Life cycle water footprint assessment of concrete production in Northwest China

As an essential natural resource on Earth, fresh water accounts for only 2.53% of the total water on Earth (Hotlos, 2008). Bruinsma (2009) predicts that water consumption for energy production will increase by 60% in 2050. The shortage of water resources, like population and environmental issues (Maja & Ayano, 2021), has become a global hot issue. Therefore, the rising water demand generates the need for more effective management of freshwater resources. For example, in arid and water-scarce areas, diversification of water supplies better ensures the stability of the water quality, even in emergencies (Boryczko & Rak, 2020). However, diversification of water supplies faces enormous challenges, such as the overuse of groundwater in arid areas (Luker, 2017) and the general exclusion of promoting the use of reclaimed water (Wu et al., 2020). Allan (1998) first proposed the concept of water footprint (WF) in 1998, and then it expanded to different types of consumed water, which led to the full application of the WF method in water consumption management in various industries worldwide (Gu et al., 2015). WF can be a useful tool to track the use of water in the environment through the product life cycle (Chen et al., 2010).

Concrete is one of the most common building materials, playing a core role in the construction industry (Kuzman et al., 2018). The whole life cycle of concrete production requires a massive volume of water. To reduce the impact of the concrete production on the environment, it is necessary to quantify its measurement in order to improve the water management levels of the concrete industry. Although WF has been used in many industries, such as crop and food production (Deng et al., 2018; Rossi et al., 2019; Karandish et al., 2021), meat (Naoum et al., 2018), and virtual water flows in global food trade (Vos et al., 2019), the application of WF in the concrete industry is still less. The researchers measured the direct water consumption in the concrete industry. Chen et al. (2010) examined the direct water intensity of cement production in France and estimated its value at 0.2 m³/ton. Study shows that the direct water intensity of cement production is 0.185–0.808 m³/ton (Elmer, 2015). Although direct water consumption measurement provides useful water management data in the concrete industry, it cannot reflect the concrete production's virtual water information. In the concrete industry, Netz & Sundin (2015) put forward the evaluation method of concrete WF in 2015. Mack et al. (2016) proposed a simplified method for measuring WF of concrete in 2016. Bardhan (2011) calculated the consumption of water resources in Indian construction. Hosseinian & Nezamoleslami (2018) calculated that the WF of concrete production in Iran is 2.321 m³/ton. Using the 2015 global concrete production data, Vergara & John (2019) estimated that the global concrete water consumption is about 180 × 10⁸m³ (63.2–304.3 × 10⁸m³), and the WF is about 0.547–2.634 m³/ton. These studies further deepen the application of WF in concrete water management, but the research on the WF of concrete production is still insufficient (Mack et al., 2016). Concrete production causes severe environmental problems due to the large amount of water consumed and wastewater produced. Unfortunately, the limited data on water consumption in the concrete industry makes it difficult for the government and relevant organizations to find the strategies and technologies needed to reduce water consumption. Therefore, it is necessary to continue to study and measure the LCWF of concrete production to inform the formulation of regional development policies.

Concrete is the core material of the construction industry. Due to the high demand, there is no other material to replace concrete (Naik, 2008). Since it cannot be replaced, more sustainable ways to produce and use concrete must be developed to meet the needs of society and to reduce the environmental impact. Therefore, it is of great practical significance to conduct an empirical study on the LCWF of concrete production in Northwest China, where water resources are extremely scarce. In this study, based on Life Cycle Assessment (LCA)'s concept, Water Footprint Network (WFN) methods were used to evaluate the LCWF of concrete production in Northwest China.

In Northwest China (Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang), the ecological environment is fragile due to the inherent shortage of water resources, insufficient surface water, and overuse of groundwater (Wang et al., 2018a, 2018b). However, Northwest China accounts for 6% of the country's concrete production and 41.7% of the cement. Therefore, Northwest China is chosen as the research area for this study.

According to the position of the analyzed product or service in a life cycle, LCA was divided into three different research boundaries (Jacquemin et al., 2012) (see Figure 1). During the production process (for example, for placement and maintenance), the water consumption of concrete varies significantly depending on the end-use (such as bridges, buildings, structures). There is a lack of statistics for this part of the data. Similarly, there is a lack of corresponding data statistics for demolition and recycling of concrete final products at the end of life. We believe that the water consumption caused by the use and removal of concrete final products in specific buildings or structures can be calculated. Therefore, this study mainly involves the WF of the actual production process.

The LCWF of concrete production was divided into direct and indirect (virtual) water consumption in the concrete production process. So, it was divided into direct water footprint (DWF) and virtual (indirect) water footprint (VDF). The DWF includes the direct water consumption of concrete production and enterprise daily operation, the sum of blue water, green water, and ash water. The VWF includes the sum of all inputs in the production process of an enterprise (Hosseinian & Nezamoleslami, 2018). The model framework is shown in Figure 2.

is the number of input elements. is the WF of input factors. The value of in this paper mainly comes from the relevant literature and the indirect calculation based on the relevant literature.

β is the sensitivity coefficient of LCWF of concrete production to parameter P. is the variation of parameter P. is the corresponding change of LCWF when the parameter P changes.

The WSI can be divided into four grades: Mild pressure (0, 0.3]; Moderate pressure (0.3, 0.6]; Heavy pressure (0.6-1]; Extreme pressure [1, +∞).

According to the registration information of China Industrial and Commercial system, there are a total of 4705 concrete producers in operation in Northwest China (as of November 31, 2019). To accurately represent the average level of concrete production in Northwest China, the sample size is selected according to the stratified random sampling method of regional stratification. Due to the similar resources and environment in Northwest China, and the similar level of production technology, there is no significant difference among the provinces. Therefore, the sample size is determined by the simple random sampling method, and then the samples are allocated in a fixed proportion among the layers. Among them, the confidence interval is 95% (Z = 1.96), the error limit is e = ±5%, the design effect is B = 1, the expected response rate is r = 1, and it is assumed that the variance P is the maximum, P = 0.5. Therefore, the total sample size and the number of samples required by the provinces are shown in Table 1.

Cement and coarse aggregate are essential raw materials in concrete production. The LCWF of cement and coarse aggregate is regarded as the VWF of input materials in concrete production. According to the inspection data of each concrete producer, the cement and coarse aggregate supply companies are identified as targets for further investigation.

According to Figure 2, the DWF of concrete, cement, and coarse aggregate mainly includes the water consumption of machinery and equipment, the water used in the production process, and employees' daily direct water consumption. The type of direct water consumption varies among producers, but the final total consumption is the total annual water consumption of the enterprise. This part of the data can be obtained through a sample survey of enterprises.

According to formula 3, the WF of energy consumed by concrete, cement, and coarse aggregate can be further calculated. Among them, the annual power consumption of concrete, cement, and coarse aggregate, the amount of diesel used by machinery, the mileage of commuter cars, the consumption of coking coal, the amount of natural gas, and the transportation distance of various input raw materials are obtained through the investigation of the sample enterprises.

1. WF of energy. In China, gasoline, diesel, and coal are the main types of energy for industrial production, heating, and power generation (Khanna et al., 2019). In 2018, 70% of China's electricity was generated from coalfired power plants (Yuan et al., 2016). The electricity produced by different energy sources in China is merged into the State Grid, which provides electricity to enterprises. However, according to the survey, the power production in Northwest China is mainly thermal power, so the WF of electricity in this study mostly refers to thermal power data. The WF of each energy type is shown in Table 2.

2. The WF of working meal. There is a lack of literature records of WF of meals in China, and it is not easy to calculate WF for different enterprises. Spiess (2014) has estimated the WF of a typical working meal based on food market prices, food costs, and WF of food ingredients. Therefore, the WF per meal can be calculated by the following formula:

   

   (7)

• 3.

  WF of input raw materials. The raw materials put into concrete production mainly include cement, coarse aggregate, sand, fly ash, and admixtures.

In this study, ordinary Portland cement, which is the most widely used type of cement, is used as the subject. The sample enterprises mainly adopt the new dry process for production. Cement production includes the acquisition, processing and preparation of raw materials (mainly limestone and clay), the processing and calcination of clinker, and the transportation, storage and recycling of cement products (Schneider et al., 2011). Cement is one of the most critical raw material inputs. The LCWF is decomposed and calculated concerning the composition of concrete production (see Figure 2).

Coarse aggregate accounts for 60–75% of the concrete volume, so WF within coarse aggregate is significant. The coarse aggregate production stage involves raw material taking, washing, material classification, transportation, and storage. Water is consumed at every step. Water consumption depends on different parameters (such as extraction process, stone source, equipment used, and type of energy consumption), making it very difficult to calculate the WF of total production. According to the classification of shape, the coarse aggregate can be divided into pebbles and gravel. Pebbles are mainly dredged from the riverbed, while the gravels are broken from mine gravels (Limantara et al., 2018). Due to the difficulties in adopting stones and environmental protection requirements, the concrete industry has no longer used pebbles as the source of coarse aggregate for concrete production in recent years. Therefore, only data for gravel are collected in this study and only water consumption for cleaning, sorting, crushing and transportation during the production process is considered. The WF of coarse aggregate is decomposed and calculated regarding the composition of the LCWF of concrete production (see Figure 2).

WF of sand, fly ash, and admixtures. In Northwest China, most of the sand used as the raw material of concrete is directly obtained from the river channel, so this study only considers the WF of the fuel consumed by mining and excavation and the water content of the sand itself. The moisture content of sand for concrete is determined to be 4% (Chinese Standard JGJ52-2006, 2006). The primary sources of fly ash are coal-fired power plants and urban central heating boilers, treated as waste, and their WF is generally calculated through electric. Therefore, this study only considers the moisture content of fly ash itself. The moisture content of fly ash for concrete is determined to be 1% (Chinese Standard GB1596-2017, 2017). Simultaneously, the WF of the admixture used in concrete production varies significantly among different sources, and due to the lack of literature and the relatively small consumption in concrete production, the VWF caused by the concrete admixture itself is not considered in this study.

Through the survey of 355 enterprises, the average yearly concrete production capacity in Northwest China is 4.16 × 10⁴ m³, and the average yearly direct water consumption is 7,367.08 m³ (2018). The production scale and production technology level of concrete enterprises in the five provinces of Northwest China are similar, with no significant difference.

The LCWF of concrete production in Northwest China is 2,159 L/m³. Among them, DWF is 177.03 L/m³, and VWF is 1,981.97 L/m³. Thus, it can be seen that there is little difference in the technical level of concrete production in northwest China, so the DWF is basically at the same level, accounting for only 8.2% of the LCWF. Thus, VWF is the most critical factor affecting LCWF. In VWF, the water footprints of Meals, Cement, and Aggregate are 760.01 L/m³, 379.78 L/m³, and 776.61 L/m³, respectively, making those parameters are the most important components of the VWF, so they are also the main controlling factors. The calculation results of the LCWF of concrete production in each province are shown in Table 4.

Based on the production data of cement and coarse aggregate, we further calculated the LCWF of cement and coarse aggregate (see Figure 3).

According to the calculation, the contribution of each parameter of LCWF of concrete, cement, and coarse aggregate production are shown in Figure 4.

1. The contribution of each parameter in concrete production to the LCWF is shown in Figure 4(a). The DWF of the concrete output is 177.03 L/m3, which is lower than that of Holcim (2015) and Hosseinian & Nezamoleslami (2018). The VWF is 1,981.97 L/m3, which is slightly lower than Holcim (2015) and Hosseinian & Nezamoleslami (2018).

2. The contribution of each parameter in cement production to the LCWF is shown in Figure 4(b). Cement has the most massive DWF (702 L/ton), followed by electricity (337.17 L/ton), working meal (191.31 L/ton), and coking coal (115.69 L/ton).

3. The contribution of each parameter in coarse aggregate production to the LCWF is shown in Figure 4(c). In the LCWF of coarse aggregate production, the contribution of electricity WF was the highest (323.66 L/ton), followed by DWF (216.23 L/ton) and working meal WF (170.21 L/ton). In this study, the DWF of coarse aggregate production is half of that resulted from Hosseinian & Nezamoleslami (2018), (460–540 L/ton), and the food and working meal WF is also significantly lower than that of Hosseinian & Nezamoleslami (2018). (376.8 L/ton).

By comparison, it is found that the DWF of concrete, cement, and coarse aggregate production in Northwest China is the same, which shows that the production technology of concrete, cement, and coarse aggregate is mature and the DWF is stable. DWF is not the main factor affecting LCWF.

It can be seen from Figure 4, the VWF accounts for 91.8%, 48.79%, and 69.59% of the LCWF of concrete, cement, and coarse aggregate, respectively. These data show that the large water consumption of the concrete industry is closely related to VWF, which is the focus of LCWF assessment.

As can be seen from Figure 3, the VWF of concrete, cement, and coarse aggregate production varies significantly from province to province. The VWF of the concrete output in Gansu, Ningxia, and Qinghai is high; the VWF of cement production in Xinjiang is high; as for the coarse aggregate production, the VWF of Ningxia, Qinghai, and Xinjiang are high. Generally speaking, Gansu, Ningxia, and Qinghai have higher levels of virtual water use, which may be related to the relatively underdeveloped economy and poor transport infrastructure of these three provinces, which has led to more indirect water investment. Figure 3 shows that when the LCWF of cement production is the same, the LCWF of concrete production in each province increases with the LCWF of coarse aggregate production. It shows that coarse aggregate plays a decisive role in the LCWF of concrete production.

It can be seen from Figure 4(a) that in addition to direct water consumption, working meal and other material inputs are the main factors affecting the LCWF of concrete production, and the impact of energy on the LCWF of concrete production is almost negligible. Due to the differences in diet structure and labor input, the WF of working meal provided by the sample concrete factory in this study is higher than Hosseinian & Nezamoleslami (2018) and Spiess (2014). Cement, coarse aggregate, and working meal account for the largest proportion of the LCWF for concrete production, as they account for 88.76% of the LCWF and for 96.69% of the VWF, which highlights the importance of reducing the WF of working meal, cement and coarse aggregate production. The WF of working meals is very large, which indicates that the concrete industry in Northwest China is still labor-intensive. Further improvement of mechanization is the key to reducing the WF of working meals. Simultaneously, although the production of cement is nearly twice as large as the production of coarse aggregates, the LCWF of coarse aggregates is much higher in concrete production due to the fact that coarse aggregates account for 48.13% of the direct material input to concrete production. Therefore, recycling and reuse of coarse aggregate may be the solution to reduce the WF in the future.

As can be seen from Figure 4(b), DWF contributes 51.21% of the cement production WF, followed by electricity, working meal, and coking coal. Most of the direct water is used to wash raw materials and flush equipment. Reducing the amount of water used for raw material cleaning in cement production can effectively reduce the WF of cement production. At the same time, according to the research of Hosseinian & Nezamoleslami (2018), the WF of natural gas is 9.251 L/m³, while the WF of coking coal is 4.8 m³/ton, so the ratio of calorific value per kilogram of natural gas to coking coal is 1:0.8. The combustion efficiency of natural gas is 90%, and the combustion efficiency of coking coal is 40%, so burning 1-ton coking coal is equivalent to the efficiency of burning 320 m³ of natural gas, while the water consumption of burning 320 m³ natural gas is 2.96 m³, and the price is only 2/3 of coking coal. Natural gas has advantages in terms of water consumption and cost. Among the 355 samples, only 69 enterprises use natural gas as fuel energy, accounting for about 24.13%. Therefore, converting coal to natural gas can become an effective way to reduce the WF of cement production, which can be regarded as an energy-saving, consumption-reducing, and economical method. Besides, this study only considers the WF of thermal power generation. However, photovoltaic power generation and wind power generation in Northwest China have begun to be connected to the grid, and power generation through renewable energy will also significantly reduce the power and WF of cement production.

For coarse aggregate production, it can be seen from Figure 4(c) that the power WF contributes the most in coarse aggregate production, followed by DWF and work meal WF. Since all kinds of machinery and equipment need a great deal of energy in coarse aggregate production, in order to reduce the WF of coarse aggregate production, we should first consider improving the production process and reducing energy consumption. For example, renewable energy such as photovoltaic and wind power is used as the primary energy. The second method is to reduce the waste of water when flushing the equipment and washing raw materials so as to reduce the sewage rate. Simultaneously, the DWF and working meal WF of coarse aggregate production in Northwest China are significantly lower than the current research results (Mellor, 2017; Hosseinian & Nezamoleslami, 2018). The result of working meal WF may be caused by the increase in the proportion of mechanized production in the coarse aggregate production industry and the difference in production scheduling between the coarse aggregate production industry and the concrete enterprises that do not need 24-hour production.

Due to the requirements of concrete setting speed and pouring speed, concrete is transported over very short distances. Therefore, concrete production in Northwest China can be regarded as the actual usage of concrete. Water resources are used locally from utilization to release for the water cycle, which does not cause the outflow of water resources. As shown in Figure 5(a), taking 2018 as an example, the total production of concrete in Northwest China is 16,459 × 10⁴m³. According to the calculated LCWF of concrete production, the total water consumption of the concrete output in Northwest China in 2018 is 35,710 × 10⁴m³, of which the direct water consumption is 2,914 × 10⁴m³, and the virtual water resources indirectly provided by the society is about 32,796 × 10⁴m³. The total consumption of water resources in the northwest provinces is shown in Figure 5(b).

We can estimate the amount of cement needed for concrete production by statistical data, which is obviously on the high side in northwest China (see Figure 5(c)) with a production surplus of 15,731 × 10⁴ton. Considering inventory and other uses, the actual production surplus will be lower than this estimate. As shown in Figure 5(c), except for Qinghai, where cement production is equal to the actual demand, all other provinces have higher cement production than the actual need. The cement output of Ningxia is as high as 14.6 times of the actual request. The overproduction of cement has brought about excessive consumption of water resources. According to the calculation in Figure 5(d), the total water consumption of cement production in Northwest China is 31,470 × 10⁴m³. After deducting water consumption for concrete production, the excessive consumption of water resources brought by cement production is 24,035 × 10⁴m³, of which the direct water consumption is 12,308 × 10⁴m³, and the virtual water resources are about 11,727 × 10⁴m³, equivalent to 67% of the water used for concrete production. Among them, Ningxia is the province with the most severe outflow of water resources. In 2018, Ningxia diverted water from the Yellow River up to 582,270 × 10⁴m³. Based on this calculation, the excessive consumption of water resources caused by cement production accounts for about 5.41% of the water transferred from the Yellow River. The cement products will be sold to other areas, forming a virtual water trade that brings an outflow of virtual water. Taking Ningxia as an example, it is equivalent to drawing water from the Yellow River and then subsidizing it to other provinces, which is very detrimental to water resource utilization.

The sensitivity coefficient of the parameters affecting the LCWF of concrete production is calculated according to formula 5. Figure 6 shows that the first three factors that have the most significant influence on the LCWF of concrete production are: Coarse aggregate, Meals, and Cement. To reduce the LCWF of concrete production, we should first consider reducing the WF of coarse aggregate, cement and working meal. Among the WF of coarse aggregate and cement, the electricity WF of coarse aggregate and the DWF of cement are the most important factors affecting the LCWF of concrete production. Therefore, to reduce the WF of coarse aggregate production, we should first consider improving the production process to minimize the VWF of electricity, using renewable energy such as photovoltaic power, wind power, and natural gas primary energy. In order to reduce the WF of meals, it is necessary to adjust the structure of the diet. To reduce the WF of cement production, we should first consider reducing the water consumption used to wash raw materials and clean the equipment, and improving the recycling and reuse of sewage in the cement production process.

The water resources data of Northwest China is obtained from the 2018 China Water Resources Bulletin (see Table 5), and the WSI of Northwest China was calculated according to formula 6.

It can be seen from Table 5 that the shortage of water resources in Northwest China is under Heavy pressure (WSI = 0.67). Qinghai and Shaanxi are under Mild pressure (WSI = 0.05) and Moderate pressure (WSI = 0.5), respectively, while Gansu is under Heavy pressure (WSI = 0.67). It is worth noting that Ningxia (WSI = 9.01) and Xinjiang (WSI = 1.28) are under Extreme pressure. The sustainable development of water resources in Northwest China is under great pressure, and the development of the concrete and cement industry aggravates this situation. The development of the concrete and cement industry is facing great unsustainability and further changes are needed. Take Ningxia as an example: the water intake of Ningxia in 2018 is 4.5 times the total available water resources. It can be found by further calculating the proportion of water used in the concrete and cement industry to industrial water. From the perspective of water use, including virtual water, the ratio of water used by concrete and cement in Northwest China accounts for 13.86% of industrial water use, of which Shaanxi 14.88%; Gansu 9.77%; Ningxia 33.49%; Qinghai 10.48%; Xinjiang 9.54%. It shows the importance of the concrete and cement industry in the economic structure of Northwest China. Ningxia, in particular, under the extreme pressure of water resources, we can see the fragility and unsustainability of Ningxia's water resources. The proportion of water used in concrete and cement production reached 33.49%, exacerbating this extreme vulnerability. Therefore, the optimization and adjustment of industrial structure in extreme water shortage areas will also be an essential and effective means to reduce water resource consumption.

As can be seen from Table 6, the total amount of water resources in Northwest China accounts for 8.8% of the country, but its water consumption accounts for 14.2% of the country. From the perspective of economic sector classification, agriculture accounts for the largest proportion of water use, which is in line with the actual development of the northwest region. From the results of WSI, Northwest China, East China, and North China are the regions with significant pressure on water resources, and the northwest region ranks third. With the rapid development of industry in Northwest China, there is a contradiction in the distribution of water resources among the agriculture, industry, and tertiary industry. Although the proportion of industrial water consumption is still low, the demand is increasing. Therefore, while ensuring economic development and water matching, the development of water-saving industry is the focus of the development. The current situation of cement production essentially brings about the outflow of water resources, so it is necessary to impose certain restrictions on the cement industry depending on water resources.

Product's LCWF reflects the inevitable relationship between water resource consumption, pollution, and environmental impact, facilitating the comparison of products from different regions. Concrete is a core material for the construction industry because it is low-cost, easy to use, and the main components can be produced locally. Therefore, it is impossible to transfer concrete production to areas with mild WSI to reduce water resource consumption, and virtual water trading is a good approach under current conditions. The water scarcity problem is mainly due to the mismatch between the spatial distribution of water resources, economic development and other significant production factors (Zhang & Anadon, 2014). Since concrete production cannot be replaced, more sustainable methods of producing and using concrete must be developed to meet the needs of society while reducing the impact on the environment. Recycling of coarse aggregate and recycled concrete may be the best way to reduce virtual water when using concrete in construction. Recycling coarse aggregate and recycled concrete will reduce the LCWF of concrete production and lessen the WSI. It is the most effective way to maintain the sustainable development of the concrete industry in the future.

This study puts forward a comprehensive evaluation model of LCWF of concrete production based on the influence of energy consumption, direct water consumption, and input factors. By accounting for the LCWF of concrete, cement, and coarse aggregate production, the following conclusions are drawn. (1) Concrete industry's vast water consumption is closely related to VWF, which is the focus of LCWF assessment. (2) The output of cement in Northwest China is obviously on the high side and overproduction, which brings about the excessive consumption of water and the outflow of water resources. (3) The first three factors that have the most significant influence on the LCWF of concrete production are Coarse aggregate, Meals, and Cement. (4) The shortage of water resources in Northwest China is under severe pressure. Qinghai, Shaanxi, Gansu are under mild pressure, moderate pressure and heavy pressure respectively, while Ningxia and Xinjiang are under extreme pressure. The sustainable development of water resources in Northwest China is under high pressure, exacerbated by the growth of the concrete and cement industries.

According to the conclusions of this study, the following policy implications can be obtained. (1) As the concrete industry is still labor-intensive, we can improve production automation through technological upgrading and further reduce the number of personnel in the production process to reduce the WF of meals. (2) In terms of energy input, expanding the utilization range of renewable energy such as photovoltaic power and wind power, and changing coal to gas can effectively reduce the LCWF of concrete production. (3) Raise the threshold of the sand and gravel industry, standardize and rectify the existing sand and gravel enterprises, and eliminate some of the backward production capacity. (4) Recycling coarse aggregate and expanding the use of recycled concrete will reduce the LCWF and WSI of concrete production. (5) With excessive cement production in Northwest China, we can optimize the regional industrial structure to reduce water resource consumption on the basis of virtual water trading.

This study has some innovation in research content and method based on some related literature: (1) based on direct water and virtual water, constructed LCWF model of concrete production in Northwest China; (2) sensitivity analysis is used to analyze the sensitivity of LCWF of concrete production; (3) the water resources consumption of concrete production is compared with that of cement. It found that the WSI of Northwest China is under heavy pressure as cement is in a state of over-production.

Authors declare no conflicts of interest for this research review article.

This research was funded by Natural Science Foundation of Inner Mongolia, grant number 2019LH07002, Research Program of Science and Technology at Universities of Inner Mongolia Autonomous Region, grant number NJSY18131, National Natural Science Foundation of China, grant number 72001167.

All relevant data are included in the paper or its Supplementary Information.