Based on the random forest model and utilizing the sediment and water process data as well as climatic measurement data from four typical hydrological stations in the Zuli River Basin from 2001 to 2020, this study conducted the simulation and attribution analysis of the dynamic characteristics of sediment and water processes at the basin scale. The results show that the random forest model has good applicability for sediment and water simulation in the Zuli River Basin, with high simulation accuracy and a Nash-Sutcliffe Efficiency (NSE) greater than 0.5 for the vast majority of stations, but the simulation of data peaks is relatively poor. During the period from 2001 to 2020, the climate in the Zuli River Basin generally exhibited a trend towards warming and increased humidity, with an overall increasing trend in runoff and a decreasing trend in annual sediment transport. From 2010 to 2020, the changes in runoff in the Zuli River Basin are primarily influenced by climate change, with a contribution rate of 197.59%; whereas the changes in sediment are mainly reduced due to human activities, with a contribution rate of 162.46%.

Understanding the trend of sediment load variation and future projections in the Ningxia-Inner Mongolia reach of the Yellow River is crucial for its management. This paper analyzed and summarized the characteristics of sediment load changes in the main stream and tributaries of this reach, and identified the causes for the sediment reduction. The characteristics of the sediment load variation in the Ningxia-Inner Mongolia reach (1960-2020) included: a decrease in sediment inflow, with a more pronounced reduction in the mainstem; altered sediment contribution ratio between mainstem and tributaries, showing increased proportional input from tributaries; distinct temporal inflection points in sediment changes for both mainstem and tributaries, with recent reduction phases being largely synchronous and; recent sediment load variations approximately follow a lognormal distribution. The main driving factors influencing these changes of sediment inflow included the detention of the main stream reservoir, water and soil conservation measures in the tributaries, and the variation of rainfall. Considering long-term rainfall stability with cyclical fluctuations and diminishing reservoir sediment retention capacity over time, this study based its analysis on actual sediment yield conditions during 2000-2020, by evaluating reduced sediment interception capacity and newly implemented conservation measures, it is projected that the annual sediment load at Xiaheyan Station on the mainstem, and annual sediment input from interval tributaries will both increase moderately compared to recent levels.

In order to effectively support scientific decision-making for ecological protection in the Yellow River Basin, with digitization, networking and intelligence as the main lines, and based on the characteristics of the ecological environment in the basin, a digital twin platform for ecological protection in the Yellow River Basin was constructed. A model framework including information infrastructure, data base, model and technology simulation, business applications was designed. By classifying and compiling monitoring data on water resources, water ecology and water environment, it established a comprehensive data system for the entire basin. Integrated data collection and monitoring, big data integration and analysis, cloud computing and edge computing integration, digital twin modeling and simulation and other technologies were used to realize the data aggregation and presentation of ecological environment factors such as ecological environment status, natural reserve distribution, ecological flow guarantee and so on. Through platform ecological data visualization, ecological simulation and ecological perspective application practice, the water function zoning and statistical situation of the Yellow River Basin, the evolution of key protected animals and plants in the estuarine delta, the evolution of land use types and the inundation range of the Jingtai Stone Forest caused by the Heishanxia Water Control Project were demonstrated, enhancing the multi-source data aggregation governance and scientific decision-making for ecological protection in the Yellow River Basin.

In order to scientifically evaluate the spatial-temporal characteristics of water-saving levels, identify key influencing factors, reveal the spatial equilibrium patterns in the Yellow River water-receiving area of Henan Province, provide a scientific basis for optimizing water resources allocation and formulate differentiated water-saving policies, a water-saving evaluation index system was built, encompassing five dimensions of comprehensive, agricultural, industrial, domestic, and ecological with 12 quantitative indicators. The TOPSIS method was employed to dynamically assess the water-saving levels of 14 prefecture-level cities in the Yellow River water-receiving areas of Henan Province from 2014 to 2023. Additionally, based on the DEMATEL method and spatial equilibrium analysis, the key influencing factors and their regional differentiation characteristics were systematically analyzed. The results indicate that a) from 2014 to 2023, the average water-saving level in the Yellow River water-receiving areas of Henan Province is classified as Level IV on a 5-level evaluation scale, indicating a relatively low level, with significant spatial heterogeneity in water-saving levels among prefecture-level cities. b) Using the DEMATEL method, five main influencing factors are identified and ranked by their degree of impact: the proportion of planned water users (C2) > water consumption per 10 000 yuan of GDP (C1) > effective utilization coefficient of farmland irrigation water (C5) > water consumption per 10 000 yuan of industrial added value (C7) > proportion of saved water (C3). c) The spatial equilibrium of water-saving in the study area generally remains relatively stable but exhibits a slight declining trend over time, reflecting potential risks of regional imbalance in water-saving development. Therefore, it is essential to strengthen water-saving policy guidance and optimize water resources allocation to ensure sustainable development.

In order to empower the promotion of the new stage of high-quality development of water conservancy, and continue to promote the construction of the digital twin Yellow River, we constructed the digital twin-based Jiaozuo Qinhe “Four Predictions” integration platform, applied it to the Qinhe River Basin in Jiaozuo section, and explored the key technologies and application scenarios for the construction of digital twins of the Qinhe River in Jiaozuo section. The key technologies included the construction of hydraulic professional models and high-fidelity digital twin simulation engines, which utilized a high-performance aggregate hydrological model, a high-efficiency flood evolution model, and a visual disaster assessment model for flood forecasting, preview simulation and disaster assessment, and built an integrated data base for “air, space and ground” through multi-dimensional and multi-temporal and spatial scales. In addition, it formed an integrated platform for the “Four Predictions” of the Jiaozuo Qinhe River, which included dynamic monitoring, forecasting, early warning, rehearsal and preplanning, and described the platform’s infrastructure, functions and application achievements. After the platform is put into application, it will realize the dynamic interaction, real-time integration and simulation of the flood forecast and dispatching results of the Qinhe River in the digital map, can more accurately assess the flood situation, achieve timely early warning and rapid response, gain the greatest initiative for flood disaster prevention, effectively enhance the flood monitoring and early warning capabilities and scientific decision-making capabilities of the Jiaozuo Qinhe River Basin, and at the same time provide construction ideas and case references for the construction of the “Four Predictions” platform in other small and medium-sized river basins.

While the cascade reservoir group is exerting comprehensive benefits, it has significantly altered the carbon and nitrogen cycle paths of rivers, forming a “source-sink duality” of greenhouse gases (GHGs). This paper systematically reviewed the progress and challenges of research on the GHGs source-sink effect of gradient reservoirs in sediment-laden rivers. In terms of monitoring technology, existing technologies such as flux chambers and eddy covariance complement each other, thereby enhancing GHGs flux observation capabilities, furthermore, acoustic surveys and sediment coring techniques have optimized the assessment of carbon burial. However, the precision of multi-source data monitoring and data fusion continues to constrain the accurate evaluation of source-sink effects. Concerning the spatial-temporal distribution patterns of these sources and sinks, GHGs fluxes exhibit distinct longitudinal gradients along the cascade reservoirs and vertical stratification within the water column; specifically, the drawdown zone emerges as a significant hotspot for enhanced emissions due to frequent wet-dry alternation; the cascade reservoirs trigger the accumulation of GHGs by extending hydraulic retention times, altering dissolved oxygen states, and transforming organic matter composition. Regarding the underlying mechanisms influencing these GHGs dynamics, sediment plays a pivotal role: density currents transport and deposit external organic carbon, serving as a crucial substrate, while sediment resuspension disturbances critically affect redox microenvironments at the sediment-water interface. Simultaneously, hydrodynamic conditions directly govern CO2 diffusion efficiency across the air-water interface, influence CH4 bubble transport pathways and dissolution within the water column, and regulate N2O production dynamics via impacts on nitrification and denitrification processes. Notably, in highly sediment-laden rivers like the Yellow River, suspended sediments uniquely promote the proliferation of methanogens directly within the water column, fostering a distinct emission pattern characterized by methanogenesis occurring in the water itself, rather than solely in the sediments. For optimization and regulation, GHGs models have evolved from empirical statistics to mechanism-machine learning fusion, but the existing multi-objective optimization models still lack quantification of water-sediment-GHGs coupling mechanisms. In view of the above issues, it is urgent to build a tracking observation system of GHGs source-sink effects in the middle reaches of the Yellow River in the group of terraced reservoirs, the spatial-temporal variability of GHGs sources and sinks and their key influencing factors, elucidate the biogeochemical process of water-sediment-GHGs interactions, and optimize the water-sediment regulation model of the cascade reservoirs.

In order to investigate the characteristics of microbial community and influencing factors in aquatic bodies during the rainy season in coastal cities, based on Illumina MiSeq sequencing, this paper studied the distribution characteristics of microbial community structure and its correlation with environmental factors in Xixiang River, Shenzhen, and analyzed the effects of environmental factors on microbial community structure. The results show that the nitrogen and phosphorus pollution of Xixiang River is serious, and the water quality of Xixiang River is poor Class V with the risk of eutrophication. There is no significant difference in microbial diversity among sampling sites. At the phylum level, Proteobacteria, Bacteroidetes and Actinobacteria are the dominant species, and Proteobacteria is the first dominant species with a relative abundance of 62.25%-77.28%. At the level of genera, the dominant genera at all sampling points are not completely consistent.  NH+4-N, TDS, NO-3-N, TN and TOC in river water are strongly correlated with the relative abundance of microorganisms, among which the microorganisms are  the most affected by NO-3-N, NH+4-N and TDS. The upstream is affected by the rehydration of reclaimed water from Gushu Sewage Plant, and the bacteria mainly removes organic matter, while the sewage and wastewater in the middle and downstream provide a good living environment for denitrifiers in the water body, resulting in lower nitrate concentration and increased NH+4-N concentration. In the rainy season, the Pearl River estuary is easily supported by the tidal tide of the Lingdingyang Sea area, which is suitable for the growth of Marivita.sp.

With the advancement of water-saving technologies, China's industrial water use efficiency has been continuously improved, and the total industrial water use has shown a downward trend. The LMDI model was applied to analyze the driving factors of changes in industrial water use, and the rebound effect was used to quantitatively analyze the actual offset degree of the effectiveness of technological water-saving measures. An empirical analysis of the long-series panel data of Henan Province and its 18 prefecture-level cities from 2014 to 2022 shows that the improvement of industrial water use efficiency has effectively reduced the industrial water use, with the role of this factor being significantly enhanced. Meanwhile, economic scale and industrial structure have also contributed to a significant reduction in industrial water use, though there are large differences in their driving directions across different years and prefecture-level cities. On the whole, industrial water use in Henan Province is in a state of partial rebound effect. Driven by technological progress, the rebound effect value has decreased year by year, and the decoupling phenomenon between industrial water use and industrial added value has become obvious since 2019. Additionally, industrial water use in all 18 provincial-level cities also experience partial rebound, but there are significant regional differences (the rebound effect is relatively low in Xuchang, Pingdingshan and Kaifeng, while it is higher in Luohe, Anyang and Jiyuan). In order to alleviate the pressure of water resources shortage, reduce water environmental pollution and improve the level of industrial water conservation, controlling the rebound effect of industrial water use is essential. In order to achieve the high-quality development of Henan’s industrial economy, suggestions are put forward, including continuing to promote the improvement of industrial water use efficiency, coordinately advancing the formulation of industrial policies and water-saving policies, and building a regionally differentiated collaborative governance system.

After the operation of reservoir impounding and sediment retention, it will cause the downstream river channel to be eroded and cut down, and even change the river pattern, which will have a certain impact on the safety of water-related projects. In order to master the rules of river sediment erosion downstream the reservoir, based on the water and sediment as well as cross-sectional observation data since the application of the Xiaolangdi Reservoir for sediment control, this paper analyzed the cumulative scouring volume, average scouring depth, the maximum water depth and the morphological changes of longitudinal and cross-sectional sections of the lower reaches of the river, and studied the influence of water and sediment and boundary conditions on scouring efficiency. The results show that from October 1999 to October 2022, the downstream river channel has accumulated a total of 2.276 billion m3 of erosion, showing the characteristics of “more erosion at the upper and less at the lower and uneven distribution along the course”. Among them, 70% occurs above Gaocun and 30% below Gaocun. The longitudinal gradient of the river channel increases in the upper and lower sections, while it decreases in the middle section, making the entire longitudinal profile more concave. The river channel has undergone significant widening and downward cutting. The increase in river width is larger at the top and smaller at the bottom, while the increase in water depth is smaller at the top and larger at the bottom. The riverbed has coarsened, with the median particle size increasing by 6% to 79%, and the scouring efficiency of the river channel has significantly decreased. The channel erosion efficiency below the reservoir is closely related to the average flood flow and the cumulative erosion volume in the preceding period. When the accumulative erosion volume above Huayuankou reach reaches 0.6 billion t and the discharge is 2 000 m3/s, the future erosion efficiency will be reduced to -2.9 kg/m3.

Through applying technologies such as digital twin, artificial intelligence, mobile communication, and the Internet to enable intelligent management of the Yellow River conservation and governance, the construction of digital twin Yellow River can realize the real-time monitoring, precise early warnings, efficient command and dispatch, and scientific decision-making. In order to provide reference for the digital twin Construction in the Yellow River Basin and nationwide, this paper analyzed the “Three Yellow Rivers” construction system of Yellow River Conservancy Commission (YRCC), particularly the progress and achievements in the construction of digital twin Yellow River. Focusing on the improvement of capabilities of flood control project safety and flood defense in the lower reaches of the Yellow River, this paper proposed the key priorities for digital twin flood control project construction: strengthening the integrated “sky-space-earth-water-project” monitoring and sensor network, establishing data aggregation, governance standards, and update mechanisms, improving the data foundation for flood control projects, accelerating the integration of mathematical models with flood control project data, enhancing the “Three Yellow Rivers” linkage and business applications.

This study investigated the spatial-temporal correlation characteristics between groundwater level changes and the intensity of human activities in the western plain of Changji Prefecture from 2000 to 2020. Groundwater depth interpolation was conducted using the ordinary Kriging method, and a Human Activity Intensity (HAI) model was built by integrating multiple indicators such as population density, regional gross domestic product, electricity consumption, and the proportion of construction and cultivated land area. The study employed the centroid migration model and bivariate local spatial autocorrelation analysis to analyze the distribution characteristics and mutual relationships of groundwater depth and human activity intensity in space and time. The results indicate that the overall groundwater level in the study area shows a trend of initial decline follows by an increase. Specifically, the area with groundwater depth greater than 60 meters is accounted for 6.96% in 2000, is increases to 15.51% in 2015, and then decreased to 12.22% iny 2020. The proportion of areas with high human activity intensity increases from 1.03% in 2000 to 3.41% in 2020, and the proportion of areas with medium to high intensity increased from 35.36%  in 2000 to 50.72% in 2020. The correlation analysis reveals a significant positive spatial agglomeration between the spatial-temporal changes of groundwater levels and human activity intensity, especially in the southern region where there is a high concentration of population and economic activities, exhibiting a high-high agglomeration characteristic. Additionally, the centroids of both groundwater depth distribution and human activity intensity are generally migrated from the southeast to the northwest direction.

With the proposal of the major national strategy of ecological protection and high-quality development in the Yellow River Basin, the research on the Yellow River Basin has become increasingly abundant. This paper mainly reviewed the literature on the high-quality development of the Yellow River Basin from the perspectives of its connotation, level measurement and spatial pattern, constraining factors, multi-dimensional studies, development paths and future directions. The results show that a) the connotation of high-quality development in the Yellow River Basin has been continuously enriched, expanding from economic development to the areas such as regional coordination, rural revitalization and cultural-tourism integration. b) By building different evaluation index systems, the measurements of the high-quality development level of the Yellow River Basin indicate that its spatial pattern is closely related to the administrative hierarchy of cities. c) Constraints to high-quality development in the basin include water resources conflicts, difficulties in industrial structure adjustment and transformation, and insufficient support from scientific and technological innovation capabilities. d) Studies on the high-quality development of the Yellow River Basin have been carried out around dimensions such as industrial structure optimization, technological innovation, urban and city cluster development, and cultural-tourism integration. e) The development paths mainly focus on strengthening ecological environment and industrial layout, promoting economic structure transformation and upgrading, and accelerating institutional and mechanism innovation. f) Future research should focus on interdisciplinary studies, multi-dimensional coupling and coordination research, comprehensive integration of multi-source data methods, and refined policy research based on zoning, grading and classification.

The integrated control of small watersheds in the Yellow River Basin is an important part of implementing the national strategy of ecological protection and high-quality development in the Yellow River Basin. In view of the control issues caused by the stepped landform, the characteristics of the climate-vegetation transition zone and the vulnerability of the cascading ecology within the basin, the core contradictions of the comprehensive management of small watersheds were analyzed from three perspectives. In terms of the natural system dimension, it was reflected in the dual pressures of resources and the environment faced by the upstream, midstream and downstream respectively. In terms of the technical system dimension, it was reflected in the lag between the rigid constraints of the resources background and the adaptability of management technologies. In the dimension of the management system, it was reflected in the fragmentation of cross-departmental rights and responsibilities, weak cross-regional collaborative capabilities, and the imbalance between protection and development goals. Building upon this analysis, this study proposes pathways to address the systemic challenges in integrated small watershed management: establishing a resilience enhancement pathway centered on “soil and water conservation, resources efficiency improvement, ecological restoration and climate adaptation”; developing a dynamic intelligent zoning and digital twin decision support system; devising a cross-domain authority-responsibility integration mechanism and a trilateral compensation system for water quality, quantity and sediment; and creating an eco-industrial value-added chain with ecological credit conversion channels. These proposals provide scientific underpinnings for overcoming the systemic challenges in integrated small watershed management within the Yellow River Basin and exploring pathways to improve its quality and efficiency.

In order to reasonably predict the future water demand situation in Henan Province, considering population scale and economic size, this study built a water demand forecasting method based on social characteristics to project the changes in water demand from 2025 to 2035. The results indicate that a decreasing trend in the total water demand for Henan Province in the future, with the total water demands for 2025, 2030, and 2035 being 21.795 billion m³, 21.198 billion m³, and 20.369 billion m³, respectively, representing reductions of 8.11%, 10.63%, and 14.03% compared to 2020. The proportion of water demand across different sectors is ranked as agriculture>domestic>ecological>industrial, with the share of water demand in agriculture and industry showing a decreasing trend, while the share of water demand in domestic use and ecological needs is increasing. The future population change in Henan Province is relatively small, with urban population growth significantly impacting domestic water demand. There is a negative correlation between economic growth and changes in water demand in Henan Province. The increase in water resources utilization efficiency is a primary factor contributing to the reduction in agricultural and industrial water demand, and is also the reason behind the overall decreasing trend in water demand. The rising level of urbanization and improvements in economic and social living standards are important driving forces behind the changes in future water demand in Henan Province.

In order to scientifically evaluate the level of high-quality development in the Yellow River Basin and explore its spatial differentiation and influencing factors, and then provide a breakthrough point for policy selection to promote high-quality coordinated development in the Yellow River Basin. Based on the concept of strong sustainable development, using the panel data of 81 prefecture-level cities in the Yellow River Basin from 2011 to 2020, measuring the total factor productivity by the the super-efficiency mixed measurement model to represent the level of high-quality development. The Dagum Gini coefficient and Moran index were employed to examine the spatial-temporal differentiation characteristics, and the spatial Durbin model was utilized to explore the influencing factors and spatial spillover effects. The results show that a) the high-quality development level of the Yellow River Basin shows a fluctuating upward trend. The phased deterioration of efficiency change effects hinders high-quality development. Technological progress effect is the main driving force for high-quality development, but it has already shown signs of weakening. b)The spatial differences in the level of high-quality development in the Yellow River Basin are significant and are increasing. The spatial non-equilibrium of total factor productivity within and among the upstream, midstream and downstream regions has shown a synchronous strengthening trend, and the inter-regional differences are the primary cause of the spatial differences in high-quality development in the basin. c)There is a significant spatial correlation in the level of high-quality development among the prefecture-level cities in the Yellow River Basin. d)The influencing factors like environmental regulation, technological innovation, opening up and urbanization are the stable driving factors for high-quality development, and each of these driving factors has different spatial spillover effects.

The riparian vegetation buffer is an interlocking aquatic and terrestrial ecosystem, which plays an important role in soil carbon sequestration. Soil samples were collected from five plant communities, such as Imperata cylindrica, Artemisia argyi, Tamarix chinensis, Cynodon dactylon and Alopecurus aequalis in the riparian zone of the lower Yellow River. The soil organic carbon (SOC) of the samples were measured and analyzed. The correlation analysis, principal component analysis, linear regression, and linear regression analysis were used to analyze the correlation between soil SOC content is soil physicochemical properties, and the important factors affecting soil organic carbon content were explored in riparian zones. The results show that the soil SOC content was different in vegetation types, and the range of variation  is 0.25-6.73 g/kg, the mean value is 1.91 g/kg. The soil SOC content of Cynodon dactylon and Alopecurus aequalis at 0-30 cm each soil layer is significantly higher than that of  Imperata cylindrica, Artemisia argyi and Tamarix chinensis. The soil SOC content in the riparian zone is positively correlated with the content of soil water content, soil available potassium, soil available phosphorus, available nitrogen, soil clay and silt, is significantly negatively correlated with soil sand content. It can be seen that soil bulk density, soil moisture content, soil available potassium and available nitrogen contribute 92.3% to the variation of 0-30 cm soil organic carbon content.

In order to curb the severe wind erosion in the Great Bend Region of the Yellow River and reduce aeolian sediment input, this study investigated the aerodynamic mechanisms of sand transport and clarified the underlying principles and scientific issues of sand arrestation. Turbulent velocity distribution formulas incorporating basal roughness and concentration distribution models typical of desert environments were applied to analyze the wind speed, sediment concentration, and sand flux during dust events. The validation against field observations and theoretical calculations reveals a clear inverse relationship between sediment flux and surface roughness, indicating that sand-control engineering effectively weakens both wind erosion intensity and sand production under strong winds. Further calculations using threshold velocity formulas that account for drag effects show that the entrainment threshold of sand particles is positively related to surface roughness. Natural desert surfaces with low roughness exhibit lower threshold velocities, intensifying wind erosion, whereas the installation of sand barriers increases surface roughness, which simultaneously raises the entrainment threshold and reduces near-surface wind velocity. Under this dual effect-enhanced threshold and diminished near-ground wind speed-the activity of aeolian sand is significantly suppressed. This demonstrates that the principle of sand-control engineering lies in increasing surface resistance to restrain sand mobility. In contrast, smooth-surfaced control structures hinder particle deposition atop dunes, thereby suppressing dune migration and expansion. Based on the principle of minimum energy dissipation in natural systems, we have proposed aligning sand-control layouts along the most stable dune ridge lines. Meanwhile, the overall configuration of control structures, optimized under the principle of minimum resistance, not only modifies near-surface wind fields but also exhibits multi-directional adaptability through its crest morphology.

In order to reveal the decoupling effects of carbon emissions in urban agglomerations within the Yellow River Basin (YRB) and identify their key influencing factors, this study aimed to provide a scientific basis for formulating effective carbon reduction policies, and advanced the national strategy of ecological protection and high-quality development in the YRB. In this study, the Tapio decoupling index model and LMDI model were comprehensively adopted to analyze the decoupling status of carbon emissions of urban agglomerations in the YRB and their temporal and spatial changes, identify the key factors affecting the carbon decoupling of urban agglomerations, and propose suggestions for the green and low-carbon development of urban agglomerations. The results show that: a) from 2001 to 2020, the carbon emissions of all urban agglomerations in the YRB are on the rise, and the carbon emissions from high to low are Shandong Peninsula urban agglomerations, Central Plains urban agglomerations, Guanzhong Plain urban agglomerations, Lan-Xi urban agglomerations, Jinzhong urban agglomerations, Hu-Bao-E-Yu urban agglomerations and Ningxia Yanhuang urban agglomerations. b) The type of carbon decoupling in urban agglomerations in the YRB is mainly weak, but the carbon decoupling index generally increases from time to time. Therefore, according to the existing economic development pattern, the economic development of urban agglomerations in the YRB is difficult to get rid of the dependence on fossil energy in the short term, and the road to carbon emission reduction is still a long way to go. c) Energy intensity, per capita GDP, technological level, and the population carrying capacity of the real economy are the main factors influencing the decoupling of carbon emissions in urban agglomerations within the YRB. Improving energy efficiency and technological capability, optimizing industrial structure, and enhancing the population carrying capacity of the real economy can effectively promote an ideal decoupling between economic growth and carbon emissions.

This paper systematically elaborated on the transformation of Yellow River research from the traditional “Three Yellow Rivers” paradigm (Prototype Yellow River, Model Yellow River, Digital Yellow River) to the new paradigm of “Authentic Yellow River”. Over the past 20 years, the construction of the “Three Yellow Rivers” elevated the scientific research and technological development of the Yellow River to a higher level, playing a crucial role in the governance and protection of the Yellow River. With the breakthrough of artificial intelligence technology, this paper proposed the concept of “Authentic Yellow River”, which took the real Yellow River as the foundation, combined artificial intelligence technology to build dynamic scenarios, achieved instant question-and-answer and precise analysis, and broke through the bottleneck of traditional research paradigms. Through the case of the Beijing-Hangzhou Grand Canal crossing the Yellow River project, it demonstrates the application logic of “Authentic Yellow River” in the governance and protection of the Yellow River, pointing out that the “Authentic Yellow River” provides methodological innovation for Yellow River research through the integration of “artificial intelligence + scenario”.

Reservoirs are significant emission sources of methane (CH4), and methanotrophs can mitigate their emissions by oxidizing CH4. In order to comprehend the community characteristics, gene abundance, and assembly processes of methanotrophs in the reservoir sediments of the Huangshui River Basin on the Qinghai-Tibet Plateau, surface sediments from eight reservoirs were collected respectively in May 2023 (dry season) and August 2023 (wet season). Based on real-time fluorescence quantitative PCR and sequencing technology of the functional gene pmoA of methanotrophs, the abundance, community composition, and the diversity of methanotrophs were analyzed, and the assembly processes of methanotrophs were analyzed using the neutral community model. The results show that at the phylum level, the methanotrophs in the Huangshui River Basin are mainly composed of Proteobacteria. At the genus level, they are mainly constituted by Methylocystis and Methylobacter. The α diversity is manifested as being higher in the wet season than in the dry season, while the β diversity is not significant. The abundance of the pmoA gene in the dry season is significantly higher than that in the wet season. The assembly processes of methanotrophs in reservoir sediments are dominated by stochastic processes, among which drift is the most powerful. Temperature, total nitrogen, and pH are the main factors influencing the methanotroph community, and sediment pH is the dominant environmental factor for the abundance of the pmoA gene.

In order to accurately identify the leakage risks of Tanshan Reservoir under complex conditions, this study proposed a multi-method integrated leakage detection framework. It combined various detection techniques, including the simulated flow field method, geophysical methods (such as high-density resistivity, transient electromagnetic, and ground-penetrating radar), and tracer techniques, along with geological surveys for comprehensive analysis. The study results indicate that the leakage mainly occurs through fractures in the dam foundation rock from the reservoir interior to the downstream, eventually emerging at the stone masonry sidewall of the energy dissipation basin. The main leakage channel lies within the completely weathered limestone layer beneath the dam body. The integrated leakage detection approach proposed in this study-revealing the overall leakage distribution through the simulated flow field method, identifying water-rich and high-moisture zones using geophysical methods, and determining leakage sources and paths with tracer techniques-can provide a reference for detecting similar complex leakages and offers technical support for emergency response and reinforcement efforts.

Integrated hydro-wind-photovoltaic development in river basins represents a crucial initiative for implementing low-carbon and green development principles. However, intensified climate change currently leads to increased volatility in renewable energy outputs, heightened uncertainty regarding future power generation potential, and challenges in accurately predicting the eco-environmental benefits of integrated energy bases. These factors pose significant obstacles to the efficient utilization of basin clean energy resources and ecological conservation. Therefore, the Cihaxia integrated water-wind-photovoltaic base in the upper reaches of the Yellow River was taken as the research object. The quantile mapping method, CNN-LSTM-Attention deep learning prediction model and improved theoretical output calculation method of water-wind-photovoltaic were used to screen high-precision future climate model data applicable to the study area. The daily average output of water-wind-photovoltaic power and its multi-time scale complementarity during the planned operation period of the integrated base (2035-2065) were predicted, and the ecological and environmental benefits of the integrated base were estimated. The key findings are: a) From 2035 to 2065, the predicted average annual power generation is 10.13 billion kW·h for hydropower, 1.187 billion kW·h for wind power, and 43.785 billion kW·h for photovoltaic power. Under four SSP scenarios, the average daily inflow to the Cihaxia Hydropower Station is increased by 0.97, 1.74, 1.25, and 1.99 m3/s respectively. The average daily theoretical hydropower output is increased by an average of 2.23 MW, while wind and photovoltaic outputs experience slight average decreases of 0.29 MW and 0.80 MW respectively. b) The annual power generation correlation coefficients are -0.22 for hydro-wind, 0.18 for hydro-photovoltaic, -0.10 for wind-photovoltaic and -0.03 for hydro-wind-photovoltaic combined. The hydro-wind combination exhibits stronger inter-annual complementarity, with the strongest intra-annual complementarity occurring during winter. c) The annual power generation from the integrated base can potentially replace 55.102 billion kW·h of coal-fired power, reducing carbon emissions by 48 million tonnes per year. Furthermore, replacing coal power with wind and photovoltaic generation saves approximately 108 million m3 of water annually. Additionally, the base is projected to reduce the average annual potential evapotranspiration of the underlying surface by 284.56 mm and increase the average annual Net Primary Productivity (NPP) of the ecosystem by 97.04 gC/m2.

In order to effectively integrate raster-based remote sensing monitoring results of soil erosion with plot-based (land use patches) soil and water conservation management, and to enhance the application of dynamic soil erosion monitoring outcomes in conservation practices, this study proposed a method and standard for categorizing land use patches based on practical experience in remote sensing image interpretation. Using land use patches derived from dynamic soil erosion monitoring as the basic evaluation unit and raster-based monitoring results as the foundation, land use patches were classified into five types of non-urgent treatment, desirable treatment, preventive protection, industry management, and non-soil erosion. The categorization criteria were determined based on topographic slope, vegetation cover, proportion of soil erosion area, and whether the responsible entity for soil erosion prevention and control was clearly defined. In Yunyang County, where hydraulic erosion dominates, the application results align with local conditions and meet the requirements of soil and water conservation management.For regions with mixed wind and hydraulic erosion, areas dominated by other erosion types, or locations with special conservation management needs, the categorization methods and standards can be adjusted accordingly during land use patch classification.

In order to explore the impact of green technology innovation in the cities of the Yellow River Basin on carbon emission intensity and its spatial spillover effects, and to provide references for promoting ecological protection and high-quality green low-carbon transformation in the Yellow River Basin, based on the panel data from 79 prefecture-level cities in the Yellow River Basin from 2005 to 2021, the study measured the level of green technology innovation and carbon emission intensity in these cities, applied a spatial Durbin model for regression analysis and robustness testing, and analyzed the impact of green technology innovation on carbon emission intensity along with its spatial spillover effects, regional heterogeneity, and resources endowment heterogeneity. The results indicate that a) the level of green technology innovation in the cities of the Yellow River Basin has steadily improved while carbon emission intensity has significantly decreased during the study period, with both showing significant spatial inequity. b) Urban green technology innovation can effectively reduce carbon emission intensity, and the indirect carbon reduction effect on neighboring cities is greater than the direct carbon reduction effect in the city itself. c) The carbon reduction effects of urban green technology innovation exhibit regional heterogeneity and resources endowment heterogeneity, with downstream cities showing significant spatial spillover effects from green technology innovation while upstream cities have not yet formed significant spatial spillover effects. Resources-based cities have a greater direct effect of green technology innovation, while non-resources-based cities have a larger indirect effect. d) The spatial spillover effect of urban green technology innovation on carbon emission intensity shows an N-shaped variation with increasing geographical distance, with the optimal geographical distance being 600 km. Recommendations: a) Continue to increase support for green technology innovation in the Yellow River Basin, enhancing the training and introduction of talents for green technology innovation to improve innovation capabilities. b) Build a collaborative green technology innovation system across multiple cities and improve the mechanisms for green technology diffusion, etc., to fully leverage the spatial spillover effects of green technology innovation from central cities, downstream cities, and non-resources-based cities, enabling green technology innovation to play a greater role in overall carbon reduction and emission reduction in the basin.

In 2024, the Yellow River Basin faced a severe and complex situation of drought resistance and flood control. The Yellow River Conservancy Commission (YRCC) implemented scientific scheduling, carried out emergency drought-relief regulation of the key reservoir on the main stream of the Yellow River with Xiaolangdi Reservoir at its core, and conducted three flood-season water-sediment regulation processes. This achieved multiple objectives include drought relief, flood control and deposition reduction. Based on the inflow water-sediment data and reservoir operation process, this study analyzed the sediment discharge of the Xiaolangdi Reservoir and the reservoir area’s siltation-erosion changes in 2024. The results are as follows: The annual sediment inflow and outflow volumes are 309.8 million tons and 205.2 million tons respectively, with a sediment discharge ratio of 66.2%. During the three flood seasons, a total of 199.7 million tons of sediment is discharged through water and sediment regulation, accounting for 97.3% of the total sediment discharge for the year. In 2024, the reservoir area is silted up by 116.5 million m³, of which, the main stream and tributaries account for 78.6% and 21.4% of the siltation respectively, with most siltation occurring between elevations of 220 m and 235 m. After the 2024 flood season, the sandbar at the mouth of the Zhenshui River (located 16.39 km from the dam) reaches a height of 5.63 m, resulting in approximately 40.7 million m³ of reservoir capacity in the Zhenshui River being un-utilizable due to the sandbar blockage. It is recommended that in future reservoir operations, measures should be taken to reduce the height of the sandbar at the mouth of the Zhenshui River or slow down its rate of rise, in order to enhance the comprehensive benefits of the reservoir. The scheduling practice in 2024 shows that in order to ensure the water supply for drought resistance downstream, no water and sediment regulation dispatching was carried out before the flood season. However, during the flood season, through scientific optimization, Xiaolangdi Reservoir can still achieve good sediment discharge results.

The source region of the Yellow River (SRYR) is an important water conservation and runoff production area in the Yellow River Basin. In the context of climate change, evaluating the water source conservation and runoff evolution trends in the SRYR is of great practical significance for basin management. This research collected DEM, land use, soil, meteorological, and measured runoff data from the SRYR, and built a high-precision ecological hydrological model (soil and water assessment tool, SWAT) and statistical downscaling model (SDSM) to calculate and analyze the temporal, spatial, and ecosystem scale changes in water conservation capacity and the evolution trend of runoff in the SRYR. The results show that the average water conservation capacity of the SRYR from 1970 to 2024 is 126.4 mm, with a distribution pattern of less in the east and more in the west, and less in the south and north, with significant spatial differences. The contribution rate of grassland to the water conservation of the SRYR is the highest, at 78.2%. The rainfall runoff method shows that the precipitation in the SRYR has increased and decreased by 15%, and the runoff has increased and decreased by 26.82% and 21.15% compared to the baseline level. The simulation results of the ecological hydrological model show that comparing to the current year (2010-2023), the recent (2025-2050) and long-term (2051-2100) changes in runoff in the source area are -2.22% to 6.51% and -0.23% to 10.23% respectively. 

The application of Artificial Intelligence (AI) in water resources management aims to solve complex issues such as water scarcity, water environmental pollution, and water ecological degradation. Its core idea is to utilize the data processing, pattern recognition and predictive analysis capabilities of Artificial Intelligence to build an intelligent solution for water resources management. This paper mainly studied the current development status, key technologies and practical application effects of AI technology in the field of water resources, and explored its application potential in the three core fields of hydrological prediction and analysis, water quality monitoring and assessment, and water resources management and optimization. Based on comprehensive research, the main algorithms and typical applications of AI in water resources management were analyzed, such as the application of models like Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU) and Transformer in hydrological prediction, the application of Convolutional Neural Network (CNN) in water quality monitoring and the application of algorithms like DDPG and DQN in reservoir regulation. The practical applications of AI technology in scenarios such as water level prediction, flood forecasting, inversion of water quality parameters and intelligent irrigation were discussed. The future development direction of AI application in the field of water resources management was prospected, emphasizing the need to enhance the integration of physical mechanisms and data-driven methods, improve model transparency, and provide technical support for building a smart water resources management system.

In actual application on the Yellow River, the change law of sediment scouring and silting is greatly significant for the sediment and flood control, but the law in short time scale has not received enough attention. In this paper, a roll prediction model for short-term scour and silt was established at Bayangaole and other stations on the Yellow River. Sense the error from the model was obviously smaller than that of the directly borrowed section, it was verified that the sediment content, velocity and water depth were correlated with the short-term changes of the section area, and the model parameters were continuous across the years. Further study and analysis of variables had provided theoretical support for optimizing in production. In addition, this paper deeply discussed the factors that affect the quality of the model, and put forward the technical direction for improving the model.

Xiaolangdi Reservoir is an important strategic project for flood control and security in the lower reaches of the Yellow River. The reservoir has been put into operation more than 20 years, great changes have taken place in the water and sediment situation of the Yellow River,the flood-carrying capacity of the lower reaches’ riverbed, the sediment deposition in the reservoir, and the flood control situation in the floodplain areas. In particular, the strategy of ecological protection and high-quality development of the Yellow River Basin and the people-oriented development concept have been promoted. New requirements are put forward for flood control security in the beach area of the lower Yellow River. At present, Xiaolangdi Reservoir is in the late stage of sediment containment, and the flood control storage capacity of the reservoir is larger than the design, so it is possible to use the surplus flood control storage capacity for flood control and beach protection. The research shows that under the premise of ensuring the flood control safety of Xiaolangdi reservoir, the upper level of flood control and beach protection operation of the reservoir is 258.8 m in the pre-flood season and 259.5 m in the post-flood season. Based on the factors such as flood transport capacity, inundation effect of beach area and the change of river flat flow, the downstream protection flow is 4 500-5 000 m3/s. Through the calculation of different combinations of small and medium flood regulation, the Xiaolangdi Reservoir can ensure that the downstream of a major flood will not overflood every 5 to 10 years in the pre-flood season, and the downstream flood will not overflood every 5 years in the post-flood season. According to the measured sediment level of the Yellow River since 2000, the mathematical model predicts that Xiaolangdi Reservoir can maintain the current beach protection standard for about 10 years. In view of the lagging safety construction of the lower Yellow River beach area, it is suggested to deepen and refine the application plan of Xiaolangdi Reservoir flood control and beach protection, and clarify the application time, water level and operation mode of reservoir beach protection, so as to provide technical support for flood control decision-making in the middle and lower reaches.

Water scarcity constitutes a critical constraint on the sustainable development of the Yellow River Basin. To provide a reference for formulating water resources security strategies in the Yellow River basin, this study systematically analyzes the fundamental characteristics, administrative situation and the historical utilization of water resources in the basin, and projects future supply-demand trends. The results indicate that: 1) The Yellow River Basin suffers from a suboptimal natural endowment of water resources, primarily characterized by severe overall shortage, uneven spatiotemporal distribution, high inter-annual and intra-annual variability, prolonged periods of consecutive dry years, and a significant declining trend in natural runoff. 2) With the continuous improvement of water resource management systems, the total water use within the Yellow River water supply zone remained generally stable from 1989 to 2023, yet notable structural changes occurred, evidenced by a substantial decrease in the proportion of agricultural water use. 3) Water use in the basin is constrained by both resource availability and management policies. Although water use efficiency has reached a relatively advanced level domestically, a significant supply-demand gap persists. The intensity of water resources utilization has exceeded the basin's carrying capacity, leading to prominent issues such as insufficient environmental flow in the mainstream and tributaries, river channel sedimentation and shrinkage, and severe groundwater over-exploitation. 4) To support national strategic goals such as ecological conservation and high-quality development within the Yellow River Basin, long-term stability and security, as well as food and energy security, water demand in the basin is projected to continue growing. It is thus imperative to accelerate the construction of inter-basin water transfer projects to secure water security for the basin through external water sources.

In order to explore the water-saving efficiency of floating balls on large water bodies in the arid regions of northwest China, this study was based on a water surface shading experiment conducted in Kunyu City of Hetian Prefecture. The experiment utilized two Φ20 standard evaporation pans, six 20 m² evaporation ponds, and one 3 000 m² evaporation pond as evaporation devices, with the covering material being HDPE black weighted floating balls with a diameter of 10 cm. The study mainly analyzed the differences in annual evaporation between the Φ20 standard evaporation pans and the 20 m² evaporation ponds and calculated the evaporation conversion coefficient between the two. Additionally, by analyzing the stability of the floating balls in the 3 000 m² evaporation pond under different wind speed conditions, the relationship between floating ball coverage area and water-saving rate was explored. The study primarily reveals that the main cause of the evaporation differences between the Φ20 standard evaporation pans and the 20 m² evaporation ponds lies in the variation in internal water temperature and the heat supply differences caused by the wall effects. The evaporation conversion coefficients are 0.587 during the non-freezing period and 0.282 during the freezing period. When the wind speed exceeds 4 m/s, the blank water area under the floating ball coverage shows a nonlinear positive correlation with wind speed, expressed by the equation y=0.070 68x1.913 69, the floating ball coverage rate and water-saving rate decrease as wind speed increases.

In order to understand the current situation of extreme precipitation events in Henan Province, scientifically recognize and master the spatial-temporal evolution characteristics of extreme precipitation events, and provide basic information support for flood and drought disaster prevention and extreme precipitation event early warning, based on the precipitation data of 26 meteorological stations near Henan Province from 1973 to 2023, an extreme precipitation threshold standard was established to screen extreme precipitation events. Meanwhile, in combination with the recommendations of the World Meteorological Organization (WMO), indicators such as extreme precipitation and continuous wet period CWD were selected. The spatial-temporal distribution of extreme precipitation indicators was statistically analyzed by using M-K trend analysis, and the period was studied by using the wavelet function. The results show that in terms of spatial distribution, the extreme precipitation threshold and eight extreme precipitation indicators in Henan Province gradually increase from northwest to southeast, but the number of extreme precipitation days and the intensity of extreme precipitation do not have obvious spatial distribution characteristics. During the time series changes, among the 8 extreme precipitation indicators, except for the continuous wet period CWD, the number of heavy rain days R25, and the number of heavy rain days R50, the remaining 5 indicators all show a decreasing trend, but they show an increasing trend after 2000. The extreme precipitation in Henan Province has periodic scale characteristics of 26-31, 16-20, and 8-11 years during the study period.

Digital intelligence plays a crucial role in promoting high-quality development in the Yellow River Basin. This study measured the digital intelligence development levels across nine provinces and regions in the Yellow River Basin from 2016 to 2023 based on text analysis. This research investigated the spatial-temporal evolution characteristics of digital intelligence development by employing methodologies such as time series analysis, Dagum Gini coefficient, local Moran’s index, spatial-temporal transition analysis and natural breaks classification. The findings indicate a steady increase in digital intelligence levels within the Yellow River Basin. However, areas with high growth rates are few, and there is significant spatial heterogeneity influenced by resources endowments, leading to notable regional development disparities that tend toward polarization. The downstream eastern regions exhibit higher development levels compared to the central and western basin areas. Consequently, this paper proposes to strengthen the construction of basic communication facilities, improve supporting policies, enhance regional cooperation, and establish data and technology sharing platforms to leverage policy “multiplier effects”, coordinate regional development and accelerate digital intelligence building.

The upper reaches of the Yellow River contribute 58% of the water volume and 12% of the sediment to the entire basin. Due to climate change and human activities, there have been significant changes in the upstream water and sediment characteristics. Based on hydrological, meteorological and other observation data, and using methods such as hydrological statistics, double cumulative value correlation curves, and Kendall rank correlation, this article analyzes the evolution law of water and sediment in the upper reaches of the Yellow River and its influencing factors. The results show that, during the period of 1956-2022, due to the increase in precipitation and temperature, the natural runoff of the main stream above Tangnaihai and the Huangshui River Basin increased by 0.005-0.011 billion m3 per year, while the hydrological stations in the Lanzhou to Toudaoguai section of the main stream and the Daxia River and Tao River basins were affected by the decrease in precipitation and changes in underlying surfaces, resulting in a decrease of 0.003-0.047 billion m3per year in natural runoff. The sediment discharge of the main and tributary control stations has shown a decreasing trend over the years. The main stream above Tangnaihai has decreased by 20 000 tons per year, the hydrological station between Lanzhou and Toudaoguai has decreased by 1.58-2.15 million tons per year, and each tributary has decreased by 50 000-470 000 tons per year. The distribution of runoff and sediment in various rivers is uneven throughout the year. The runoff is mainly concentrated from May to October, accounting for 67.8% to 81.4% of the annual total. Sediments are mainly concentrated from June to September, accounting for 63.2% to 91.4% of the annual total. The impact of reservoir and hydropower stations on water and sediment changes is significant. Taking the Longyangxia and Liujiaxia hydropower station as an example, the joint operation of the two reservoirs after 1986 resulted in a 13.6% reduction in annual runoff and a 75.3% reduction in annual sediment discharge at Lanzhou station.

Water scarcity is the core bottleneck restricting socio-economic high-quality development and ecological protection in the Yellow River Basin. To address the water resource challenges in the basin and support decision-making for ecological protection and high-quality development, this study systematically analyzes the multi-dimensional contradictions among water, food, energy, and ecology within the basin, revealing that the current water supply and demand are in a state of “tight balance”, and the gap is expected to continue widening in the future. However, existing measures-such as water conservation, engineering storage and regulation, and regional water transfer-are constrained by the total available water resources and limited coverage, making it difficult to fundamentally alleviate the structural water shortage. Therefore, guided by the water governance principle of “prioritizing water conservation, balancing spatial distribution, adopting systematic approaches and leveraging the roles of both government and market”, water safety guarantee a four-pronged integrated approach of “water saving, water diversion, water allocation and water management” is proposed: a) Deepening water conservation across all sectors to promote a transformative shift from “efficiency enhancement” to “an efficiency-to-effectiveness revolution”. b) Advancing major strategic projects, particularly accelerating the assessment and construction of the western route of the South-to-North Water Diversion Project. c) Incorporating water resources carrying capacity as a binding constraint into territorial spatial planning, industrial layout and urban development frameworks to achieve broader spatial equilibrium. d) Leveraging the synergistic roles of government and market in water governance to optimize water allocation and ensure sustainable utilization.

The upper reaches of the Yellow River (UPYR) serve as the primary source area for the basin’s runoff. It is essential to quantify the impacts of climate change and anthropogenic activities on the variation patterns of runoff in this region to enhance effective water resource management and support informed decision-making within the Yellow River Basin. In this study, we developed the SWAT hydrological model for the upper reaches of the Yellow River, calibrating and validating it from the base period of 1964 to 1980. We systematically evaluated the effects of climate change and human activities （including water usage, reservoir regulation and land use） on runoff changes from 1981 to 2020. The findings indicate that: a) The basin is currently undergoing a significant increase in both precipitation and temperature, with precipitation levels rising at a rate of 8.11 mm per decade and a corresponding warming rate of 0.35 ℃ per decade. It is important to note that there is spatial heterogeneity in the intensity of the impacts of climate change. In the source area, the contribution rate of climate factors at the Jimai Station is 94%. In contrast, at the Toudaoguai Station, the contribution rate decreases to 21%. b) The influence of human activities on runoff attenuation exhibits spatial gradient characteristics, with variations ranging from 60% to 80% between Lanzhou and Toudaoguai section. Water abstraction and consumption are identified as the principal contributing factors, accounting for approximately 40% to 45% of this phenomenon. The establishment and operation of the reservoir have resulted in a temporal redistribution of runoff, leading to a decrease of 18.11%±6.27% during the flood season and an increase of 12.33%±4.2% during the non-flood season. c) Between 1964 and 2020, the annual runoff in the upstream region of the Yellow River experienced a decline of 148 million cubic meters per decade. An analysis of the factors contributing to runoff reveals that precipitation recharge is the primary determinant, accounting for approximately 80%±11.33%. This is followed by contributions from snow and ice melt, thawing of frozen soil and groundwater recharge. The findings of this study elucidate the nonlinear superposition effects of climate change and anthropogenic activities in the upper reaches of the Yellow River, thereby providing theoretical support for understanding the variations in upstream runoff in the context of climate change.

Yulin in Northern Shaanxi is rich in high-quality coal resources. However, large-scale coal mining has caused severe damage to regional water resources. Taking Yulin coal mine as the study area, this study built a Relative Risk Model (RRM) to quantify the stress effects of risk sources including industrial development, mining-induced collapse and groundwater drainage on four types of receptors of social water stress, underlying surface, vegetation ecology and water resources. By dividing 84 watershed risk units, the study used log-normalization of risk values and natural break classification to zone the study area into five risk levels, proposing differentiated protection strategies. The results show that the mining-induced collapse is the primary risk source (contributing 52% of the risk value). The underlying surface and vegetation ecology are sensitive receptors (accounting for 35% and 28% of the risk value respectively). Level I risk areas require prioritized surface restoration and comprehensive utilization of drained water, while Level V areas need to be given attention, and buffer zones should be delineated.

Faced with the problem of insufficient accuracy in traditional water quality parameters prediction methods when dealing with complex nonlinear water quality parameters change, a CNN-Transformer-based parallel prediction model for water quality parameters in the Yellow River was proposed, based on the periodic and nonlinear characteristics of water quality parameters change. This model predicted dissolved oxygen, permanganate index, ammonia nitrogen, and total phosphorus at the Qilipu monitoring section of the Yellow River from 2020 to 2025. The model inputed monitoring data in parallel into the CNN (Convolutional Neural Network) module and the Transformer module, respectively extracting local detail features and global dynamic features, and used a fully connected layer to map the fused features to the prediction results. Comparing the prediction performance of the CNN-Transformer model with RNN (Recurrent Neural Network), CNN, LSTM (Long Short-Term Memory), and Transformer models, the results show that, compared with the other four models, the CNN-Transformer model reduces MSE by 3.93%~10.96%, RMSE by 5.82%~9.33%, MAE by 12.44%~14.48%, and improves R² by 6.56%~26.65%, demonstrating the most outstanding performance.