Simulation study on the impact of check dams on water and sand in Xiliugou Basin and inner Mongolia section of the Yellow River

Abstract

Severe soil erosion on the Loess Plateau has led to a reduction in the area of agricultural land as well as an increase in the risk of flooding in the lower reaches of the Yellow River. Ten Kongdui (Mongolian for “Kongdui”, meaning “Great Mountain Gully”) is located in the upper reaches of the arsenic sandstone hilly and gully area. It is located in the heart of the Kubuqi sandstorm area. This area is one of the sandy and coarse sand production areas in the middle reaches of the Yellow River. It is also the main sand source area of the Inner Mongolia section of the Yellow River. The Ten Major Kongdui Xiliugou Basin is located in the upper and middle reaches of the Yellow River in the coarse sand-producing area. The gullies are deep and steep, with exposed arsenic sandstone. The chain reaction of heavy rain, flood and sediment is intense, making it a key channel for coarse sand from the Yellow River to flow into the river. To effectively address soil erosion in this area, curb the expansion of pyrite sandstone gully erosion and reduce the amount of sediment flowing into the Yellow River, it is proposed to establish an integrated engineering system of “soil and water conservation - sediment interception” within the basin. Through the measure of check dam local sediment storage will be achieved, the ecosystem functions will be restored, and the healthy life of the Yellow River will be maintained. Using distributed hydrologic modeling to explore the effects of a sand detention project in the Xiliugou watershed on watershed runoff and sand transport, the SWAT model was calibrated (1990–1999) and validated (2000–2020) using observed runoff and sediment data at Longtouguai Station, the simulated runoff and sand transport at Longtouguai Hydrological Station were found to fit well with the measured values through model simulation. The linear fitting coefficient R² exceeds 0.6, it is considered that the linear relationship between the simulated values and the measured values is reasonable, which indicates that the reservoir model in SWAT model can be used for check dam simulation, and the water and sand impacts of water and sand reduction of the new check dam project on the Xiliugou watershed are analyzed through the results of the SWAT model calculations and the impacts of further calculations on the channel siltation of the Inner Mongolia section of the Yellow River are further calculated. The results show that: 1, the construction of check dams can affect the runoff volume of the basin to a certain extent, and intercepts part of the runoff, the average annual water reduction of the newly built 79 check dams is 2.44 × 10⁶ m³. 2, it has a great influence on the sand transport in the basin, and the effect of sand reduction is obvious, the average annual sand reduction of the newly built 79 check dams is 4.09 × 10⁵ t. 3, Reduces sand content in the Yellow River and enhances flushing of existing sediment in the Nei Mongol section of the river, and reduces water demand for sediment transport. The results of this study provide reference for promoting the construction of water sand replacement project in Xiliugou Basin and the high-quality development of the Yellow River Basin.

Introduction

The Yellow River, often called the ‘mother river’ of China, is vital to the nation’s ecological security. In recent years, ecological protection and high-quality development of the Yellow River Basin have been elevated to a national strategy. Reducing sediment influx is the key to safeguarding the river¹. The Ten Major Kongdui (kongdui = flash-flood gully in Mongolian) deliver some of the highest coarse-sediment yields on the Loess Plateau and constitute the dominant source of channel-bed aggradation in the lower Yellow River, posing a serious threat to flood safety. Field observations and modelling have repeatedly shown that engineering measures can markedly modify the basin’s flow–sediment relationship. Shi et al. ² applied SWAT to the Wuding River and found that check-dam construction reduced annual runoff and sediment load by 12% and 11.7%, respectively, during 1970–1980. Su et al. ³ attributed 63.6–83.2% of the runoff decline in the Huangfuchuan River between 1960 and 1979 and 1980–1999 to anthropogenic activities, with the proportional contribution increasing over time. Likewise, Wang et al. ⁴ demonstrated that human interventions accounted for > 50% of the runoff and sediment changes in two small tributaries. These studies collectively underline the decisive role of engineering works in improving the flow–sediment regime of the Yellow River^5,6. Although the sediment-reduction benefit of check dams has been widely recognised, quantitative appraisals specifically targeting the Xiliugou catchment—one of the coarsest-sediment hotspots among the Ten Kongduis—are still lacking. The present study therefore aims: (1) to quantify the runoff and sediment retention of 79 newly planned check dams in Xiliugou using the SWAT model, and (2) to evaluate how this local intervention propagates downstream to alter sediment deposition along the Inner-Mongolia reach of the Yellow River.

The Inner Mongolia section of the Yellow River is a typical alluvial section. On its southern bank, there are the Kubuqi Desert and the Ten major tributaries (in Mongolian, “tributaries”), including the Maobula River, the Burgasetai Valley, the Heilai Valley, the Xiliugou, the Hantai River, the Haoqing River, the Hasila River, the Muhar River, the Dongliugou River and the Hustai River. Xu et al. ⁷ showed that the amount of sediment transported to the Yellow River each year is more than one-tenth of the total amount of sediment input to the Yellow River under the influence of the northwest wind that dominates the region. Most of the large and coarse particles that enter the main stream in the form of high-deposition logistics are deposited in the riverbed. It highly promoted the sediment content of the upstream flood, which led to the formation of an actual “above-ground suspended river”, a phenomenon that had a significant impact on the geographical environment of the area⁸. As the sources of the Ten Major Kongdui are all in the hilly and gully areas of arsenic sandstone with extremely severe soil erosion, flowing through the Kubuqi Desert, it is very easy to form high-sediment floods. However, floods with high sediment content will in turn cause siltation and riverbed rise downstream, thus causing serious impacts on sediment transport and riverbed evolution in the mainstem of the Yellow River (Inner Mongolia section)⁹.

A check dam is a hydraulic structure built in various levels of gullies in areas with soil erosion for the purpose of retaining silt and silting the ground^10–12. It has been widely adopted in the sandy and coarse-sand areas with severe soil erosion in the upper and middle reaches of the Yellow River, especially in the concentrated source areas of coarse sediment^13,14.General Secretary Xi Jinping pointed out in his speech at the Symposium on Ecological Protection and High-Quality Development of the Yellow River Basin that “the middle reaches of the Yellow River should emphasize soil and water conservation and pollution control, and where there are conditions, we should vigorously build dry terraces and check dams.” Scholars at home and abroad have conducted extensive research and analysis on the hydrological effects of check dams in construction areas, Guyassa et al. ¹⁵ analyzed the impact of the combination of check dams and vegetation measures on the local flood process through actual observations conducted in Ethiopia. They found that the combination of check dams and vegetation measures significantly reduced the peak flow and total flood volume in the basin, and delayed the arrival time of the flood peak; Pal et al. ¹⁶ analyzed that various combinations of the number, size and site location of check dams established in the same area could have an impact on the sediment retention capacity and sediment retention years, and found that the construction of check dams would have a significant impact on the sediment transport rate of the basin; Li et al. ¹⁷ studied the water reduction effect of 80 check dams in the Luan River Basin from 1996 to 2000 and found that the average annual water reduction of the check dams in this basin was 0.34 billion cubic meters; Castillo et al. ¹⁸ showed that check dams intercept sediment in the upstream watershed and reduce sediment transport downstream, which is important for flood control security and sediment transport in the downstream watershed. Based on the research of various scholars and long-term practical experience in soil and water conservation, it has been proved that, check dams, as a unique gully regulation project in the Loess Plateau, play crucial roles in flood regulation, erosion reduction, and sediment trapping. These functions not only effectively reduce the amount of sediment flowing into the Yellow River and mitigate sediment deposition in the lower reaches of the Yellow River but also create dammed land suitable for crop cultivation. As a result, silt dams generate substantial soil and water conservation benefits, along with significant social and economic returns in the construction regions.

At present, how to scientifically quantify the impact of check dams on watershed water and sand is the focus of the current research on check dams construction and watershed ecological environment improvement, the application of a wider range of soil and water conservation measures to evaluate the method of hydrological method and water conservation method. Base period comparison method calculates the degree of influence of precipitation during the study period and human activities on water and sediment changes in the basin by establishing the functional relationship between precipitation in the base period and runoff and sediment transport volume. The water conservation method superimposes the contribution of each water and sand reduction factor and calculates its effect on water and sand changes. Although both have made significant contributions and produced a large number of research results in calculating the impact of water and sediment reduction by soil and water conservation measures, due to the large number of influencing factors considered in their calculations and the high accuracy of the required data, the application of different parameters and data with different accuracies in actual calculations will have a significant error impact on the calculation results. With the development of the theory of sediment production and hydrological simulation technology, scholars at home and abroad have begun to calculate the results of sediment production in river basins using different hydrological models. Freeze et al. ¹⁹ first proposed the concept and framework of distributed hydrophysical models, and since then, distributed hydrological models have begun to develop; Pareta et al. ²⁰ calculated the sediment migration of the Brahmaputra River and the predicted flood water level at the Gumi site using the MIKE 21 C model; Tibebe et al. ²¹ evaluated the surface runoff generation and soil erosion rate of a small watershed (Keleta watershed) in the Awash River Basin of Ethiopia based on the SWAT model. The predicted soil loss was significantly correlated with the measured rainfall and the simulated surface runoff.

The above-mentioned research only provides a comprehensive coefficient for the rigid barrier of “dams”, and fails to answer the marginal issue of “how much sediment is reduced by adding one dam”. At the same time, it does not offer the impact of the construction of upstream check dams on the water and sediment of the Yellow River downstream. The distributed hydrological model (SWAT) with a physical basis can fully consider the impacts of climate change, human activities, etc. on the hydrological processes of river basins²². Therefore, in this paper, the SWAT model is adopted to calculate and simulate the two hydrological elements of runoff and sediment at the Longtouguai hydrological station in Xiliugou. By constructing the SWAT model, the time period from 2000 to 2020 is selected, and three scenarios are set up: no check dams, existing check dams, and existing sediment dam + planned new sand check dams. The water-sediment effect in the basin was simulated and calculated to obtain the sediment production and discharge data of the Xiliugou Basin. The research area is located in the concentrated distribution zone of the top ten Kongdai in the Loess Plateau, and is one of the typical representatives of the construction of check dams in the middle reaches of the Yellow River. It has completed a complete density gradient of “zero dam - one hundred - nearly two hundred” within 20 years, this study adopts a spatiotemporal control experimental approach of ‘same climate, three dam capacities’, using the SWAT + check dam reservoir module to quantitatively isolate the independent contribution of the dam body to runoff-sediment, providing an extensible numerical paradigm for ‘performance audit of existing projects + admission of new projects’ in the upper and middle reaches of the Yellow River.” Based on the measured data, the sediment production and discharge amounts of the ten major Kongdui and other rivers were estimated using empirical formulas. The annual sediment discharge volume of the downstream Toutaoguai Station was fitted through regression analysis. Further, the sediment discharge rate method was used to calculate the scouring and siltation volume of the Inner Mongolia River section after the construction of the siltation dam in the Xiliugou Basin.

Study area and data

Overview of the research area

The Xiliugou Basin is one of the Ten Major Kongdui (Mongolian for “mountain flood valley”) that directly flow into the Yellow River in the Inner Mongolia section of the Yellow River. It originates from the top of Zhangjia Mountain in Duihao Township, Boerjianghaizi Town, Dongsheng District, Ordos City, Inner Mongolia. Flows through Zhierjianghaizi Town and Zhan Danzhao, Zhaojunfen Township, and emptys into the Yellow River from Hebian Village, Zhaojunfen Township. The total length is 106.5 km, the total area of the watershed is 1,356.3 km², and the terrain within the watershed is higher in the south and lower in the north, the relative height difference between the north and south is approximately 500 m. The Longtougai Station is located in the middle reaches of the Xiliugou main stream, close to the lower reaches. The catchment area of the basin above this station is 1,157 km², occupy most of the total catchment area of the basin. The upper reaches of the Xiliugou River Basin are hilly areas. The surface is fragmented and full of gullies. The surface composition is mainly loess, and there is a large area of arsenic sandstone exposed in the underlying strata. The arsenic-bearing sandstone exhibits a loosely consolidated structure and is susceptible to disintegration upon exposure to water. Therefore, erosion in this area is relatively intense²³. The middle reaches of the river pass through the Kubuqi Desert. During the monsoon season, a large amount of wind-sand soil accumulates in the river course and on both sides of the riverbank, forming a wind-sand area landform. When the flood flows through this area, a large amount of sediment in the river course is carried away, causing the sediment content to further increase²⁴. The downstream area is an alluvial flood sector, with a flat terrain and wide, shallow river channels. The flat terrain and wide, shallow channels promote sediment deposition, affecting channel stability and reducing flood conveyance capacity. The location of the basin is shown in Fig. 1.

Fig. 1.

Location of the Xiliugou Basin in the Ten Major Kongdui region, Inner Mongolia.

Study regional data

The construction of large-scale check dams in Xiliugou began in 2000. By the end of 2015, a total of 113 check dams had been built in the Xiliugou Basin, including 41 main check dams and 72 medium and small-sized check dams. According to on-site investigations, 16 check dams broke during the “August 17” flood in 2016 (including 11 main check dams and 5 medium and small-sized check dams). There are currently 97 silt-retaining check dams in the Xiliugou Basin, with a dam control area of 203.68 km² and a sediment retention capacity of 1.79 × 10⁷ m³.

The on-site investigation of the construction of the check dam is as follows, 79 new sand-retaining check dams have been built in the Xiliugou Basin, with a controlled area of 138.06 square kilometers and a sediment retention capacity of 2.95 × 10⁷ m³. Among them, there are 19 medium-sized sand-retaining check dams, with a controlled area of 71.04 km² and a sediment retention capacity of 1.81 × 10⁷ m³. There are 60 small sand-retaining check dams, with a controlled area of 67.02 km² and a sand-retaining capacity of 1.14 × 10⁷ m³.

The data required for constructing the SWAT model in this paper include the DEM (Digital Elevation Model) map of the study area, soil type data, land use data, meteorological data (temperature, wind speed, sunshine duration, etc.), measured hydrological data and the conditions of the check dam in the study area (the current data of check dam is up to 2015). The relevant data and sources are shown in Table 1 below.

Table 1.

Basic data and sources of the SWAT model.

Data type	Resolution / Scale	Year	Data source
DEM	30 m	2009	Geospatial data cloud
Soil	1:1,000,000	2000	National Earth System Science Data Platform
Land utilization	1:100,000	2010	China Resources and Environment Database
Meteorology	Day	1990 ~ 2020	China Meteorological Science Data Sharing Service Network
Hydrology	Day	1990 ~ 2020	Yellow River Hydrological Yearbook
Current check dams		1993 ~ 2015	Water Conservancy Bureau of Dalaate Banner
Plan to build a new check dam		2018	The “Implementation Plan for the Pilot Project of Sand Control and Water Exchange in Ordos” (2018) approved by the Yellow River Conservancy Commission
Annual water and sediment data of the Yellow River section in Inner Mongolia	Year	2008 ~ 2020	Water resources bulletin, sediment bulletin and soil and water conservation bulletin.

Research method

SWAT model method

Based on the measured hydrological data, DEM, soil types, land use and meteorological data (temperature, wind speed, sunshine, etc.) of the study area, a SWAT model is constructed. Complete data input, parameter calibration and model checking. The existing check dams and the proposed new sand-retaining check dams are taken as the condition input model. Set up multiple scenarios, simulate and output the runoff volume and sediment transport volume under different schemes.

By adopting a three-step approach of “zero dams → existing dams → additional planned dams”, the aim is not to reproduce the measured values for each year, but to quantitatively isolate the independent contribution of the “dams” as a single variable to the hydrological process line. Scenario 1 (Zero dam): It provides a “natural benchmark” for calculating the “changes caused by the dam” rather than the “absolute error”. Scenario 2 (97/113 dams): The actual dam height, reservoir capacity and siltation thickness are input according to the acceptance data, representing the “actual management conditions”. Scenario 3 (176 dams): on the basis of Scenario 2, 79 additional “planned dams” will be added. The reservoir capacity will be determined based on the feasibility study report to predict the “maximum possible future effect”.

The SWAT model contains a large number of physical and mathematical formulas and involves dozens of parameters. During the use of the SWAT model, a large number of researchers have found that among these parameters, the changes of some parameters will have a significant impact on the simulation results of the model. These parameters are called sensitive parameters. The main sensitive parameters of the SWAT model when conducting runoff simulation include CN2 (the curve number that determines the partitioning of rainfall into surface runoff and infiltration; higher values accelerate runoff generation and constitute one of the most sensitive controls on streamflow simulation), ESCO (the soil evaporation compensation factor that modulates the difference between actual and potential evapotranspiration; larger values intensify the evaporation limitation and markedly affect runoff, particularly in arid regions), ALPHA_BF (the base-flow recession coefficient that governs the rate at which groundwater discharges to the channel; higher values yield faster base-flow response and directly control low-flow regimes), SOL_AWC (the available water capacity of the soil layer that defines the soil’s ability to store moisture; greater capacity buffers storm runoff and is sensitive across all climatic zones), GW_REVAP (the groundwater “revap” coefficient that regulates the upward flux of groundwater into the soil profile for subsequent evapotranspiration; increased values reduce base-flow contributions while enhancing evapotranspiration losses), SOL_K (the saturated hydraulic conductivity of the soil that controls the vertical percolation rate; higher conductivity decreases surface runoff and increases both lateral subsurface flow and groundwater recharge).

The research adopts the PSO algorithm of Swat-CUP to automatically invoke SWAT output to complete parameter calibration, sensitivity and uncertainty analysis. Update the parameters with New_pars.txt and iterate in a loop until R² and E_NS in Summary_Stat.txt meet the preset precision. When verifying 2016–2020, “dam failure” must be re-input into the model as a new boundary condition. Taking 2016-08-17 as the node, 16 dams were removed from the pond/res module of SWAT. From 2000 to 2015, the retention rates were fixed using the original dam system parameters (CN2, ESCO, CH_N2, SPEXP, etc.). From 2016 to 2020, only the geometry and quantity of the dam system were modified, while the rest of the parameters were locked and the model was re-run.

This paper divides the Xiliugou Basin into 35 sub-basin units based on DEM (Fig. 2.)

Fig. 2.

Sub-basin division map of Xiliugou Basin.

It can be known from the model evaluation results in Table 2 that during the rate period, the coefficient of determination R² between the simulated values and the measured values is 0.75, and the Nash efficiency coefficient E_NS is 0.7. The calibrated parameters were imported into SWAT for the runoff simulation calculation during the verification period again, and the evaluation results of the verification period were obtained: R² is 0.7 and E_NS is 0.63.

Table 2.

Simulation evaluation indicators of annual runoff volume and sediment transport volume of the model.

Time period	Simulated variable	R 2	ENS	RMSE%	RE
Calibration period(2000 ~ 2015)	Annual runoff	0.75	0.70	23.56%	11.62%
Validation period(2016 ~ 2020)	Annual runoff	0.70	0.63	18.41%	8.06%
Calibration period(2000 ~ 2015)	Annual sediment transport volume	0.72	0.68	26.73%	-14.27%
Validation period(2016 ~ 2020)	Annual sediment transport volume	0.67	0.62	19.62%	-13.13%

As can be seen from Table 2, the simulation results of the rate period and the validation period, R² ≥ 0.65, E_NS ≥ 0.60, RMSE%≤30%, the RE is all within ± 20%, the rate results are more stable²⁵, the rate parameters meet the requirements of the model, and can be used for numerical simulation of the water and sand transport process in the later stage.

Based on the data of the existing and proposed check dams, three scenarios were hypothesized to simulate the two hydrological elements of runoff and sediment beneath different check dams.

Scenario One: Assuming there are no check dams in the study area, simulate the two hydrological elements of runoff and sediment transport at the Longtouguai Hydrological Station in the study area during the period from 2000 to 2020.

Scenario Two: Assuming that the study area has 113 check dams that existed before 2016 and 97 check dams that existed from 2016 to 2020, simulate the two hydrological elements of runoff and sediment transport at the Longtouguai Hydrological Station during the period from 2000 to 2020 in the study area.

Scenario Three: Assuming that the study area has 113 check dams that existed before 2016 and 97 check dams that existed from 2016 to 2020, along with 79 planned new sand-retaining check dams (after completion), simulate the two hydrological elements of runoff and sediment transport at the Longtouguai Hydrological Station during the period from 2000 to 2020 in the study area.

By analyzing the rainfall and runoff data of the Longtouguai Hydrological Station from 2000 to 2020 in the existing data, three scenarios were assumed to simulate the two hydrological elements of runoff and sediment under different numbers of check dams, and the results were summarized.

Calculate the sediment transport volume of the top ten holes by empirical formula

Quantitative analysis of the contribution rate of sediment inflow into the Yellow River from the ten major Kongdui in Inner Mongolia to the siltation of the Yellow River is the prerequisite and foundation for studying siltation reduction technologies and implementing sediment reduction projects. However, among the ten major Kongdui, only one hydrological station is set up downstream of each of the three Kongdui of Maobulang, Xiliugou and Hantaichuan, and the monitoring data are not continuous. Therefore, there are certain difficulties in calculating the measured annual sediment transport of the ten major Kongdui. This calculation selects the research method of Lin Xiuzhi et al. ²⁶ This method combines correlation analysis with frequency curve analysis. Meanwhile, the sand production of other porosity is calculated through the area comparison method and the proportion method of sand transport modulus.

(1) Calculation of the sediment transport volume in the Bursetai Valley and Heilai Valley between Maobula and Xiliugou.

As the two Kongdui are relatively close to each other and located in the middle, the average sediment transport modulus of each year in Maobula and Xiliugou, as well as their respective drainage areas, are used to calculate the annual sediment transport volume of Bursetaigou and Heilai Gou.

(2) Calculation of sediment discharge from the Hantai River.

Construct the sediment regression relationship between the Xiliugou Basin and the Hantaichuan Basin.

When ;

When ;

In the formula, represents the annual sediment discharge of Xiliugou, with the unit of 10⁴ t; represents the annual sediment discharge of the Hantai River, with the unit of 10⁴ t.

(3) Calculation of the annual sediment transport volume of each confluence east of the Hantai River.

Due to the fact that the vegetation and geological conditions of each Kongdui basin east of the Hantai River are relatively similar, in general years, the annual sediment transport volume of the remaining Kongdui basins is directly calculated by using the sediment transport modulus of the Hantai River and the area of each Kongdui basin. When calculating the sediment yield of other Kongdui, the sediment transport modulus of other Kongdui was appropriately reduced based on the research of Zhi Junfeng and Shi Mingli and the proportion of the sediment transport modulus of Hantai River. The information of the ten major Kongdui is shown in Table 3.

Table 3.

The drainage area and river length of the ten major Kongdui.

Kong dui name	Mao bula	Buerse taigou	Heilai Gou	Xiliu gou	Hantai River	Haoqing River	Hashila River	Mu Hua Gou	Dongliugou		Husitai River
Drainage area /km2	1261.5	545.9	943.8	1193.8	874.7	213.3	1088.6	406.7		451.2	406.0
River Chief /km	110.9	73.8	89.2	106.5	90.4	34.2	92.4	77.2		75.4	65.0

Considering the impact of the heavy rain and flood on August 17, 2016, the data of Xiliugou was appropriately reduced and calculated by averaging historical data.

Empirical calculation method for erosion and siltation in the inner Mongolia river section

Adopting the empirical formula method based on the study area²⁷ and the measured water and sand data of the Inner Mongolia section of the Yellow River, the Inner Mongolia section of the Yellow River is selected for calculation, which is located in the lower part of the upper reaches of the Yellow River, and the main stream enters the river from Shizishan in Ningxia and exits the border at Magazine Township of Jungar Banner in Ordos, with a total length of about 823 km. the south bank of this section is distributed by the Qubuzi Desert and the Ten Kongtui. Comprehensively considering the data of the hydrological stations on the main stream, the section from Sanhu Estuary to Toudaoguai was selected as the calculation section for the influence of sediment control in the study area on the erosion and siltation of the main stream of the Yellow River in Inner Mongolia. The empirical formula for calculating the erosion and siltation volume of the Inner Mongolia section of the Yellow River was applied to calculate the impact of the construction of the sediment control project in the Xiliugou Basin on the erosion and siltation of the Inner Mongolia section of the Yellow River.

It should be noted that since over 99% of the annual sediment flowing into the Yellow River in the study area comes from the flood season, the average annual sediment reduction of each river section calculated through empirical formulas can be approximately regarded as equal to the average sediment reduction during the flood season.

In response to the sediment transport and riverbed scouring and siltation issues in the Inner Mongolia river section, Wu Bao sheng et al. established a sediment transport volume calculation method based on the “more inflow, more discharge” formula for sandy rivers. This method takes into account the influence of the incoming sediment volume from the upper station, the cumulative sedimentation volume in the early stage, the critical sediment transport volume, and the sediment particle size of the main and tributary streams on the sediment transport volume. Based on the above method, using the measured water and sediment data of the Sanhu Estuary - Toudaoguai River section from 1953 to 2010, the following calculation formula for the annual sediment transport volume of the Toudaoguai Station was obtained through regression analysis and fitting:

1

In the formula, represents the calculated annual sediment transport of the first turn (10⁸ t); represents the measured annual sediment transport volume (10⁸ t) in the Sanhu Estuary plus Santou interval (excluding Kongdui). represents the measured annual sediment transport volume (10⁸ t) of the ten major Kongdui; represents the measured annual water volume of the Toudaoguai Station (10⁸ m³); represents the calculated annual scouring and siltation volume (10⁸ t) of the Sanhu Estuary to Toudaoguai River section. The subscript “t” represents the first turn, and the symbol “n” indicates the year.

Based on the sediment discharge during the flood season and non-flood season at the outlet section of the river section calculated by formula (1), the sediment discharge during the flood season and non-flood season of the given river section can be further calculated by the sediment discharge rate method. The specific calculation formulas are as follows:

2

In the formula, represents the average annual or flood season scouring and siltation volume of the river section (in billions of tons); W_sj refers to the average annual or flood season sediment import volume of the river section (in billions of tons); The W_sz branch refers to the average annual or flood season sediment inflow (in billions of tons) of tributaries. W_sp discharge refers to the average annual or flood season sand discharge volume (in billions of tons) of the interval drainage ditch. W_sf wind refers to the average annual or flood season when yellow wind enters the river section and turns into sand (in billions of tons). W_sc represents the average annual or flood season sediment discharge volume of the river section (in billions of tons).

The section from Sanhu Estuary to Toudaoguai River has no drainage channel for water discharge, and the yellow wind turning into sand is mainly carried by the flood through the Ten Major Kongdui. Therefore, there is no need to consider W_sp and W_sf.

R² test method

The coefficient of determination R² is used to measure the extent to which a regression model explains the variation of a continuous dependent variable. It is used to test whether there is a significant association between two categorical variables or whether the distribution of a categorical variable significantly differs from the theoretical distribution. In scientific research, the coefficient of determination R² is a core indicator for evaluating the quality of model fitting. It can be used to assess the effect, make decisions, control risks, modify models, and conduct research. When testing the rationality of a model or empirical formula, the coefficient of determination R² is only used to evaluate the goodness of fit of the “regression equation” to continuous data, the larger the value, the more reasonably the equation fits the sample data. The relationship between the annual scouring and siltation volume of the river section using the empirical formula and the measured values was calculated by using the coefficient of determination R² to determine the rationality of the empirical formula. The calculation formula of the coefficient of determination R² method is as follows:

3

In the formula: is the observed value, is the predicted value, and is the mean of the observed values. The closer it is to 1, the stronger the explanatory power of the regression equation for sample variation.

Results

Effectiveness of new check dams in the Xiliugou watershed for water and sand retention

Taking the Xiliugou Basin as the object, a running-sediment model capable of simulating the action of a check dam was constructed and calibrated based on the reservoir module of the SWAT model. Set up three scenarios and compare the differences in labor and sand production among them.

Scenario one simulation result

The simulation results of Scenario One are shown in Fig. 3.

Fig. 3.

Comparison of simulated and measured (a) annual runoff and (b) sediment transport volume in Scenario 1 (assuming no check dams).

It can be seen from Fig. 3(a) that the simulated runoff values during the period from 2000 to 2020 increased significantly. The measured average annual runoff volume during this period was 1.73 × 10⁷ m³, while the simulated value was 2.19 × 10⁷ m³, with a difference of 26.72%. The reason for this is that the vast majority of the check dams in the study area were built after 2000. In Scenario One, Since the measured data from 2000 to 2020 already include the retention function of the check dam, Scenario 1 is bound to overestimate the runoff and sediment. Its sole function is to provide an upper limit reference.

It can be seen from Fig. 3(b) that the changing trend of the simulated annual sediment transport value during the period from 2000 to 2020 is basically consistent with the measured changing trend of the annual sediment transport. During this period, the simulated annual sediment transport value differs greatly from the measured value by 79.7%. The reason for this is that in Scenario One, the simulation of the annual sediment transport volume during the period from 2000 to 2020 did not take into account the influence of sediment reduction by check dams, and thus would be larger than the measured value.

In Scenario One, the simulated value of the average annual runoff during the period from 2000 to 2020 increased by 4.62 × 10⁶ m³ compared with the measured average annual runoff volume. The simulated value of the average annual sediment transport during the period from 2000 to 2020 increased by 7.06 × 10⁵ t compared with the measured average annual sediment transport volume.

Simulation results of scenario two

The simulation results of Scenario Two are shown in Fig. 4.

Fig. 4.

Comparison of simulated and measured (a) annual runoff and (b) sediment transport volume in Scenario 2 (assuming partial check dams).

It can be seen from Fig. 4a that the simulated runoff value decreased during the period from 2000 to 2020. The measured average annual runoff volume during this period was 1.73 × 10⁷ m³, while the simulated value was 1.28 × 10⁷ m³, with a difference of 26.01%. The reason is that the vast majority of the check dams in the study area were built successively around 2000. The model simulates the existing check dams that have been built. But in fact, these check dams have not yet been fully completed. Therefore, the measured runoff volume is larger than the simulated runoff volume value.

It can be seen from Fig. 4b that the simulated value of annual sediment transport during the period from 2000 to 2020 is not much different from the measured value. The reason is that in Scenario Two, the simulated value of annual sediment transport is affected by the sediment reduction effect caused by the setting of check dams, and the simulated value is smaller than the measured value.

Scenario Two is the sole bridge connecting “observed reality” with “policy scenarios”, and it can be used to quantify the “historical contribution of existing dams” and the “upper limit of new dam increments”. Without it, the SWAT output cannot be transformed into figures that can be directly cited by the management department. Scenario Two reproduces sediment load well, but overestimates annual runoff during the period of dense dam construction, mainly due to the static CN value and the linear base flow assumption. Future work will introduce dynamic CN and MODFLOW coupling to improve the rapid runoff generation and groundwater compensation process. At the same time, since the design reservoir capacity was adopted in the model establishment without considering siltation, it may lead to an overly large interception data of the dam reservoir. In subsequent studies, the terrain after siltation was updated dam by dam one by one, and the dam was dynamically opened and closed according to the “remaining reservoir capacity - water level” curve to reduce errors.

In Scenario Two, the simulated value of the average annual runoff during the period from 2000 to 2020 decreased by 4.52 × 10⁶ m³ compared with the measured average annual runoff volume. The simulated value of the average annual sediment transport during the period from 2000 to 2020 decreased by 1 × 10⁵ t compared with the measured average annual sediment transport volume.

Simulation results of scenario three

The simulation results of Scenario Three are shown in Fig. 5.

Fig. 5.

Comparison of simulated and measured (a) annual runoff and (b) sediment transport volume in Scenario 3 (assuming full check dams).

It can be seen from Fig. 5(a) that the simulated runoff value decreased during the period from 2000 to 2020. The measured average annual runoff volume during this period was 1.73 × 10⁷ m³, while the simulated value was 1.03 × 10⁷ m³, with a difference of 40.29%. The reason is that due to the functions of water reduction, regulation and storage, and flood peak reduction of check dams, setting up check dams in the study area will lead to a reduction in the simulated annual runoff.

It can be seen from Fig. 5(b) that the simulated value of annual sediment transport during the period from 2000 to 2020 differs significantly from the measured value. The simulated value has decreased by 57.43%. The reason for this is that in Scenario Three, the simulated value of annual sediment transport is affected by the sediment reduction effect caused by the setting of the check dams, and the simulated value is smaller than the measured value, indicating that the sediment reduction effect of the check dams is obvious.

In Scenario Three, the simulated value of the average annual runoff during the period from 2000 to 2020 decreased by 6.96 × 10⁶ m³ compared with the measured average annual runoff volume. The simulated value of the average annual sediment transport during the period from 2000 to 2020 decreased by 5.09 × 10⁵ t compared with the measured average annual sediment transport volume.

Search for the data of the maximum peak flood flow at the Longtouguai Hydrological Station in Xiliugou from 1964 to 2016 for a total of 38 years (among which, the data from 1991 to 2005 is scarce), and analyze the reasons why the runoff and sediment transport in 2006 and 2008 were slightly more prominent than those in other years in the annual runoff comparison chart and the annual sediment transport comparison chart. After calculation, the average value of the series of flood peak flow is 778m³/s. Among the analysis and research years from 2000 to 2020, the maximum peak flood flow in 2006 was 1090m³/s, and in 2008 it was 1100m³/s. SWAT can only provide a “reasonable trend” for “extreme peaks”, and cannot be expected to accurately reproduce the extreme values of flood peaks²⁸. Due to the large flood flow and high sediment content, errors will occur between the measured runoff and sediment transport volume and the simulated data.

Based on the simulation results of the SWAT model, the average annual sediment discharge calculated by setting 113 existing check dams during the period from 2000 to 2020 was subtracted by setting 113 existing check dams and 79 newly built check dams, resulting in an average annual sediment discharge of 4.09 × 10⁵ t for the 79 newly built check dams.

SWAT simulation (2000–2020) indicates, the 79 new check dams would reduce mean annual runoff by 2.44 × 10⁶ m³.

Calculation of the sediment transport volume of the top ten Kongdui

Since there is only one hydrometric station below each of the Maolun, Xiliugou, and Han Tai channels among the ten Kongdui, the monitoring data is not continuous, the area ratio method only needs one parameter of the catchment area, so it can also be used in the zero data area. The method of ratio of the sediment yield model directly enlarges the sediment discharge intensity according to the area, which is suitable for the basins with lack of sediment data but similar topography and climate. Therefore, the sediment yield of other Kongdui was calculated by using the historical hydrological data of the Longtouguai Hydrological Station in Xiliugou through the area comparison method and the sediment transport modulus ratio method. The calculation results are shown in the following table (Table 4).

Table 4.

Top ten porous sand transport table.

Year	Sediment transport volume(104 t)	Year	Sediment transport volume(104 t)
1998	16427.7	1999	1404.37
2000	17.24	2001	16.11
2002	5.95	2003	4894.00
2004	29.28	2005	6.58
2006	1004.08	2007	19.27
2008	1044.61	2009	2.39
2010	10.38	2011	1.23
2012	18.37	2013	14.83
2014	60.41	2015	1.69
2016	1370.00	2017	46.18
2018	70.25	2019	2.44
2020	8.48

Calculation of erosion and siltation in the inner Mongolia section of the yellow river

The sediment inflow and inflow in the Inner Mongolia River section (from Sanhu Estuary to Toudaoguai River section) mainly come from the Ten Major Kongdui on the south bank of the Yellow River. Since the sediment interception project set in this study is located in Xiliugou, the water and sediment conditions of the study area and other Kongdui need to be treated differently.

According to the current situation and the study conditions of the dry river three lakes estuary station and interval tributary water and sand series data, using the formula (1) can be calculated to get the annual sand loss of Toudaoguai station, after calculating the 2008–2020 Toudaoguai hydrological station in the simulation of the new 79 check dams after the average annual sand loss of 6.27 × 10⁷ t. On this basis, by using Formula (2), the flood season silt removal volume of the section from Sanhu Estuary to Toudaoguai River can be calculated. Furthermore, the average annual silt reduction volume of the Inner Mongolia river section after the construction of the sand control project can be obtained as 7 × 10⁵ t.

R² test result

The annual sediment discharge results of the Toudaoguai Station in the Inner Mongolia river section were calculated by using the sediment discharge formula. Further, the annual sediment flushing and siltation volumes of the Inner Mongolia River section (from Sanhu Estuary to Toudaoguai River section) from 2008 to 2020 were calculated by the sediment discharge rate method. The correlation coefficient between the calculated annual sediment flushing and siltation volumes of the river section and the measured values is 0.92, thirteen data points were selected for calculation, and the abnormal data points of the 2016 extraordinary flood were removed, indicating that the method used in this paper is relatively reliable. This formula can be used to calculate the scouring and siltation impact on the Inner Mongolia river section after the water and sediment reduction of the Xiliugou check dam project.

Discussion

The sediment reduction of check dams in water and soil conservation engineering measures

In this paper, by establishing a SWAT model to simulate the water and sediment reduction in the Xiliugou Basin caused by the construction of 79 newly-built check dams, it is concluded that the average annual sediment retention capacity of the 79 newly-built check dams is 4.09 × 10⁵ t, and the average annual water retention capacity is 2.44 × 10⁶ m³.

As the main river structure on the Yellow River and the Loess Plateau rivers, the check dams can not only essentially raise the reference level, prevent gully cutting and source erosion, but more importantly, effectively store sediment and prevent sediment accumulation in the river channel. Vigiak Olga et al. ²⁹ based on the SWAT slope sediment yield module, a sediment budget model at the scale of the upper Danube River basin was constructed, confirming that the “sediment retention - erosion reduction” mechanism still holds in basins of > 10⁴ km². In Sect. 3.1 of this paper, the same approach is adopted to embed the silt dam module into SWAT and complete the verification in the 1,356.3 km section of the Xiliugou Kongdui, which echoes Vigiak’s cross-basin conclusion formation method - scale double. Gao et al. ³⁰ summarized the construction intensity of check dams in the middle reaches of the Yellow River and evaluated the sediment reduction efficiency in different periods. Section 4.1 of this paper, based on the dam coefficient data from 2000 to 2020, calculates that the average annual sand retention capacity of the 79 newly built check dams in the study area is 4.09× 10⁵ t. The results showed that from 1952 to 2011, the cumulative sediment reduction of check dams in the middle reaches of the Yellow River was 805.6 million tons. Ran et al. ³¹ found that compared with other soil and water conservation measures, the construction of check dams for sand interception has a more obvious effect on reducing the entry of sediment into the Yellow River. The research conducted by Ran et al. ³² in the Huangfu River Basin of the middle and upper reaches of the Yellow River found that the amount of sediment intercepted by the check dam in the Huangfu River Basin from 1990 to 1996 was approximately 9.7 × 10⁵ t. In conclusion, whether it is macro long sequence statistics, measure contribution ranking, or event scale measurement in typical river basins, previous studies have quantitatively confirmed the significant sand-retaining function of check dams. Based on multi-source data and the construction of the SWAT model, this paper provides independent evidence of the good sand-retaining effect of the check dam, which forms a cross-scale and cross-method mutual confirmation with the above conclusion.

In recent years, numerous research results have shown that the impact caused by human activities is an important reason for the significant reduction in the amount of sediment flowing into the Yellow River from the Loess Plateau in the middle and upper reaches of the Yellow River^32–35. Among them, the construction of check dams plays an important role in reducing erosion and sand production, as well as intercepting sediment flowing into the Yellow River³⁶. How to accurately calculate the sediment retention benefit of check dams is of great significance for soil and water protection and governance in this area, as well as the operation and maintenance of check dams. Therefore, many scholars have conducted extensive research on the calculation method of sediment retention capacity of check dams and achieved rich results. The existing methods include:1. The field investigation method, although it has a relatively strong credibility, is considered time-consuming and labor-intensive in actual work, and the number of check dams is relatively large, and it is greatly affected by the climate during the investigation period. At the end of 2014, Ordos City launched a city-wide investigation into the benefits of sediment retention by check dams and a hidden danger investigation.2. The topographic surveying method, by means of the measured topographic map of the channel where the check dam is located or by using 3 S technology, etc., extracts the information of the check dams and generalizes or interpolates the topography of the channel. Then, combined with the siltation elevation of the dam in different years, the total siltation sediment volume of the check dam is estimated by using volume calculation formulas such as prisms or cones^37,38. Zeng et al. ³⁹ combined unmanned aerial vehicle (UAV) photogrammetry with simulated inundation analysis and proposed a new method for estimating dam-blocking siltation sediments in complex terrains. Five models were established through regression analysis. The verification showed that the errors in the volume estimation of the optimal models of single-retaining check dams and regional retaining check dams were 12–13% and 2–3%, respectively. However, since this method relies on high-precision DEM data and is unable to calculate the sand-blocking data for each period and simulate the impact of check dams construction, this method is not applicable to this study. 3. Model simulation method, the model simulation method refers to the simulation of hydrological processes in river basins based on physical mechanism hydrological models, which can distinguish the influence of various elements on water and sediment changes. The models used include models such as SHE, SEDD, SWAT, MIKE, VIC, etc. During the development process of these models, different functional modules were also provided according to the actual situation, with more simulation functions, and have been widely applied in many fields. In this paper, the SWAT model is adopted for simulation, which can better reflect the erosion status of the basin. Moreover, by setting different working conditions to control the variables, the contribution of check dams to the sediment transport changes in the basin can be obtained. Three main methods are commonly used to estimate sediment retention: (1) field investigation, (2) topographic survey, and (3) model simulation. Each has advantages and limitations. In this study, the SWAT model was adopted because it can distinguish the influence of check dams under different scenarios.

With the increase in the number of check dams, the proportion of check dams in reducing runoff and sediment is also increasing, indicating that the contribution of check dams to the reduction of annual runoff and sediment was greater from 2000 to 2020 than from 1990 to 1999. Therefore, check dams are the main factor in retaining water and sediment in the Xiliugou Basin.

However, according to the statistics of 1,200 check dams in the middle reaches of the Yellow River, it takes 12 to 15 years for the sediment retention efficiency to drop from approximately 80% to 30%. Meanwhile, in the subsequent improvements, the time-dependent trapping efficiency, TE should be verified and the correction coefficients should be corrected. Therefore, if the goal is to reduce the coarse sediment of the Yellow River in a “long-term, efficient and controllable” manner, in the subsequent construction of check dams, a cascade dam system combined with regular sediment removal should be considered.

The impact of check dams construction on the Inner Mongolia section of the Yellow River on the scouring and siltation of tributaries of the Yellow River

The annual sediment discharge of the Toudaoguai Station was calculated by using the empirical formula of the annual sediment discharge of the Sanhu Estuary - Toudaoguai River section. According to the sediment discharge rate method, the annual sediment discharge of the Sanhu Estuary - Toudaoguai River section from 2008 to 2020 was calculated by using the established sediment discharge formula. It was calculated that the average annual sediment reduction of the Inner Mongolia section of the Yellow River after the construction of 79 new sand control check dams was 7 × 10⁵ t.

The source of sediment in Inner Mongolia section of the Yellow River is complex, coarse sediment mainly comes from the desert along the Yellow River and its underlying arsenic sandstone area, the Yellow River through the Ulanbuh Desert and Kubuqi Desert is about 620 km, sand particles in the desert as well as the arsenic sandstone are easy to be disintegrated when they meet water, and enter into the Yellow River through the action of wind and water, which becomes a major source of near-source coarse sediment; Due to the relatively low gradient of the Inner Mongolia section of the Yellow River (approximately 1/10,000–2/10,000), it is highly prone to the formation of an above-ground suspended river caused by sediment deposition. Meanwhile, once the sediment in the Inner Mongolia section enters the lower reaches, it will affect the water-sediment relationship in the lower reaches of the Yellow River, leading to the continuous shrinkage of the main channel in the lower reaches and a significant decline in flood discharge capacity. The sediment deposition and the “suspended river” phenomenon in the lower reaches of the Yellow River pose a serious threat to the economic development and social stability of the areas along the river^40,41. After the source of coarse sand was cut off, the average annual dredging cost for the lower reaches of the Yellow River could be reduced by 110 million yuan. Moreover, the water depth of the Grade IV channel could be maintained from 180 days to 220 days, taking into account flood control, navigation and the safety of the lives and property of 920,000 people in the floodplain area. Therefore, the Xiliugou check dam is not a local water and soil conservation project, but a source control node for the “desert-suspended river” chain disaster of the Yellow River. Its sediment reduction benefits can be gradually magnified along the 620-kilometer desert section downstream, providing a solution to the entire Yellow River’s sediment predicament.

To sum up, the sediment in the Inner Mongolia section of the Yellow River has multiple impacts on the river. It not only directly affects flood control safety but also has profound influences on ecology, economy and society^42–44. Therefore, solving the problem of sediment deposition in the Inner Mongolia section of the Yellow River is not only considered from an environmental perspective, but more importantly, it is jointly addressed from economic and social aspects. The construction of soil and water conservation projects in the Xiliugou Basin can effectively help reduce sediment deposition in the Inner Mongolia section of the Yellow River. From the calculation results, the construction of check dams in the Xiliugou Basin can effectively reduce sediment inflow into the Inner Mongolia section, thereby mitigating siltation and enhancing flood control capacity.

Deficiencies and prospects

In this paper, the SWAT model is adopted for the simulation of the water-sediment effect in Xiliugou. This model takes the reservoir as the outlet point of the sub-basin to describe that during the confluence process, the runoff and sediment from the upper reaches of the river channel are intercepted, and the runoff and sediment outflow flows into the lower reaches of the river channel. Check dams and small reservoirs have similar characteristics, both playing a role in retaining sediment runoff, preventing soil erosion and conserving soil and water. However, there are also differences between the two. On the one hand, during the process of intercepting sediment in check dams, silt fields will form inside the dam. These silt fields may subsequently grow plants, playing a certain role in reducing water and sediment. On the other hand, the service life of check dams is unlike reservoirs, check dams have shorter service lives (20–30 years for main dams, shorter for small and medium dams). If the check dam is not dredged after it is full, its function of flood and sand retention will be significantly reduced or even fail. When water and sand come from the upstream, they will directly flow into the main river trunk. In the future, when the current SWAT reservoir module cannot fully capture the temporal evolution of check dams. Future improvements should include a specialized check dam module to simulate sediment storage capacity over different life stages.

One of the key measures to achieve sediment reduction in the rivers of Inner Mongolia is to reduce the sediment intake of the ten major Kongdui. When calculating the sediment production data of the ten major Kongdui, methods such as Lin Xiuzhi were adopted to calculate the sediment production of the remaining Kongdui based on the sediment production data of Xiliugou. However, due to the scarcity of observational data of the ten major Kongdui, it is quite difficult to accurately assess their annual sediment transport. In addition, the construction of the Xiliugou check dam may also interfere with the final results to some extent. Therefore, future research should focus on (a) improving data availability for tributary sediment loads, (b) refining models for check dam life-cycle performance, and (c) integrating hydrological, ecological, and socio-economic impacts of sediment reduction^45,46.

Conclusions

This study adopted a distributed hydrological model, with the period from 1990 to 2000 as the rate period and the period from 2001 to 2020 as the research period. The SWAT model was used to calculate the impact of the construction of 79 sediment check dams in the Xiliugou Basin of the Ten Major Kongdui on the water and sediment in the Xiliugou Basin, and the empirical formula was used to calculate the sediment production in other basins of the Ten major Kongdui. Based on this, the sediment discharge formula was used to calculate the scouring and siltation influence of the newly-built check dams on the Inner Mongolia section of the Yellow River (from Sanhu Estuary to Toudaoguai section), and the rationality of the formula was verified through R². The main conclusions are as follows:

1. When other parameters remain unchanged, the check dam has significantly reduced the average annual runoff volume of Xiliugou. The reduction rate increases with the expansion of the check dam scale, with the maximum reaching 23%. During the period from 1990 to 1999, no check dam was built in Xiliugou. In Scenario 2 (partial check dams), the percentage decrease in the simulated average annual runoff compared with the measured average annual runoff is 10.87%. In Scenario 3 (full check dams), the simulated average annual runoff volume decreased by 23.11% compared with the measured average annual runoff volume. During the period from 2000 to 2020, the average annual runoff volume simulated in Scenario 3 decreased by 5.02 × 10⁶ m³ compared with the measured average annual runoff volume, and the average annual runoff volume simulated in Scenario 3 decreased by 2.44 × 10⁶ m³ compared with that simulated in Scenario 2.

2. When other parameters remain unchanged, the simulation results of sediment transport volume with and without sediment check dams vary greatly, and the sediment reduction effect of sediment check dams is obvious. During the period from 1990 to 1999, in Scenario 2, the percentage reduction in sediment between the simulated annual average sediment transport and the measured annual average sediment transport is 12.63%. In Scenario 3, the percentage reduction in sediment between the simulated annual average sediment transport and the measured annual average sediment transport is 26.48%. During the period from 2000 to 2020, the average annual sediment transport simulated in Scenario 3 decreased by 4.09 × 10⁵ t compared with that in Scenario 2.

3. The newly-built soil and water conservation projects in the study area have a significant impact on the erosion and siltation of the main stream of the Yellow River. The rationality of the formula was tested through R², and the result reached 0.92. The construction of 79 check dams in the study area can reduce siltation by 7 × 10⁵ t in the Inner Mongolia River section (from Sanhu Estuary to Toudaoguai River section).

Check dams “reduce water and sediment” - runoff can be reduced by up to 23%, and sediment transport can be cut by up to 26%. If 79 new check dams are built as planned, the main stream of Inner Mongolia alone can reduce sediment accumulation by 700,000 tons, with a significant demonstration effect.

Author contributions

Conceptualization, W.Z.; Z.Z.; Q.Q. and W.F.; data interpretation and methodology, X.Z. and W.F.; validation, W.H.; Y.W.; P.T. and Y.L.; software, W.L. and W.F.; original draft preparation, X.Z.; funding acquisition, R.H.; D.Z. and W.Z.; All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by Key Special Project of the “Science and Technology Revitalization of Mongolia” Action (grant number 2022EEDSKJXM004-4), National Natural Science Foundation of China (grant number 42401022), Key Research and Development and Technology Transfer Program Project of Inner Mongolia Autonomous Region (2025SYFHH0219), Ordos Major Science and Technology Project - Research on the Integrated Scheduling Technology of Recycled Water and Other Multiple Sources in Ordos City (ZD20232323), Special project of basic scientific research business expenses of China Academy of water resources and hydropower (Grant No.MK0145B012021), Key R&D and Achievement Transformation Program of Inner Mongolia Autonomous Region (2025YFHH0005).

Data availability

Data available on request from the authors. The data that support thefindings of this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.