will focus on how the restorative reconstructed hydraulic structures can determine the critical hydraulic thresholds required for structural protection to prevent erosion and damage, and how the Froude number can be used to assess the navigational safety of river sections, aiding in the evaluation of ancient shipping capacity. Building on this, the research will further translate hydraulic analysis results into actionable heritage protection strategies, including cultural heritage reuse, prioritization of critical node repairs, and optimization of visitor flow management.

To enhance understanding of these complex hydrological evolution processes, this study employs graphical visualization techniques to intuitively demonstrate the impact of hydraulic facility operation on river hydraulic characteristics (Fig. 1). Through this comprehensive analysis, the study reveals the critical role of the restored Nanwang Hub in responding to hydrological variations and integrated water resource management, providing a scientific basis for optimizing existing hydraulic facility functions and improving the overall efficiency of comprehensive water resource management in the canal.

Fig. 1: Research framework.

Schematic representation of the research design, including data collection, modeling, and analysis steps.

Overview of the Nanwang Water Diversion Hub

As the highest point along the entire Beijing-Hangzhou Grand Canal during the Ming and Qing dynasties, the Nanwang Water Diversion Hub, known as the “Water Ridge” of the canal³⁵, utilized the dredging of three lakes as a water pivot, constructing sluices and dams to regulate water volume and ensure unimpeded waterborne transport. It stands as one of the most critical hydraulic hubs in the entire canal system, its core function embodied in efficient water resource management and ensuring smooth waterborne transport. Built in the 9th year of the Ming Yongle reign (1411 AD), following detailed surveys and planning by hydraulic engineering experts, a large dam was constructed at Daicun, and a diversion project was established, ingeniously channeling the Wen River westward into the canal system to achieve rational water resource allocation (Fig. 2).

Fig. 2

Ancient map of the Ming-Qing Nanwang Water Diversion Hub (Adapted by the Research Team Source: Yao Hanyuan: History of the Beijing-Hangzhou Canal, P199). Important hydraulic structures are indicated.

At Nanwang Town, the Wen River’s flow was scientifically divided by the hub: seventy percent flowed north into the Zhang and Wei Rivers, while thirty percent flowed south into the Yellow and Huai Rivers. This diversion ratio fully considered the water volume demands of the two major northern and southern water systems, reflecting the high wisdom of ancient hydraulic engineering³⁶. This project not only successfully regulated water volume but also ensured stable waterborne transport, hailed as a great innovation in Chinese hydraulic history.

Currently, the Nanwang Water Diversion Hub site is being transformed into a regional archeological park (Fig. 3), focusing on conserving and showcasing the hydraulic engineering heritage of the Beijing-Hangzhou Grand Canal³⁷. The heritage area is divided into multiple functional zones, including a core conservation zone, a public archeology experience zone, and a historical environment restoration zone, ensuring the long-term preservation of its structure and function through scientific restoration and modern management³⁸.

Fig. 3: Current state of the Nanwang Water Diversion Hub heritage area.

a Daicun Dam; b Liulin Sluice; c Nanwang Diversion Hub Site; d Shili Sluice; e Siqian Sluice; f Xujiankou Water Gate; g Yuankou Sluice; h Changming Water Gate. Photograph showing the present condition of the heritage site, including preserved sluices, dams, and channels.

Spatial layout of the Nanwang Water Diversion Hub

The layout of the Nanwang Water Diversion Hub during the Ming and Qing dynasties demonstrated a complex and sophisticated water resource management system. According to the History of Ming - Treatise on Rivers and Canals, the Huitong River section during the Ming and Qing dynasties stretched 166.2 km from Nanwang Lake north to Linqing, with a drop of 29.43 m; south to Zhenkou (opposite Xuzhou) it was 160.66 km long, with a drop of 37.932 m. This means that the terrain of the Beijing-Hangzhou Grand Canal gradually sloped downwards to the north and south from the Nanwang area, forming a natural flow gradient, making water resource management and regulation particularly crucial within this region. Consequently, a very high density of hydraulic facilities appeared in the reach centered on Nanwang Town (Fig. 4)³⁶.

Fig. 4: Map of Ming-Qing Beijing-Hangzhou Grand Canal hydraulic structure locations.

a Distribution of Water engineering facilities in the Beijing-Hangzhou Grand Canal; b Nanwang Water Diversion Hub Area. Locations of major hydraulic structures such as sluices, dams, and diversion hubs during the Ming-Qing period.

These hydraulic facilities were mainly distributed along the Ming-Qing Huitong Canal, with a few water gates (Doumen) and dams surrounding the North Five Lakes, Nanwang Lake, Shushan Lake, Mata Lake, Songjiawa, and along the Wen River. These facilities collectively formed the Nanwang Water Diversion Hub, safeguarding water resource management and smooth waterborne transport in the Nanwang area, ensuring the long-term stable operation of the Nanwang Hub as the core hydraulic hub of the Beijing-Hangzhou Grand Canal during the Ming and Qing dynasties.

Data sources

To comprehensively reconstruct the layout of the Nanwang Water Diversion Hub, this study initially restorative reconstructed a basic information database of water engineering facilities along the Grand Canal during the Ming and Qing dynasties based on clear data screening criteria. In terms of time scale, the study focused on the Ming and Qing dynasties (1368–1912), as historical records from this period are relatively complete. In terms of data sources, the research team reviewed over 500 historical documents, including official records, local annals, and ancient maps, and excluded documents with low credibility. For each facility location, at least two independent historical sources must corroborate the information. Following these criteria, the research team selected 1146 valid sites (Fig. 4) and collected detailed information on the water engineering facilities, including their names, primary functions, river segments they belonged to, river channel structures, and latitude and longitude coordinates.

The research team utilized the coordinate extraction function of the Ovitalmap Geographic Information Service Platform (www.ovital.com) to successfully obtain the precise geographic coordinates of the hydraulic engineering facilities and classified them in detail based on their functions. To ensure the accuracy of of the restorative reconstruction data, the research team conducted a field survey of the entire Nanwang Water Diversion Hub to further verify the current locations of the hydraulic structures. They also used a DJI Mavic 3 multispectral edition multi-rotor drone to conduct aerial photography—oblique photogrammetry under stable weather conditions, with flight parameters set to 80% forward overlap and 70% side overlap. All images were processed using Smart3D ContextCapture software, with a point cloud density estimated at ~1000 points per square meter, generating high-precision point cloud models. These high-precision 3D point cloud models³⁹ provided reliable data support for the restorative reconstruction and hydraulic simulation of key hydraulic structures (Figs. 5–10).

Fig. 5: 3D point cloud model of Daicun Dam.

The red box highlights the main remains of Daicun Dam.

Fig. 6: 3D point cloud model of Liulin Sluice.

The red box highlights the main remains of Liulin Sluice.

Fig. 7: 3D point cloud model of Nanwang Hub Site.

The red box highlights the main remains of Nanwang Hub Site.

Fig. 8: 3D point cloud model of Shili Sluic.

The red box highlights the main remains of Shili Sluice.

Fig. 9: 3D point cloud model of Siqianpu Sluice.

The red box highlights the main remains of Siqianpu Sluice.

Fig. 10: 3D point cloud model of Xujian Doumen.

The red box highlights the main remains of Xujian Doumen.

In addition, the river network data used in this study was primarily sourced from The Historical Atlas of China. To achieve spatial integration between historical maps and modern digital geographic data, the research team conducted high-resolution scans of the relevant map sheets from The Historical Atlas of China and completed geographic matching processing using ArcGIS 10.8. During the matching process, geographic control points with co-located relationships between historical and modern maps (e.g., streets and alleys in Nanwang Town, water diversion gates, and river mouths) were selected. The Rubber Sheeting Transformation method was used to overlay these co-located geographic control points, transforming the historical map into modern digital elevation model (DEM) data in the WGS84 coordinate system. Based on the registered modern digital elevation model (DEM), combined with the hydrological analysis tools in ArcGIS 10.8, a modern hydrological map highly consistent with the historical river network was successfully extracted. The geographic base map used originates from the EARTHDATA data service platform managed by NASA (https://search.asf.alaska.edu/). The Nanwang Hub topographic map cited in this study was redrawn from Cai Fan’s Hydraulic Engineering of the Beijing-Hangzhou Grand Canal. Furthermore, combining HEC-RAS simulation experiments with terrain and river network data, the team generated data on the hydrological impact of the restored hydraulic structures in the Nanwang area. Specific steps are detailed in the model construction section.

This study encompasses various types of hydraulic facilities built within the Nanwang Water Diversion Hub during the Ming and Qing dynasties, including flow control, water retention, and overflow facilities. Their primary functions were to optimize channel width, depth, flow velocity, and integrated water resource management. In this study, these individual hydraulic facilities form a diversion hub, whose overall operational mechanism will be discussed further in subsequent sections. Specific parameters and dimensions of the Nanwang Hub hydraulic facilities are shown in Table 1.

Table 1 Information on water engineering facilities in the Nanwang Channel

Research methodology

This study utilized HEC-RAS software developed by the US Army Corps of Engineers Hydrologic Engineering Center (HEC) for model computation and simulation experiments⁴⁰. Existing research indicates that HEC-RAS is an advanced hydrodynamic calculation program widely used for simulating and analyzing river hydrodynamics, sediment transport, and water quality changes^41,42,43, while also capable of simulating the impact of hydraulic structures like bridges, culverts, and weirs on flow behavior^44,45 (Fig. 11).

Fig. 11: HEC-RAS research methodology flowchart.

Schematic flowchart showing the steps of hydraulic modeling, data input, calibration, and analysis.

The HEC-RAS model was used to simulate the hydrodynamic characteristics of the restored Nanwang Water Diversion Hub, including flow velocity, water depth variation, and flow impact on the riverbed⁴⁶. Research by scholar Brunner shows that the HEC-RAS model has significant advantages in computational result accuracy; compared to commonly used hydrodynamic models like TUFLOW, MIKE FLOOD, and SOBEK, HEC-RAS offers higher computational efficiency and reliability⁴⁷.

The core algorithm of the HEC-RAS 1D model is primarily based on the energy equation, specifically the Bernoulli equation, which describes energy conservation of water flow in a channel. The basic form of the Bernoulli equation is as follows:

$${Y}_{1}+{z}_{1}+\frac{{a}_{1}{v}_{1}^{2}}{2g}={Y}_{2}+{z}_{2}+\frac{{a}_{2}{v}_{2}^{2}}{2g}+{h}_{e}$$

In the equation: ${Y}_{1}$, ${Y}_{2}$ represent the water depth at cross-section 1 and 2, respectively (m); ${z}_{1}$, ${z}_{2}$ represent the elevation of the main channel at cross-section 1 and 2, respectively (m);${a}_{1}$, ${a}_{2}$ represent the velocity coefficients at cross-section 1 and 2, respectively; g represents gravitational acceleration (m/s²); ${h}_{e}$ represents head loss (m). By analyzing different channel cross-sections, HEC-RAS can simulate water movement under various hydrological conditions.

The Manning’s Roughness Coefficient (n) is a crucial parameter in the HEC-RAS model, used to describe the roughness or frictional resistance of flow in a channel or river⁴⁸. The Manning’s Roughness Coefficient is a dimensionless number. In this study, the specific values for Manning’s n were set based on recommendations in the HEC-RAS Hydraulic Reference Manual 6.5 (HEC-RAS User Manual).

Similar to other ancient canals, and considering the historical context of unmechanized construction during the Ming and Qing dynasties in China, which aimed to minimize anthropogenic intervention in the natural riverbed environment⁴⁹, and combined with recommended reference values from the HEC-RAS Hydraulic Reference Manual (HEC-RAS User Manual)⁵⁰, Manning’s n values were set for the canal banks and channel respectively: Canal banks as mature field crops, n = 0.050; For the artificially excavated, winding, and sluggish earthen canal channel with dense weeds or aquatic plants in deep channels, n = 0.033 (Table 2).

Table 2 Manning’s n values (suggested references from HEC-RAS Hydraulic Reference Manual 6.0 3-1)

Model validation

To ensure the accuracy and applicability of the Nanwang Water Diversion Hub HEC-RAS model constructed in this study, the research team selected historical flood events with detailed hydrological records for validation. According to historical hydrological records, on September 13, 1964, an extreme flood event occurred in the Nanwang region, considered the most severe flood disaster in the region’s history. The flood peak flow was high, and the event lasted for an extended period, demonstrating strong representativeness and extremity. In terms of historical data, this flood has relatively complete official hydrological records and documentary accounts from the Wenshang County Meteorological Bureau in Jining City, Shandong Province. Relevant data on rainfall and river responses are relatively comprehensive, providing a solid data foundation for the construction of high-precision models. The measured peak discharge of this flood reached 6930 m³/s, with average rainfall of 520.9 mm in August, 147 mm in September, and a total annual rainfall of 1394.8 mm, far exceeding the multi-year average of 641.71 mm for the area. (Data source: Wenshang County Meteorological Bureau, Jining City, Shandong Province).

Using this flood event as a benchmark, the research team input the 1964 rainfall conditions and relevant upstream boundary flow parameters into the constructed HEC-RAS model without altering the model structure. Simulation results showed that the model-calculated peak discharge was 6714.32 m³/s, deviating from the measured value by only 3.1%. This error falls within the reasonable range for large flood simulation studies, indicating the model’s high reliability in accurately reconstructing the hydraulics of extreme historical floods.

Model construction

The research team established a fundamental database for the Nanwang Water Diversion Hub to comprehensively assess its hydrodynamic characteristics and the efficacy of its hydraulic facilities. This database includes precise geographic coordinates of facility sites, construction timelines, engineering dimensions, and other information (Table 1). Through on-site creation of 3D point cloud models, the current state of some hydraulic facilities was verified, and supplementary data on their preservation status and precise geographic distribution was added. This content provided accurate data support for subsequent restorative hydrological simulation experiments.

Based on the topography, channel morphology, and flow direction of the Nanwang area, combined with DEM (Digital Elevation Model) data, the topographic information of the Nanwang Water Diversion Hub was restored, revealing key information such as terrain relief, slope changes, and river flow direction. This information was then imported into HEC-RAS to further reconstruct the geometric model of the channel and hydraulic facilities (Fig. 12). During the hydrological model restoration process, the team set reasonable hydrological boundary conditions based on the actual characteristics of the Nanwang Hub, including key parameters such as inflow, flow velocity, and water level, ensuring the model could realistically reflect the hydrological changes under natural and human-regulated conditions.

Fig. 12: Geometric model of the Nanwang area channel and hydraulic structures.

Representation of channel, channel cross-section used in HEC-RAS simulations.

Based on HEC-RAS software, the restorative simulation experiments of the model were realized. This study focused on the impact of the restorative reconstructed hydraulic structures on the channel path and water resource management in the Nanwang area. Simulation results not only demonstrate the overall regulatory capacity of the restored structures but also specifically evaluate their efficiency in integrated water resource management.

Finally, through comparative analysis of the restorative simulation results, the research team was able to precisely assess the role of the restored hydraulic structures in the complex river hydraulic characteristics of the Nanwang area. These findings help optimize the functions of existing hydraulic facilities, enhancing the hydrological regulation and water resource management capabilities of the canal’s hydraulic facilities.

Ethics approval and consent to participate

Manuscripts reporting studies not involve human participants, human data or human tissue.

Results

Flow characteristics analysis

Through simulation experiments using HEC-RAS software, the research team obtained a series of important hydrological parameters for each hydraulic facility site within the restorative reconstructed Nanwang Water Diversion Hub, including energy grade elevation, flow area, water surface elevation, Froude number, etc. Specific calculation results are presented in tabular form (Tables 3–6). These parameters provide the basis for analyzing the flow characteristics of the Nanwang area, illustrating the role of hydraulic facilities in regulating flow and enhancing navigational capacity.

Table 3 Channel energy grade elevation comparison data table

Table 4 Channel flow area comparison data table

Table 5 Channel water surface elevation comparison data table

Table 6 Channel Froude number comparison data table

Before the overall hub restorative reconstruction, the average energy grade elevation of the channel was 28.239 m, with site data points showing significant fluctuations (Fig. 13). The flow area was 585.791 m², and the average water surface elevation was 28.202 m (Fig. 13). This data indicates uneven energy distribution and unstable flow conditions within the channel, with high fluctuations in kinetic and potential energy. Such hydraulic conditions easily lead to localized energy concentration, causing phenomena detrimental to navigation. Furthermore, the maximum Froude number reached 1, indicating the presence of hydraulic jump phenomena and excessively fast local flow velocity, resulting in intense energy conversion unfavorable for stable ship navigation, while also exacerbating local riverbed erosion and sedimentation issues.

Fig. 13: Comparative analysis of channel energy grade elevation in the Nanwang Water Diversion Hub.

Plot comparing energy grade elevations along the channel under different scenarios.

Following the comprehensive restoration of the hydraulic facilities, significant changes occurred in the local channel’s flow characteristics. Analysis of the hydrological simulation results (Fig. 14) shows that the average energy grade elevation of the channel became 31.283 m, a year-on-year increase of 10.7% compared to before reconstruction (Fig. 15). Site data points became more stable (Fig. 13). The average flow area increased to 1863.298 m², a year-on-year increase of 218.082% (Fig. 15). The average water surface elevation rose to 31.279 m, a year-on-year increase of 10.910% (Fig. 15). These results indicate that the restorative reconstruction of the hydraulic structures effectively increased the channel’s water level and the potential energy of the flow her enhancing the overall channel capacity. The higher hydraulic head energy endowed the Nanwang area with stronger water diversion, storage, and drainage capacities, significantly improving its hydraulic conditions.

Fig. 14: Channel water volume simulation diagram.

Left panel: Channel water volume before setting up a facility; right panel: Channel water volume after setting up a facility.

Fig. 15: Channel average energy comparison analysis chart.

Left panel: Comparison of Average Energy Grade Elevation; middle panel: Comparison of Average Flow Area; right panel: Comparison of Average Water Surface Elevation.

Notably, the restorative reconstructed Nanwang Hub, by elevating the channel water level, increased the navigable water depth and formed a stable hydraulic head difference. This played a crucial role in ensuring the smooth passage of large-tonnage vessels, effectively mitigating the risk of ship grounding or impassability during low-water periods or under conditions of uneven energy distribution. Simultaneously, the redistribution of water flow energy by the hydraulic facilities effectively alleviated localized energy over concentration, thereby reducing the frequency of hydraulic jump occurrences and enhancing channel stability.

Analysis of changes in river hydraulic characteristics

To more intuitively represent the changes in river hydraulic characteristics during the restorative reconstruction simulation experiments, the research team integrated data from the Nanwang Hub before and after reconstruction, employing graphical methods for visualization.

The Nanwang area differed significantly before and after the restorative reconstruction of hydraulic facilities, directly impacting vessel navigation. Moreover, the geographical elevation of the channel gradually decreased north and south from the central Nanwang area (Fig. 16). Combined with Fig. 17 analysis, it can be seen that before restorative reconstructing the hydraulic facilities, the continuity of flow was stronger in the northern channel of the Nanwang area. However, flow continuity in the southern channel progressively weakened, and water volume gradually decreased. Concurrently, localized flow power and shear stress showed significant increases, reaching maximum values of 259.20 W/m² and 82.91 N/m², respectively (Fig. 18). This data indicates excessively fast local flow velocity (Fig. 18) and strong sediment transport capacity, leading to riverbed siltation and blockage, ultimately causing the channel to dry up. This phenomenon reflects that before restorative reconstructing the hydraulic facilities, the hydrological environment in the southern channel of the Nanwang area was insufficient to support canal continuity and stable vessel passage.

Fig. 16

Cross-section of the Nanwang Water Diversion Hub (Adapted by the Research Team, Source: Assessment of the Composition and Value of China’s Grand Canal Heritage, P72). Shows the cross-section of the channel.

Fig. 17: Channel water volume simulation diagram.

Left panel: Channel water volume before setting up a facility; right panel: Channel water volume after setting up a facility.

Fig. 18: Comparative analysis of channel shear stress.

Left panel: Channel shear stress before setting up a facility; right panel: Channel shear stress after setting up a facility.

Specifically, after restorative reconstruction, each hydraulic facility within the Nanwang Hub was interconnected, forming a complete and tightly integrated hydraulic engineering network. This hub conveyed water flow to the southern part of the Nanwang area, ensuring flow continuity and smooth passage for vessels. Simultaneously, the hydraulic facilities widened the top width of the northern channel, reducing flow power by 72.02% and shear stress by 20.10% (Fig. 18). The reduction in flow power and shear stress inhibited riverbed erosion through a dual mechanism: first, expanding the area below the critical shear stress threshold, thereby reducing erodible area; second, attenuating sediment transport capacity, promoting the deposition of fine particles to form a protective surface layer. This indicates a significant improvement in flow stability and increased water volume, ultimately achieving a positive cycle for riverbed stability.

Analysis of hydraulic structure efficacy characteristics

Through HEC-RAS computational analysis, the research team obtained cross-sectional data for the Nanwang Hub before and after restorative reconstruction. Integrating this with topographic data and presenting it visually enabled a precise analysis of the efficacy and operational mechanism of its hydraulic structures. The restorative reconstructed Nanwang Hub contains 26 flow control facilities, accounting for 92.6% of the entire hub, indicating their pivotal role. The hub’s hydraulic structures can be categorized into three types: flow control facilities, overflow facilities, and water retention facilities (Table 1). Among these, overflow facilities primarily discharge excess water from the channel, prevent flooding, ensure stable water levels, and protect downstream safety. Water retention facilities primarily ensure the normal operation of navigation and coastal agricultural irrigation. Since the Nanwang Hub has only one instance each of an overflow facility and a water retention facility, their numbers are too small to fully represent their functional characteristics. Therefore, this study primarily discusses the operational mechanism of flow control facilities and their impact within the channel.

Before restorative reconstructing the hydraulic structures, the flow in the overall channel of the Nanwang area exhibited strong continuity but insufficient water volume in the northern channel. The average top width was 254.962 m, average water depth was 3.428 m, average channel velocity was 0.7 m/s (Fig. 19), and average shear stress was 13.209 N/m². This data indicates that the northern channel was relatively narrow, restricting flow diffusion and intensifying friction between water and the riverbed, leading to noticeable riverbed erosion. This erosion further enhanced sediment transport capacity, causing siltation in parts of the channel, thereby hindering the normal passage of large or high-tonnage vessels.

Fig. 19: Channel cross-section flow velocity.

Left panel: Channel cross-section flow velocity before setting up a facility; right panel: Channel cross-section flow velocity after setting up a facility.

After restorative reconstructing the hydraulic facilities, navigation was ensured. A series of flow control facilities, including Shizi Sluice, Siqian Sluice, Liyun Sluice, Liulin Sluice, Xujiankou Doumen, Shili Sluice, etc., were constructed in the Nanwang area. Coordinated operation of these facilities significantly optimized the local channel’s river hydraulic characteristics. Specifically, the flow control facilities widened the top width to 537.819 m, increased water depth to 6.505 m, while controlling channel velocity to 0.2 m/s (Fig. 19), reducing shear stress by 64%. These changes mean that the flow control facilities greatly increased the lateral flow space of the water body, reduced concentrated flow pressure within the channel, optimized top width, depth, and velocity, and significantly enhanced the channel’s navigational capacity, creating more favorable conditions for high-tonnage vessel passage. Moreover, sediment transport capacity was substantially reduced. This not only diminished the erosive effect of flow on the riverbed but also effectively prevented siltation, enhancing channel stability and traffic ability.

Uncertainty and reliability analysis

This study strives to ensure the rigor of data sources and the accuracy of spatial precision in model construction. However, due to limitations in historical records and measurement technology, a certain degree of uncertainty remains unavoidable. In terms of boundary condition setting, due to the lack of systematic hydrological observation data from the Ming and Qing dynasties, this study systematically compiled historical documents such as the “Compilation of Historical Flood Investigation Materials in China” and the “Ming History: River and Canal Annals,” combined with actual measurement data from modern hydrological conditions in similar river basins, to reconstruct the typical boundary water levels and flow processes in the upstream and downstream areas of the study region. This process may involve certain temporal precision errors and water level amplitude errors. In terms of boundary condition settings, considering the historical context of the construction of the Grand Canal during the Ming and Qing dynasties, which relied on non-mechanized construction methods, the aim was to minimize human intervention in the natural riverbed environment at the time. In the absence of complete cross-sectional and roughness coefficient measurement data, this study determined the Manning’s roughness coefficient for the river channel based on terrain type and vegetation coverage, referencing the recommended values in the HEC-RAS Hydraulic Reference Manual (HEC-RAS User Manual), and adopted a unified zoning assignment method. This method is suitable for overall river regulation needs, but may deviate from historical actual conditions in local small-scale cross-sections.

To minimize the impact of the aforementioned uncertainties on simulation results, the research team assessed the response of core output indicators to changes in roughness, water level boundaries, and cross-sections. The assessment results showed that the aforementioned uncertainties caused some fluctuations in the water surface line and flow velocity values at local nodes, but the overall hydrological evolution trends exhibited consistent characteristics. Therefore, these uncertainties are within an acceptable range and do not affect the reliability of the scientific conclusions regarding the operational mechanisms of the Nanwang Water Diversion Hub.

Discussion

This study conducted a restorative reconstruction and numerical simulation of the Ming-Qing Nanwang Water Diversion Hub based on an HEC-RAS model, quantitatively revealing its operational mechanisms in regulating water levels, stabilizing flow, enhancing navigation safety, and inhibiting riverbed erosion for the first time. The results provide solid scientific support for the re-evaluation of the Beijing-Hangzhou Grand Canal as a world-class linear cultural heritage and inject a new technological perspective into understanding the historical evolution of ancient hydraulic engineering.

By integrating 3D point cloud data with the HEC-RAS geometric model, this study achieved a digital reconstruction of the Nanwang Hub and visually recreated its “low-tech yet highly adaptive” energy dissipation mechanism in simulations. Compared to the French Midi Canal, the Midi Canal primarily employs a discrete lock system for energy dissipation (Fr ≈ 1.2), with lock group spacings of approximately 3–5 kilometers. Its operational mechanism centers on “linear lifting,” primarily designed to meet the demand for smooth navigation. In contrast, the Nanwang Water Diversion Hub employs a continuous stepped energy dissipation system (Fr ≈ 0.3), combined with a matching lock group spacing of 2.1 ± 0.3 kilometers. This effectively reduces erosion and wear on the riverbed and lock structures, significantly extending the lifespan of the facilities, while also accommodating multiple functions such as water diversion, navigation, and irrigation. This achievement can be directly applied to heritage parks and exhibition halls. Through dynamic digital imagery and interactive platforms, the public can intuitively experience the engineering wisdom and historical landscape of ancient hydraulic hubs, further enhancing societal recognition of the Grand Canal’s hydraulic culture.

Based on simulation results, this study clarified the key role of the Nanwang Hub in elevating channel water levels, stabilizing flow regimes, and reducing hydraulic jump phenomena. Simultaneously, its comprehensive enhancement of the channel’s water diversion, storage, and drainage capacities ensured continuous navigation during dry seasons and long-term sustainable operation of the canal. Flow control facilities, by rationally distributing flow energy, not only alleviated localized energy concentration but also reduced shear stress and flow power by ~20% and 72%, respectively, further enhancing overall riverbed stability.

The above simulation analysis indicates that the Nanwang Water Diversion Hub already possessed ecological regulation functions similar to the modern “sponge city” concept during the Ming and Qing dynasties. It replenished groundwater through channel seepage (infiltration), regulated flood peaks via diversion dams and sluice gates (detention), ensured canal transportation and irrigation through elevated water storage (storage and utilization), and discharged excess water in multiple directions during flood seasons (discharge). These resilient functions align closely with modern water management principles. From a cultural heritage reuse perspective, this study proposes a “three-zone integrated” conservation plan: 1. Core Conservation Zone: Preserve original sluice and dam ruins to establish a “Water Conservancy Culture Experience Area.” Utilize digital projections and virtual tours to enhance visitors’ immersive understanding of ancient hydraulic engineering wisdom; 2. Public Archeology Experience Zone: Install visual hydrodynamic demonstration devices and interactive activities simulating critical events like flood scenarios and lock operations to deepen public understanding of canal functions; 3. Historical Environment Restoration Zone: Leverage existing canal ruins to create an ecological education base and low-impact sightseeing routes, integrating heritage preservation with eco-tourism and rural revitalization.

In future restorative reconstructed research, this study’s methodology can be extended to other significant canal heritages (e.g., Rome’s Canal de Rome, France’s Canal du Midi). Through restorative simulation and cross-regional comparison, the synergistic mechanisms and modern value of ancient hydraulic hubs can be revealed, providing a model for the scientific conservation, maintenance, and revitalization of global linear hydraulic cultural heritage. Concurrently, combining heritage management with “digital twin” technology can enable real-time monitoring of site health and risk early warning, offering historical insights and design inspiration for enhancing urban water environment resilience and flood control.

However, when applying this research method to other canal heritage sites, several challenges may arise. There are significant differences in the types of hydraulic structures, operational mechanisms, and topographical conditions across different regions, necessitating targeted adjustments and recalibration of hydrological parameters (such as Manning’s Roughness Coefficient, boundary conditions, and hydraulic structure settings) in the HEC-RAS model. Additionally, the preservation of reliable historical documentation varies across different heritage sites, with some regions potentially lacking clear spatio-temporal information or missing documentary records, which may limit the precise restoration of hydraulic structures. Furthermore, each heritage site has distinct environmental differences and varying degrees of natural erosion. Some sites lack sufficient ground-based physical information, which is unfavorable for conducting high-precision real-world modeling. Therefore, during the transfer of this research method, it is essential to thoroughly assess the local historical documentation foundation, hydrological and geographical characteristics, and the current state of heritage preservation at the site, and flexibly adjust the modeling precision and technical approach to ensure the scientific rigor and accuracy of the method system.

In summary, this study not only provides innovative pathways for the digital conservation, public education, and ecotourism development of the Nanwang Water Diversion Hub but also lays a theoretical and technical foundation for the revitalization and sustainable development of the Beijing-Hangzhou Grand Canal’s hydraulic cultural heritage. It holds significant reference value for the management and conservation of similar global tangible cultural heritage. This study contributes not only to preserving the physical form of the canal heritage but also to transmitting its historical and cultural significance, ensuring its sustainable development in a modern context.