Characteristics of temporal changes and influencing factors of carbon dioxide and methane fluxes at the water-gas interface of the Inner Mongolia section of the Yellow River

Abstract

Global warming has evolved into a global environmental concern. Despite the significant impact of rivers on greenhouse gas (GHG) emissions, there remains a lack of wide quantification of emissions at high temporal resolution. This research investigated the characteristics of the CO₂ exchange flux (FCO₂) and CH₄ exchange flux (FCH₄) at the interface of water and gas, as well as the primary determinants that impact these fluxes, at both the seasonal and 24-h scales, in the Inner Mongolia section of the Yellow River (IMYR). FCO₂ was 52.52 ± 78.26 mmol·m⁻²·d⁻¹ in summer, which was higher than autumn (30.81 ± 51.24 mmol·m⁻²·d⁻¹) and spring (-96.09 ± 264.31 mmol·m⁻²·d⁻¹). FCH₄ peaked in summer (928.45 ± 513.31 μmol·m⁻²·d⁻¹) and declined in the fall (326.76 ± 576.31 μmol·m⁻²·d⁻¹). The lowest FCH₄ was recorded in spring (61.75 ± 190.26 μmol·m⁻²·d⁻¹). Wind speed (WS) and organic carbon mineralisation were the primary determinants of FCO₂ at the seasonal scale, whereas temperature and dissolved oxygen (DO) were the primary determinants of FCH₄. Monitoring FCO₂ and FCH₄ at night revealed higher average levels than during the day. FCO₂ was regulated by levels of dissolved inorganic carbon (DIC) and total phosphorus (TP) levels over a 24-h period, whereas FCH₄ was regulated by levels of dissolved oxygen (DO) and total nitrogen (TN). Throughout the majority of the time period, the IMYR river segment emits carbon dioxide (CO₂) and methane (CH₄) to the atmosphere. Due to the fact that emission patterns and driving mechanisms of riverine CO₂ and CH₄ differ across time scales, it is necessary to conduct more precise observations with high spatial and temporal resolution in order to acquire the most reliable data and comprehend more targeted approaches to carbon emission reduction.

Keywords:

• Inner Mongolia section of the yellow river
• CO₂ flux
• CH₄ flux
• temporal variability
• influencing factors

1. Introduction

Over the past few decades, the global average surface and oceanic temperatures have continued to rise, and one of the main contributors to global warming is the continued increase in atmospheric CO₂ and CH₄ concentrations (Borges et al. 2019). In the study of global change, the issue of the global carbon cycle has emerged as a focal point.

The transportation of carbon from terrestrial to marine ecosystems is facilitated by rivers, which constitute a critical link in the global carbon cycle (Cole et al. 2007). A portion of the carbon is deposited in the river channel when it enters the river network system via a complex process involving physical, biogeochemical, and anthropogenic disturbances in its terrestrial forms. The remaining carbon is degraded and ultimately released into the atmosphere in the form of CO₂ or CH₄ (Ran et al. 2017). Riverine GHG equilibrium concentrations undergo oversaturation and are subsequently discharged into the atmosphere (Ström and Christensen 2007). Annually, rivers across the globe are thought to discharge approximately 1.5 Tg·C of CH₄ and 1800 Tg·C of CO₂ into the atmosphere, respectively (Bastviken et al. 2011; Raymond et al. 2013).

Different spatial and temporal dimensions of carbon transport and transformation patterns are influenced by the geographical location, climatic background, land use, lithology, and hydrological conditions of rivers around the world. At present, considerable progress has been made in the calculation and measurement of the mechanisms underlying CO₂ and CH₄ emissions from rivers. However, the absence of high-frequency and high-precision data in the temporal and spatial dimensions complicates the precise quantification of specific emissions (Ran et al. 2015; Reiman and Xu 2018). When nocturnal CO₂ emissions are incorporated, global riverine CO₂ emissions to the atmosphere may be estimated to increase by 0.20 to 0.55 Pg·C·yr⁻¹ (Gómez-Gener et al. 2021). Furthermore, Chen et al. (Chen et al. 2021) demonstrated intricate daily fluctuations in methane fluxes during a 36-h continuous monitoring of methane fluxes in an urban section of the Suzhou Creek in Shanghai, China.

Over time, a succession of studies pertaining to greenhouse gas (GHG) emissions in rivers across the globe were conducted, resulting in the accumulation of a certain volume of data. The Amazon Basin contributes 43% to the global CO₂ emissions from inland waters, with an annual CO₂ emission of 1.39 Pg·C. The annual CH₄ emission from the basin’s major rivers ranges from 0.40 to 0.58 Tg·CH₄·yr⁻¹, accounting for 22–28% of the global riverine methane emissions (Sawakuchi et al. 2014; 2017). Ran et al. (Ran et al. 2016) determined that the Yellow River Basin as a whole emitted approximately 7.9 ± 1.2 Tg·C·yr⁻¹ annually in CO₂. Additionally, it was determined that the diffusive flux of methane in the estuary of the Yangtze River peaked at 27.9 ± 11.4 μmol·m⁻²·h⁻¹ during the rainy season and 36.5 ± 24.4 μmol·m⁻²·h⁻¹ during the dry season (Li et al. 2023).

The Yellow River ranks second-longest in China and fifth-longest in the world (Wu et al. 2018). The Yellow River (IMYR) in Inner Mongolia is a lengthy river section with a wide basin area. In addition to grassland, desert, and other geomorphological features of the region, it travels through a variety of area types, including industrial, agricultural, tourism, and other forms of urban areas. Consequently, the IMYR serves as a paradigmatic river segment distinguished by its geographical attributes. In the past, this section lacked comprehensive data on GHG emissions, and no in-depth analysis of their temporal characteristics had been conducted. Examining the characteristics of the variations of FCO₂ and FCH₄ at the water-gas interface of the Inner Mongolia section of the Yellow River (IMYR) across various time scales and identifying the primary drivers of those variations are the primary research objectives of this paper. The study’s findings can serve as a theoretical foundation for the study of riverine GHG emissions and furnish data support for the investigation of the release characteristics of riverine GHGs at short-term scales.

2. Materials and methods

2.1. Situation in the study area

The Yellow River source is the eastern portion of the Tibetan Plateau; it flows from west to east through nine provinces, including Qinghai, Sichuan, Gansu, Ningxia, Inner Mongolia, Shanxi, Shanxi, Henan, and Shandong (Gao and Wang 2017; Wu et al. 2018). The IMYR commences at Bitter Water Gully, situated at the confluence of Shizuishan City and Wuhai City, and terminates at Toudaoguai, Toketo County, in the eastern direction. Its geographical coordinates are 37°35′∼41°50′ N, 106°10′∼112°50′ E. In addition to its 830 km length, the IMYR is primarily situated in the northernmost region of the Yellow River Basin (Su et al. 2015). Seasonal climates are apparent. Winter is cold and protracted, while spring and autumn receive less precipitation (Zhang et al. 2022). Five main deserts comprise the western portion of Inner Mongolia, and for many years, the Yellow River Basin has been laden with sand and silt carried by the wind, resulting in a high sand content. This section of the river is extremely typical and representative. The climate of the Inner Mongolian Plateau is classified as a semi-arid continental monsoon, characterised by a gradual east-to-west transition from a mesothermal semi-arid zone to an arid zone (Li et al. 2021).

In this study, gas and water samples were gathered at six sampling sites located in the IMYR catchment, spanning from the western to the eastern regions: Wuhai (WH), Linhe (LH), Ulateqianqi (QQ), Baotou (BT), Toketo County (TX), and Laoniuwan (LNW) (Figure 1). Wuhai City is situated in the Yellow River’s higher reaches. In addition to its abundant coal resources, it is an important industrial city in Inner Mongolia. The environmental degradation caused by the frequent mining and handling of coal resources (Liu et al. 2020; Gao et al. 2021), may affect CO₂ and CH₄ emissions in the Wuhai section of the IMYR. Linhe and Wulateqianqi Banner are situated in the agriculturally developed Hetao Plain. The level terrain and fertile soil offer favourable circumstances for agricultural cultivation (Gao et al. 2021). Agricultural production on a large scale also influences the gas fluxes of CO₂ and CH₄. Baotou is a vital economic centre in Inner Mongolia and a significant industrial city in China (Gao et al. 2021). Industrialisation may have an impact on the diffusion of CO₂ and CH₄. The river in Toketo County is subject to a certain level of pollution due to the presence of advanced pharmaceutical factories, chemical plants, power plants, and livestock industries (Dong et al. 2015). Laoniuwan is renowned as a natural resort due to its magnificent scenery, and its enormous influx of visitors may also affect the river’s CO₂ and CH₄ emissions. The carbon emission patterns of urbanised rivers are fraught with uncertainty as a result of human activities; therefore, it is critical to expand the spatial scope of pertinent studies to clarify their carbon cycle mechanisms. The IMYR watershed encompasses numerous cities; therefore, conducting measurements and calculations of CO₂ and CH₄ emission fluxes at the water-air interface in each river segment will furnish not only empirical evidence for pertinent research but also establish a theoretical framework applicable to other urbanised rivers characterised by comparable climatic conditions.

Figure 1. Map of sampling locations delineated with red dots along the Yellow River in Inner Mongolia (IMYR).

2.2. Sample collection

From March 2023 to November 2023, a monthly systematic sampling and monitoring campaign was conducted in six reaches of the IMYR, namely Wuhai (WH), Linhe (LH), Wulateqianqi (QQ), Baotou (BT), Toketo County (TX), and Laoniuwan (LNW), to acquire 54 sets of samples for this study. In Inner Mongolia, March, April, and May are designated spring months. June, July, and August are designated as summer months. September, October, and November are designated as autumn months. Each seasonal data point presented in this article is an average of the corresponding months. During the frigid months of December, January, and February, Inner Mongolia experiences extreme cold, and IMYR experiences an ice closure period that prevents sample collection at the site. This paper conducts a synthesis of findings from pertinent literature and sources in order to evaluate the potential impact of the absence of winter data on the conclusions drawn. Furthermore, in order to examine the diurnal fluctuations of each factor, a total of 48 sets of samples were gathered during 24-h sampling in the TX and BT reaches situated in the ‘Golden Triangle’ development zone in Inner Mongolia in late summer and early autumn of 2022. Late summer and early autumn constitute a special transitional period from the summer’s heat and humidity to the autumn’s cooler and drier conditions, during which both the summer climate and a portion of the autumn weather persist. Situated in the core of the economic development zone, the TX and BT sampling sites are the most representative watersheds of the IMYR’s bias for urban-type streams; thus, they contribute significantly diurnal variation study.

2.3. Analysis of samples

2.3.1. Assessment of water environment parameters

At each sampling event, wind speed (WS) and air temperature (T_A) parameters supplied by the Inner Mongolia Meteorological Service (IMMS) were measured in the field. The measurements of dissolved oxygen (DO), water temperature (T_W), and pH were conducted in the water column at the sampling location utilising a portable water quality monitor (Model DZB-712, INESA-Shanghai). The water samples were gathered in polythene receptacles and stored at 4 °C in the refrigerator for one week prior to the commencement of all subsequent measurements.

Following the filtration of water samples through a 0.3 μm ultrafine glass fibre filter membrane, the membrane was subjected to continuous baking at 50 °C for a duration of 24 h. The determination of the total suspended solids (TSS) involved utilising the disparity in membrane mass prior to and subsequent to the pumping process. A total organic carbon (TOC) analyser was employed to ascertain the concentrations of dissolved inorganic carbon (DIC) and dissolved organic carbon (DOC). The determination of total phosphorus (TP) and nitrogen (TN) was accomplished using alkaline potassium persulfate digestion-ammonium molybdate spectrophotometry and ultraviolet spectrophotometry, respectively, in accordance with the Chinese national standard method (http://www.sac.gov.cn/).

2.3.2. Measurement and calculation of gas fluxes

The gas samples were collected and detected using the floating chamber gas chromatography method (Duchemin et al. 1999; Zhang et al. 2020). Outside-wrapped in tinfoil, the translucent glass sampling box was shielded from the damaging effects of sunlight. The sampling box and 100-mL syringe, which are both equipped with a three-way valve, are linked via a 1.50-metre-long gas conduit. It is important to note that for the gas exchange concentration to surpass the gas chromatograph’s detection limit, the sampling box must be placed at the surface of the water for a duration of 30 min prior to gas collection. For a period of 30 min, gas samples were collected every 5 min. After sequentially numbering and injecting the collected gases into the gas collection bag with a syringe, they were returned to the laboratory for subsequent measurements. The fluxes of CO₂ and CH₄ are computed using Equation (1)(1) $$F=\frac{K\times F_1\times F_2\times V}{A\times F_3\times F_4}$$(1) after the concentrations of these substances are determined using a meteorological chromatograph (6820 Meteorological Chromatograph System) to measure CO₂ and CH₄ and plot a ‘concentration-time’ fitted line graph. (1) $$F=\frac{K\times F_1\times F_2\times V}{A\times F_3\times F_4}$$(1) where F is the exchange flux of GHGs at the water-gas interface, and K is the slope of the fitted straight line (ppm·min⁻¹). A is the surface area of the interface in m², and V is the effective volume of the floatation vessel in m³. F₁ is the unit conversion coefficient from ppm to μg·m⁻³ and is calculated as follows: 1798.45 for CO₂ and 655.47 for CH₄. The conversion coefficient between time intervals is denoted by F₂, which is 1440 for every minute and day and 60 for every minute and hour. The unit conversion factor from μg to mg is denoted as F₃. It is not required for the calculation of the gas flux of CH₄. The mass-to-amount conversion factor for substances is denoted by F₄: 44.0095 for CO₂ and 16.043 for CH₄. The text provides values in mmol·m⁻²·d⁻¹ and mmol·m⁻²·h⁻¹ for the CO₂ exchange flux at the water-gas interface (FCO₂), whereas the CH₄ exchange flux (FCH₄) is expressed in μmol·m⁻²·d⁻¹ and μmol·m⁻²·h⁻¹. In cases where F is positive, rivers are considered ‘sources’ of atmospheric GHG emissions. Rivers function as ‘sinks’ that absorb atmospheric GHGs when F is negative.

2.4. Data processing and statistics

Data processing and statistical analysis were performed using Excel, SPSS 18.0, and Origin 2023 software. Utilising SPSS 18.0, temporal differences in FCO₂ and FCH₄ were examined via one-way analysis of variance (ANOVA). To examine the impact of different water quality parameters on CO₂ and CH₄ fluxes, Pearson’s correlation analysis was applied. Sample site maps were produced utilising the ArcGIS 10.4 GIS (ESRI®) software.

3. Results

3.1. Characteristics of changes in water environment parameters

The characteristics of ten water environment parameters that influence the production and emission of greenhouse gases in rivers were analysed at various time scales in this study. The parameters included in the analysis were wind speed (WS), air temperature (T_A), water temperature (T_W), total suspended solids (TSS), pH, dissolved oxygen (DO), dissolved inorganic carbon (DIC), dissolved organic carbon (DOC), total phosphorus (TP), and total nitrogen (TN).

3.1.1. Parameters of the water environment at seasonal scales

The variations in the IMYR water environment factors during the spring, summer, and autumn of 2023 are shown in Figure 2. The IMYR region experiences variations in wind speed (WS) ranging from 2.33 to 4.00 m/s. The average mean wind speed is 2.87 ± 0.48 m/s, which remains constant across all three seasons (Figure 2a). The measured T_A ranged in temperature from 10.50 °C to 29.00 °C, with an average value of 19.22 ± 6.56 °C (Figure 2b). The range of T_W was 5.67 °C to 27.80 °C, with an average of 15.45 ± 7.42 °C. Summer marked the peak of the T_W, while autumn marked its trough (Figure 2c). The variations in TSS are shown in Figure 2d; they ranged from 22.00 mg/L to 628.67 mg/L, with an average value of 264.51 ± 180.26 mg/L. It indicates IMYR has a high concentration of TSS, with the maximum concentrations occurring during the summer. The water body exhibited a 48.06 ± 0.21 range of 7.40 to 8.31, with an average value of 8.06 ± 0.21. The watershed exhibited an alkaline water quality, devoid of any notable seasonal fluctuations (Figure 2e). The DO level exhibited a range of 11.95 mg/L to 25.03 mg/L, with an average value of 16.34 ± 3.70 mg/L. DO was perpetually present in high concentrations, with its peak value occurring during the summer months of June to August, when it was oversaturated (Figure 2f). The concentration of DOC varied from 4.04 mg/L to 43.08 mg/L, with an average value of 9.23 ± 9.08 mg/L (Figure 2g). The DIC ranged from 3.19 to 4.72 mg/L, with an average value of 3.77 ± 0.47 mg/L, which exhibited a generally low level of concentration (Figure 2h). The TP concentration varied between 15.98 and 53.83 μg/L, with an average value of 31.49 ± 11.31 μg/L. The concentration was greatest in summer and lowest in autumn (Figure 2i). The concentration of TN varied between 5.96 mg/L and 9.07 mg/L, with an average value of 7.25 ± 0.96 mg/L. Compared to the summer and spring, TN levels were highest (Figure 2j).

Figure 2. Characteristics of seasonal variations in parameters of the water environment. where (a)∼(j) are WS (m/s), T_A (°C), T_W (°C), TSS (mg/L), pH, DO (mg/L), DOC (mg/L), DIC (mg/L), TP (μg/L), and TN (mg/L), respectively.

3.1.2. Parameters of the water environment on a 24-h scale

The 24-h variations in water environmental factors at TX and BT are shown in Figure 3. The WS recorded at the TX and BT sampling sites varied between 2.00 and 4.00 m/s, with an average of 2.56 ± 0.57 m/s (Figure 3a). The T_A for the two sampling sites varied from 11.33 °C to 26.33 °C, with an average of 17.16 ± 4.52 °C (Figure 3b). The T_W varied between 15.10 and 21.80 °C, with an average of 18.08 ± 2.07 °C. At both, T_A and T_W were higher during the day than at night (Figure 3c). The TSS exhibited a range of values from 565.67 to 3592.00 mg/L, with an average value of 1982.73 ± 752.36 mg/L. The TSS was higher at the BT sampling site than at the TX sampling site (Figure 3d). The water pH values at both locations exhibited a range of 7.53 to 7.99, with an average of 7.83 ± 0.12, indicating an alkalescent state (Figure 3e). The DO ranged from 13.07 to 24.22 mg/L, with an average of 18.88 ± 3.47 mg/L, indicating that the water was oversaturated with dissolved oxygen (Figure 3f). Low concentrations of DOC and DIC were detected at both the TX and BT sampling sites. The DOC varied between 1.10 and 4.82 mg/L, with an average of 2.48 ± 1.31 mg/L. Similarly, the DIC fluctuated between 2.77 and 3.86 mg/L, with an average of 3.33 ± 0.28 mg/L (Figure 3g, h). The concentration of TP varied between 18.71 and 33.26 μg/L, with an average value of 23.74 ± 3.66 μg/L (Figure 3i). The TN values ranged from 5.78 to 7.54 mg/L, with an average of 6.90 ± 0.56 mg/L (Figure 3j). At the BT sampling site, both TP and TN were higher in the water than at the TX sampling site. Other than temperature, no other parameters exhibited substantial diurnal differences.

Figure 3. Changes in parameters of the water environment over a 24-h period, where (a)∼(j) are WS (m/s), T_A (°C), T_W (°C), TSS (mg/L), pH, DO (mg/L), DOC (mg/L), DIC (mg/L), TP (μg/L), and TN (mg/L), respectively.

3.2. Seasonal and spatial changes of FCO₂ and FCH₄

The seasonal variation of FCO₂ at six sampling sites within the IMYR is shown in Figure 4a. Using one-way ANOVA, the seasonal changes in FCO₂ among the six sampling sites in the IMYR were not statistically significant (p > 0.05). However, discernible variations in the mean FCO₂ were observed across seasons (Figure 4a). In terms of FCO₂ values, the IMYR study area is an overall ‘sink’ for atmospheric CO₂, with values ranging from −681.32 to 195.60 mmol·m⁻²·d⁻¹, with a mean value of −4.25 ± 174.64 mmol·m⁻²·d⁻¹. Notably, the mean FCO₂ value at the TX sampling site in spring was −96.09 ± 264.31 mmol·m⁻²·d⁻¹. During this time, the basin acts as a ‘sink’ for atmospheric CO₂. This is primarily attributable to the exceptionally negative mean FCO₂. FCO₂ has an average value of 52.52 ± 78.26 mmol·m⁻²·d⁻¹ during the summer of 2023; it serves as a ‘source’ of atmospheric CO₂. In autumn, the average value of FCO₂ was 30.81 ± 51.24 mmol·m⁻²·d⁻¹, indicating that it was a contributor to atmospheric CO₂ emissions. The IMYR recorded its maximum mean FCO₂ value during the summer, with autumn following suit. The minimum value was recorded in the spring. LNW has the highest mean FCO₂ across all sampling sites during all three seasons, as determined by a comparison of FCO₂ values. This is followed by QQ, WH, LH, BT, and TX. The overall average FCO₂ value for the entire sampling period is negative at the TX sampling site, which functions as a net ‘sink’ for atmospheric CO₂ due to the fact that the FCO₂ values are negative and substantial in absolute terms during the spring.

Figure 4. Seasonal characteristics of FCH₄ and FCO₂ (a) and (b), respectively. The average values of FCO₂ and FCH₄ during the spring, summer, and autumn seasons are denoted by the dashed blue, black, and red lines in figures (a) and (b), respectively.

The seasonal variation of FCH₄ at six sampling sites in the IMYR is shown in Figure 4b. The seasonal variations in the average FCH₄ values at the six sampling sites did not reach statistical significance (p > 0.05). For three seasons, the FCH₄ at each site ranged from −219.81 to 1865.14 μmol·m⁻²·d⁻¹, with a mean value of 438.99 ± 584.89 μmol·m⁻²·d⁻¹. Consequently, the watershed can be regarded as a net source of atmospheric CH₄ emissions (Figure 4b). The study area exhibited mean FCH₄ values of 61.75 ± 190.26 μmol·m⁻²·d⁻¹ in spring, 928.45 ± 513.31 μmol·m⁻²·d⁻¹ in summer, and 326.76 ± 576.31 μmol·m⁻²·d⁻¹ in autumn. Summer had the highest mean FCH₄ among the three seasons, followed by autumn, with the lowest being spring. Comparing the mean of FCH₄ at the six sampling sites in the IMYR reveals that all six sites are ‘sources’ of atmospheric CH₄, with LH having the highest mean and WH having the lowest mean of FCH₄.

3.3. Characteristics of FCO₂ and FCH₄ changes at the water-gas interface on a 24-h scale

A series of 24-h sampling activities were carried out at the TX sampling location from August 27th to the 28th, 2022, and at the BT sampling site from September 7th to the 8th, 2022. From 10:00 am on one day to 9:00 am on the following, hourly samples of gas and water were collected. At three-hour intervals, the average value of the measured data was analysed.

3.3.1. Characteristics of FCO₂ changes at the water-gas interface on a 24-h scale

During the 24-h period, the FCO₂ at the TX and BT sampling sites varied significantly (Figure 5). The mean FCO₂ at TX peaked at 7.04 ± 2.56 mmol·m⁻²·h⁻¹ between 04:00-07:00, minimum to −8.16 ± 15.86 mmol·m⁻²·h⁻¹ between 19:00-22:00, and averaged 0.63 ± 4.33 mmol·m⁻²·h⁻¹ over the scale of 24 h. This indicates that the TX is a source of CO₂ emissions into the atmosphere during brief time intervals. The average FCO₂ in the BT was generally higher than that in the TX, with minimum values of −6.53 ± 31.58 mmol·m⁻²·h⁻¹ and maximum values of 18.52 ± 20.27 mmol·m⁻²·h⁻¹ occurring between 10:00–13:00 and 13:00–16:00, respectively, and a mean value of 3.99 ± 7.12 mmol·m⁻²·h⁻¹. Short time intervals indicate that the BT is a source of atmospheric CO₂ emissions (Figure 5a). An examination of the mean FCO₂ levels during the day and night at the two sampling sites revealed that the night-time value (2.84 mmol·m⁻²·h⁻¹) was 1.6 times higher than the daytime value (1.79 mmol·m⁻²·h⁻¹) (Figure 5b).

Figure 5. Characterisation of FCO₂ changes over a 24-h period. The temporal variation is shown in figure (a). The comparison of daytime and nocturnal fluxes is illustrated in figure (b), where the horizontal line in the box-and-line diagram represents the median and black dots and triangles represent the mean and outliers, respectively. The box’s top and bottom margins represent 75% and 25% of the dataset, respectively, while the top and bottom whisker lines represent the 90th and 10th percentiles, respectively. The FCO₂ at the TX and BT sampling sites is represented by the red and blue bars and boxes, respectively.

3.3.2. Characteristics of FCH₄ changes at the water-gas interface on a 24-h scale

The changes of FCH₄ in the TX and BT sampling sites during the 24-h scale interval were significant, with FCH₄ in the TX much higher than in the BT (Figure 6). The mean FCH₄ data from the TX showed the maximum value of 98.66 ± 33.42 μmol·m⁻²·h⁻¹ at 10:00–13:00 and the lowest value of −15.72 ± 37.68 μmol·m⁻²·h⁻¹ during 16:00–19:00, with an average value of 40.64 ± 45.96 μmol·m⁻²·h⁻¹ throughout a 24-h period. As a result, the TX is a net source of atmospheric CH₄ emissions over short time periods. The BT reach had the highest value of 40.78 ± 47.46 μmol·m⁻²·h⁻¹ between 10:00-13:00 and the lowest value of −44.45 ± 65.34 μmol·m⁻²·h⁻¹ between 04:00-07:00, with an hourly average of 3.18 ± 22.51 μmol·m⁻²·h⁻¹, indicating that the BT is also a short-time interval source of atmospheric CH₄ emissions (Figure 6a). The overall data indicated that the nightly average FCH₄ (28.37 μmol·m⁻²·h⁻¹) was 1.84 times higher than the daytime average (15.44 μmol·m⁻²·h⁻¹) (Figure 6b).

Figure 6. Characterisation of FCH₄ changes over a 24-h period. Figure (a) shows the variation over time. Figure (b) compares daytime and night-time fluxes, with black dots and triangles representing the mean and outliers, respectively, and a horizontal line representing the median in the box-and-line figure. The box’s top and bottom margins represent 75% and 25% of the dataset, respectively, while the top and bottom whisker lines represent the 90th and 10th percentiles, respectively. The red and blue bars and boxes represent the FCH₄ at the TX and BT sample sites, respectively.

3.4. Relevance over different time scales

3.4.1. Correlation analysis at seasonal scales

The findings of the connection of FCO₂ and FCH₄ with seasonal scale water quality parameters are shown in Figure 7. FCO₂ had a substantial positive association with WS (r = 0.540, p < 0.05), a substantial negative correlation with DOC (r = −0.880, p < 0.001), and no significant correlation with other water quality parameters. These findings showed that wind speed and DOC concentration in water had a considerable seasonal influence on CO₂ fluxes at the water-air interface (Table 1, Figure 7). FCH₄ showed a significant positive association with T_A (r = 0.610, p < 0.01), T_W (r = 0.470, p < 0.05), and DO content (r = 0.640, p < 0.01), but no significant correlation with other parameters. These findings indicate that temperature and DO concentration play major roles in determining CH₄ fluxes at the seasonal-scale water-gas interface (Table 1, Figure 7).

Figure 7. Heat map of the correlation between FCO₂, FCH₄, and water environment parameters at a seasonal scale. The magnitude of the correlation coefficients is denoted by the colour shades; lighter colours signify weaker correlations, while darker colours signify stronger correlations; red colours signify positive correlations, and blue colours signify negative correlations. Significant correlation is denoted by the symbols *, **, and *** at the 0.05, 0.01, and 0.001 levels (two-sided), respectively.

Table 1. Correlation analysis of FCO₂ and FCH₄ with other water environment factors at different time scales.

Time scale	Parameters	WS	TA	TW	TSS	pH	DO	DOC	DIC	TP	TN
seasonal scale	FCO2	0.540*	0.120	0.056	−0.240	0.024	0.300	−0.860***	0.320	−0.240	0.250
	FCH4	0.300	0.610**	0.470*	0.190	0.260	0.640**	−0.071	−0.160	0.280	−0.350
24-h scale	FCO2	0.086	0.034	−0.041	0.069	0.390**	0.007	0.140	0.510***	0.380**	0.180
	FCH4	0.049	−0.020	0.077	−0.290*	−0.008	−0.400**	−0.110	0.150	0.170	0.300*

3.4.2. Correlation analysis on a 24-h scale

FCO₂ exhibited a more significant positive correlation (r = 0.390, p < 0.01) with pH and TP (r = 0.380, p < 0.01) and a highly significant positive correlation (r = 0.510, p < 0.001) with DIC content, when 48 sets of data at the 24-h scale for both TX and BT sampling sites were combined (Table 1, Figure 8). On a 24-h time scale, the results indicated that water DIC, pH, and TP were the primary determinants of CO₂ fluxes at the water-gas interface during short intervals. FCH₄ exhibited a significant negative correlation (r = −0.400, p < 0.01) with DO content, a significant positive correlation (r = 0.300, p < 0.05) with TN content, and a negative correlation (r = −0.290, p < 0.05) with TSS content (Table 1, Figure 8). These results suggest that DO, TN, and TSS are significant determinants of CH₄ flux at the water-gas interface at the TX and BT sampling locations within a brief time interval. Furthermore, TN and TSS content were correlated in a highly significant positive way (r = 0.480, p < 0.001). The correlation between DO and TN content was considerably inverse (r = −0400, p < 0.01) (Figure 8).

Figure 8. Heat map of the correlation between FCO₂, FCH₄, and water environment parameters on a 24-h scale. The magnitude of correlation coefficients is denoted by colour shades; lighter colours signify weaker correlations, while darker colours signify stronger correlations; red colours signify positive correlations, and blue colours signify negative correlations. Significant correlation is denoted by the symbols *, **, and *** at the 0.05, 0.01, and 0.001 levels (two-sided), respectively.

4. Discussion

4.1. Seasonal variation of FCO₂ and FCH₄

The seasonal mean FCO₂ in the IMYR study area peaked in summer, then declined in autumn, and finally reached its minimum in spring (Figure 4a). Meanwhile, the seasonal mean FCH₄ peaked in summer, then declined in autumn, and finally reached its minimum in spring (Figure 4b). The Meteorological Bureau of the Inner Mongolia Autonomous Region reports that precipitation in Inner Mongolia is highest during the winter and summer months, as opposed to the autumn and spring months (nm.cma.gov.cn). An important source of carbon in rivers is soil erosion along riverbanks; precipitation can increase the flux of soil carbon into rivers, which can indirectly contribute to the oversaturation of CO₂ and CH₄ in the river and subsequently increase its greenhouse gas emissions (Campeau et al. 2014). Previous research on CO₂ emissions from streams and rivers in the United States has established a positive correlation between river CO₂ releases and annual precipitation. This correlation can be attributed to the significant quantity of soil CO₂ that was discharged into the river (Butman and Raymond 2011). It is important to note, however, that the dilution effect of precipitation decreases emission fluxes and concentrations of CO₂ and CH₄. And the washout and dilution effects of precipitation can, in some instances, counterbalance one another (Zeng and Masiello 2010). Long-term scale analysis of CO₂ partial pressure changes in the Yellow River Basin reveals that precipitation dilutes CO₂ partial pressure, causing it to be lower during the rainy season compared to the dry season (Ran et al. 2015).

Consequently, it can be hypothesised that the higher FCO₂ and FCH₄ observed in our study during the summer as opposed to autumn and spring may be ascribed to the substantial precipitation that occurs during the summer, leading to a significant discharge of soil carbon into the river. The predominant hydrological process is precipitation flushing. Despite the decrease in spring precipitation, early spring, snow and ice melt events increase stream flow and dilute CO₂ and CH₄ concentrations, leading to the lowest FCO₂ and FCH₄ of the season. An analysis of CH₄ levels in a river adjacent to Cambridge Bay, Nunavut, Canada, revealed a reduction of over one hundredfold in concentrations subsequent to the melting of the ice, in comparison to the time period during which the ice was covered (Manning et al. 2020). Despite the strong dilution effect caused by precipitation, the watershed in our study maintained the highest FCH₄ levels in the summer. This was likely due to the fact that elevated summer temperatures stimulated methanogenic bacteria, which increased their CH₄ production. In contrast to our findings, concentrations of CH₄ emissions in the Amazon Basin were higher during periods of dry water compared to those of abundant water. It was accounted for by the high dilution of CH₄ in abundant water and by the prolonged oxidation time of CH₄ in the deep-water column during periods of high-water levels (Sawakuchi et al. 2014). During the winter, when as the IMYR enters its frozen state, the water-gas interface experiences an obstruction due to the solidified ice. During winter, the subglacial systems of rivers generate a substantial amount of CO₂ and CH₄, which accumulate to the point where they are unable to be promptly released into the atmosphere. As the ice melts, the dissolved gases undergo a certain degree of dilution before being ultimately discharged into the atmosphere. In the absence of direct measurements and assessments of CO₂ and CH₄ fluxes at the water-air interface at the IMYR throughout the winter, annual IMYR GHG emission flux values converted from spring, summer and autumn data may be higher than the actual values.

The GHG dynamics of rivers are significantly influenced by fundamental basic hydrological indicators such as precipitation, streamflow, and flow; however, these processes are intricate and subject to variation. A deeper investigation is warranted into pertinent related studies in order to provide additional clarity regarding the relationship between river hydrological processes and GHG emissions. In most cases, variations in precipitation and river flow occur synchronously. In contrast to an urban river in the UK, where dissolved carbon dioxide concentrations increased during high flows, dilution prevented carbon dioxide saturation in the majority of river samples during high flows in the Mekong (Li et al. 2013; Gu et al. 2021). The discharge pattern of CH₄ from rivers may exhibit an ‘initial flush’ effect, characterised by brief periods of elevated discharge rates preceding the peak flow. However, these hydrological pulses can also substantially dilute nutrients and organic matter in the water column, consequently leading to a reduction in CH₄ production (Dyson et al. 2011). Additionally, greater gas transport rates occur at the water-gas interface as a consequence of more pronounced surface perturbations brought about by faster flow rates.

The aforementioned points validate both the results drawn in this paper and the speculations developed as a result of those findings. The study did not yield statistically significant seasonal differences in mean FCO₂ and FCH₄ (p > 0.05). Further detailed studies are required to quantify the characteristics of these variations; therefore, we shall further enhance the sampling frequency and continuity in future investigations.

4.2. Characteristics of daily changes in FCO₂ and FCH₄

The rates of respiration and photosynthesis of organisms in the river undergo continuous fluctuations in response to changes in sunlight intensity and temperature throughout the course of 24 h. These changes are accompanied by alterations in the river’s CO₂ concentration and partial pressure (Yates et al. 2007). Additionally, the CO₂ exchange at the interface of river water and gas will vary considerably more during the day and night. The TX sampling site exhibited consistent behaviour as a source of atmospheric CO₂ emissions from 10:00 to 13:00 on August 27 and from 22:00 to 10:00 on August 27 to August 28. In contrast, the BT sampling sites were ‘sources’ of atmospheric CO₂ emissions from 13:00 to 01:00 and 04:00 to 10:00 on September 7 to September 8 (Figure 5a). Higher CO₂ emissions occurred during the night at both sampling sites (Figure 5b). In the interim, we determined that the TX sampling site as the ‘source’ of atmospheric CO₂ on 27 August from 10:00 to 13:00, and on 27 and 28 August from 19:00 to 7:00. A source of atmospheric CH₄ emissions was observed at the BT sampling site between 10:00 and 13:00 on September 7 and from 16:00 to 4:00 on September 7–8 (Figure 6a). Night-time CH₄ emission fluxes were higher than daytime CH₄ emission fluxes (Figure 6b). Consequently, the GHG emission fluxes in the watershed will be underestimated if night-time data are disregarded.

Additionally, numerous studies demonstrate that the average FCO₂ is higher when night-time CO₂ emissions from rivers are considered. An investigation into the Lower Mississippi River revealed that the combination of diurnal gas fluxes led to significantly higher CO₂ fluxes than those derived from a solitary daily partial pressure of CO₂ (Reiman and Xu 2018). Based on an analysis of 52 consecutive years of data from 66 rivers worldwide, it has been determined that carbon dioxide emissions, when nocturnal factors are considered, are approximately 27% higher on average compared to emissions when daytime factors are considered (Gómez-Gener et al. 2021). The results of the study regarding water body CH₄ fluxes at various time scales in Shanghai’s rivers also suggest that future researchers will need to increase the sampling frequency and period in order to precisely quantify the CH₄ balance in urban rivers (Chen et al. 2021). Night-time fluxes were 39% higher than daytime fluxes, according to large-scale observations of FCO₂ at the water-gas interface in 34 European rivers by Attermeyer, K. et al. (Attermeyer et al. 2021).

Determining the precise magnitudes of global riverine CO₂ and CH₄ emissions using restricted daytime observations is a difficult endeavour. Consequently, sampling and monitoring at night are required to improve the precision of GHG emission data forecasts.

4.3. Effect of water environmental factors on FCO₂ and FCH₄

At the seasonal scale, FCO₂ and water DOC content exhibited a highly significant negative correlation (r = −0.880, p < 0.001) (Table 1, Figure 7). This suggests that as CO₂ release from the IMYR increased, DOC content decreased. In our study, DOC concentrations varied between 1.25 and 43.08 mg/L, with an average value of 7.97 ± 9.64 mg/L (Figure 2f). One significant source of CO₂ in rivers is the mineralisation of organic carbon. When terrestrial sources of DOC are mineralised in rivers, the organic carbon will be converted to CO₂ and subsequently released into the atmosphere. Consequently, this process will significantly reduce the quantity of DOC along the river system (Winterdahl et al. 2016). This view is highly consistent with our findings; therefore, it can be deduced that organic carbon mineralisation is most likely one of the major sources of CO₂ in the IMYR. FCO₂ was substantially and positively correlated with WS (R = 0.540, p < 0.05), according to the findings of the correlation analysis (Table 1, Figure 7). In general, larger WS cause a disturbance in the tranquil state of the water-gas interface, which facilitates the liberation of free-state CO₂ from the water column. The impact of flow velocity on the gas exchange rate at the water-gas interface was observed to be most pronounced when the WS was below 0.5 m/s and to diminish as the WS increased above 2 m/s (Alin et al. 2011). WS in the IMYR watershed varied between 2.33 and 4.00 m/s; therefore, we have concluded that WS modulates FCO₂ in the region’s rivers.

The T_A, T_W, and DO concentrations in the IMYR catchment were the most representative of seasonal variations in FCH₄, according to our findings. (Table 1, Figure 7). Increased temperatures within aquatic environments stimulate bacterial proliferation and activity in the sediment, thereby augmenting the CH₄ concentration in the river. This augmented CH₄ is subsequently released into the atmosphere upon supersaturation. The methanogenic process must occur in an anaerobic environment, and the concentration and flux of riverine CH₄ are typically inversely proportional to DO. Additionally, elevated DO concentrations enhance the oxidation of CH₄, thereby impeding CH₄ emissions; however, this conclusion is inconsistent with the outcomes that we have obtained. Due to the rapid flow and brief residence time of water in the IMYR river, it is hypothesised that CH₄ was eluted from the water prior to oxidation, notwithstanding the elevated DO concentration. This also indicates that seasonal CH₄ exchange fluxes at the interface of water and gas are probably governed by a complex interplay of various factors.

At the 24-h scale, FCO₂ exhibited positive correlations with pH (r = 0.390, p < 0.01), DIC concentration (r = 0.510, p < 0.001), and TP concentration (r = 0.380, p < 0.01) (Table 1, Figure 8). DIC in rivers is predominantly produced through chemical weathering processes, wherein the principal chemosynthetic states are dissolved CO₂, HCO₃^-, and CO₃^2- (Feely et al. 2004). Upon entering the river network, HCO₃^- generated via the carbonate weathering process contributes to an increase in DIC concentration and river saturation with CO₂, subsequently releasing CO₂ into the atmosphere. The positive correlation between the DIC and the CO₂ exchange flux at the water-gas interface is primarily due to this factor. An extent of pH regulation in rivers can be achieved through the DIC. While this study found a positive correlation between pH and FCO₂ (r = 0.390, p < 0.01), this correlation was not as significant as that of DIC. Nitrogen and phosphorus are essential nutrients in aquatic systems and are critical indicators for assessing the level of water pollution. The growth and reproduction of aquatic organisms are stimulated by eutrophication, which also accelerates the discharge of CO₂ from rivers. Phosphorus is primarily deposited in the Yellow River by sediment that is carried into the river (Hu et al. 2023). In the Inner Mongolia portion of the watershed, the TP concentration was relatively low, ranging from 7.79 to 53.83 μg/L, but it fluctuated considerably. Consequently, the impact on FCO₂ cannot be disregarded.

The FCH₄ value exhibited a statistically significant inverse correlation with the DO (r = −0.400, p < 0.01) and TSS (r = −0.290, p < 0.05) contents, respectively, on a 24-h time scale. Additionally, it demonstrated a significant positive correlation (r = 0.300, p < 0.05) with the TN content (Table 1, Figure 8). Numerous investigations have been undertaken to elucidate the correlation between concentrations of river CH₄ and DO. The predominant source of CH₄ in rivers is sediment, as is customary. In addition to facilitating methanogenesis under anaerobic conditions and increasing CH₄ production, the low DO concentration inhibits the oxidation of CH₄, which results in its supersaturation in rivers and subsequent emission into the atmosphere (Yang et al. 2012; Hu et al. 2018). The reversal of the relationship between FCH₄ and DO at the 24-h and seasonal scales in this study demonstrates the significance of our multi-timescale study. At the 24-h scale, the negative correlation between TSS and FCH₄ in the water column is because a high concentration of TSS can impede gas exchange at the water-gas interface. However, there was a small correlation observed between FCO₂ and TSS or FCH₄, suggesting that the impact of TSS on FCH₄ was insignificant. Water quality can be assessed using TSS, and elevated concentrations of TSS can result in water pollution (Bilotta and Brazier 2008). At the 24-h scale, TSS concentration was positively and substantially correlated with TN (r = 0.480, p < 0.001) (Figure 8); TN concentrations also influenced CH₄ flux. Nitrogen is a critical biogenic element in aquatic ecosystems; eutrophication ensues when the nitrogen concentration in the water is exceeded. According to research conducted in a minor lake in Ulansuhai, eutrophic waters are more conducive to the production and emission of methane (Sun et al. 2021).

4.4. Comparison of CO₂ and CH₄ emission fluxes in global Rivers

In recent years, research into GHGs emitted by rivers has become a world-class favourite. A list of FCO₂ and FCH₄ values in various countries and climatic zones can be found in Table 2. The majority of rivers are net ‘sources’ of CO₂ and CH₄ for the atmosphere. The average value of FCO₂ in the IMYR watershed was −4 ± 175 mmol·m⁻²·d⁻¹, which was generally low. Comparatively, the mean value of FCO₂ in the IMYR region is significantly lower than that of the Kuye River, an additional tributary of the Yellow River (Ran et al. 2016; Chan et al. 2021). Significantly influencing this result is the fact that the TX, BT, LH, and WH sampling sites at various times became atmospheric CO₂ sinks. Particularly during the spring, the TX sampling site absorbed a substantial quantity of atmospheric CO₂. This may have been largely attributable to the complex disturbance of the river ecology that occurred at that time as a result of human activities. Continental climates and FCO₂ levels are comparable between the rivers of the Quebec region of Canada and the Ket River in the Siberian Plain of Russia (Rasilo et al. 2017; Lim et al. 2022). The value of FCO₂ in the Daning River, which is a tributary of the Yangtze River, were comparable to those in the Amazon River stream (Alin et al. 2011; Ni et al. 2019). Given the economically developed nature of the Yangtze River basin, it is plausible that anthropogenically induced inputs of organic matter could stimulate in situ respiration, leading to increased CO₂ production and emission into the atmosphere. Critical to the high FCO₂ levels in tropical watersheds is T_W. The solubility of CO₂ in rivers is diminished at elevated temperatures, thereby promoting the release of CO₂ into the atmosphere (Dinsmore et al. 2013). The wide variation in FCO₂ values observed during long-term monitoring in the Victorian Alps indicates a change in the properties of CO₂ emissions from rivers (Hagedorn and Cartwright 2010).

Table 2. Global CO₂ and CH₄ emission fluxes from Rivers.

River/Basin	Country	Climate	FCO2 (mmol·m^-2·d^-1)	FCH4 (μmol·m^-2·d^-1)	Reference
Inner Mongolia section of the Yellow river	China	Temperate climate	−4 ± 175	439 ± 585	This study
Yellow River	China	Temperate climate	856 ± 409	–	(Ran et al. 2016)
Kuye River	China	Temperate climate	91	350	(Chan et al. 2021)
Daning River	China	Subtropical monsoon climate	330 ± 470	–	(Ni et al. 2019)
Rivers of the Tibetan Plateau	China	Plateau mountain climate	78	1664	(Qu et al. 2017)
Amazon River	Barazil	Tropical climate	19 ∼ 361	10 ∼ 40300	(Alin et al. 2011; Sawakuchi et al. 2014)
Victorian Alps	Australia	Tropical climate	−14 ∼ 8,220	–	(Hagedorn and Cartwright 2010)
Brisbane River	Australia	Subtropical climate	–	20 ∼ 1740	(Musenze et al. 2014)
Ket River	Russia	Continental climate	9 ∼ 114	390 ∼ 1380	(Lim et al. 2022)
Rivers of the Québec region	Canada	Continental climate	66	2000	(Rasilo et al. 2017)

The mean FCH₄ of the IMYR was 439 ± 585 μmol·m⁻²·d⁻¹, which was higher than that of the Kuye River, a tributary of the Yellow River, but lower than that of other rivers (Chan et al. 2021). It ranged from −220 to 1865 μmol·m⁻²·d⁻¹. This may be primarily due to the IMYR’s larger watershed area in comparison to the Kuye, as well as the differences in land use and vegetation traversed by the two rivers. The Amazon FCH₄ exhibits significant variability, with a maximum value that is 4000 times higher than the minimum (Sawakuchi et al. 2014). The FCH₄ of the Amazon River will be influenced in a synergistic manner by its extensive watershed, numerous plant species, substantial tributary count, and vast water flow. There is a suggestion that CH₄ emissions will be impeded and accumulate beneath the ice surface of rivers during the ice-covered period (Bussmann et al. 2021). As a result, the FCH₄ of rivers situated in regions characterised by cold and prolonged winters is comparable is similar to that of rivers in the Ket River in Russia and the Québec region of Canada (Qu et al. 2017; Rasilo et al. 2017; Lim et al. 2022). Therefore, the river’s location’s climate is a significant determinant of FCH₄ and emission patterns. At present, the investigation of CH₄ emissions from rivers across the globe is primarily concentrated in tropical and subtropical areas, with temperate rivers receiving less attention.

5. Conclusions

This study examined the attributes of changes in FCO₂ and FCH₄ at the interface of water and gas, as well as the primary influencing factors across various time scales in the IMYR. The substantial variation in average FCO₂ and FCH₄ levels between spring, summer, and autumn can be attributed to the disparities in precipitation during these three seasons. Furthermore, it is crucial to consider the impact of hydrological parameters on GHG emissions from rivers. Because variation in WS and DOC content also contributed to the seasonal difference in the mean FCO₂ in this study. The primary water environment parameters that influenced the occurrence of seasonal variations in mean FCH₄ were T_W and DO. Based on the positive correlation between DO and FCH₄ (r = 0.680, p < 0.01), it can be deduced that due to the turbulence of the water, CH₄ in the water was discharged from the river and emitted to the atmosphere prior to oxidation, despite the high DO content (r = 0.680, p < 0.01). According to the monitoring data, FCO₂ and FCH₄ were greater during the night compared to the day. Furthermore, the significant determinants and patterns of influence exhibited differences when examining the 24-h scale as opposed to the seasonal scale. FCO₂ was primarily regulated at the diurnal scale by DIC concentration, whereas FCH₄ decreased as dissolved oxygen concentration increased. In comparison to rivers worldwide, the IMYR rivers exhibit reduced values for carbon dioxide and methane exchange fluxes at the water-gas interface. The results of our study demonstrated the temporal variability and distinct emission patterns of riverine GHG emissions. Consequently, in order to obtain an accurate assessment of riverine GHG revenue and expenditure and to develop the most targeted strategies to mitigate carbon emissions, it is imperative to employ monitoring with a high temporal resolution. In the course of our subsequent research, we shall investigate other factors that influence GHG emissions in the IMYR watershed and commence further quantitative analyses of the carbon cycle within this particular area.

Authors’ contributions

Conception and writing of the full text, Aruhan; Field sampling, Aruhan, Xiaoli Wang, Wuyinga, Xia Wu, Liming Qi, Hongxia Bao, Narentuya, Yue Lang, Fang Liu; Sample measurements, Aruhan, Xia Wu, Wuyinga, Hongxia Bao; Statistics and analysis of the data, Aruhan; Supervising the revisions, Xiaoli Wang, Liming Qi; All authors have read and agreed to the published version of the manuscript.

Data availability statement

The data that support the findings of this study are available from the corresponding author, [Wang], upon reasonable request.

Additional information

Funding

This study was financially supported by the National Natural Science Foundation of China under Grant (NO. 42167027), the Natural Science Foundation of Inner Mongolia Autonomous Region under Grant (NO. 2020MS04013), the Collaborative Innovation Centre for Water Environment Safety of Inner Mongolia Autonomous Region under Grant (NO. XTCX003), the Fundamental Reasearch Funds for the Inner Mongolia Normal University (NO.2022JBXC015), Graduate students’ research & Innovation fund of Inner Mongolia Normal University (NO. CXJJS22112) and the Funds for the Project on Construction of Chemistry Key Fundamental Disciplines of the College of Chemistry and Environmental Sciences of Inner Mongolia Normal University (NO. 2023HHYC001).