Water Quality Simulation Research Articles

How Do Satellite Precipitation Products Affect Water Quality Simulations? A Comparative Analysis of Rainfall Datasets for River Flow and Riverine Nitrate Load in an Agricultural Watershed

Excessive nitrate loading from agricultural runoff leads to substantial environmental and economic harm, and although hydrological models are used to mitigate these effects, the influence of various satellite precipitation products (SPPs) on nitrate load simulations is often overlooked. This study addresses this research gap by evaluating the impacts of using different satellite precipitation products—ERA5, IMERG, and gridMET—on flow and nitrate load simulations with the Soil and Water Assessment Tool Plus (SWAT+), using the Tar-Pamlico watershed as a case study. Although agricultural activities are higher in the summer, this study found the lowest nitrate load during this season due to reduced runoff. In contrast, the nitrate load was higher in the winter because of increased runoff, highlighting the dominance of water flow in driving riverine nitrate load. This study found that although IMERG predicts the highest annual average flow (120 m3/s in Pamlico Sound), it unexpectedly results in the lowest annual average nitrate load (1750 metric tons/year). In contrast, gridMET estimates significantly higher annual average nitrate loads (3850 metric tons/year). This discrepancy underscores the crucial impact of rainfall datasets on nitrate transport predictions and highlights how the choice of dataset can significantly influence nitrate load simulations.

Chlorine dosage management in drinking water systems: comparing Bayesian optimization to evolutionary algorithms

ABSTRACT To maintain a sufficient chlorine residual in water distribution systems (WDSs), chlorine dosage needs to be optimized. The majority of previous studies that aimed to optimize chlorine dosage in WDSs considered single-species water quality (WQ) models featuring chlorine decay with simple reaction kinetics. Recent efforts have proposed using multi-species water quality (MS-WQ) models to account for chlorine interactions with various chemical and microbiological species, thus providing a comprehensive and accurate evaluation of the WQ within WDSs. Nevertheless, the key challenge of implementing MS-WQ models within optimization frameworks is their high computational cost and poor scalability for larger WDSs. Furthermore, previous optimization studies generally relied on evolutionary algorithms (EAs), which require conducting a significant number of WQ simulations. Bayesian optimization (BO) has been recently suggested as an efficient alternative to EAs for the optimization of computationally expensive functions. This study aims to present a systematic comparison between BO and other widely used EAs for the optimization of MS-WQ in WDSs. A case study featuring a real-life, midsized benchmark WDS was implemented to comprehensively evaluate all three optimization techniques. The results revealed that BO is notably more computationally efficient and less sensitive to changes in the constraints than EAs.

Enhancing Long-Term Water Quality Modeling by Addressing Base Demand, demand Patterns, and Temperature Uncertainty Using Unsupervised Machine Learning Techniques.

Water quality modelling in Water Distribution systems (WDS) is frequently affected by uncertainties in input variables such as base demand and decay constants. When utilizing simulation tools like EPANET, which necessitate exact numerical inputs, these uncertainties can result in inaccurate simulations. This study proposes a novel framework that leverages unsupervised machine learning, specifically a Gaussian Mixture Model (GMMs), to represent and integrate these uncertainties in the simulation. By classifying historical water demand into fuzzy clusters, the framework allows for certain linguistic inputs (e.g., "high" or "low" demand) to be used in water quality simulations. The framework also incorporates representative hourly demand patterns and temperature-dependent chlorine decay constants based on historical data correlations. Validations were conducted on the Anytown network using WNTR-EPANET, comparing simulated chlorine residuals with Validation data from 181 steady-state simulations. The simulation through the framework achieved a Jensen-Shannon Divergence (JSD) of less than 0.008 across all demand clusters, indicating high similarity between predicted and actual probability distributions . In comparison to other simulation scenarios tested, which exhibited increased variability (JSD > 0.18), the proposed framework demonstrated improved accuracy in representing chlorine residual distributions. The methodology is adaptable to other systems, if similar historical datasets containing key variables, such as flow rates and temperature, are provided. While the framework offers a more flexible and accurate approach to handling uncertainties in WDS, its effectiveness is contingent upon the availability of robust historical demand and temperature data for decay constant calibration.

Investigating water quality dynamics in distribution networks with dynamically adaptive connectivity

ABSTRACT Water distribution networks with dynamically adaptive connectivity offer greater operational flexibility. While this strategy has demonstrated improvements in pressure management and network resiliency, further research is needed to better understand its impact on water quality dynamics. This paper investigates the short-term variability of disinfectant residuals in a real-world distribution network operated with dynamic connectivity. We simulate water quality dynamics under two control configurations with pressure control and automatic flushing valve operations. Our simulation results inform the development of flow variability metrics to reveal the relationship between changing hydraulic conditions and increased water quality dynamics. These metrics can (i) improve observability by supporting the placement of additional water quality monitoring locations and (ii) enhance controllability by enabling the formulation of optimization problems that incorporate hydraulic surrogates for modelling water quality. Furthermore, we validate the identified regions of increased water quality dynamics using continuous disinfectant data from a large-scale experimental programme. Our findings emphasize the need for further analytical and experimental research to manage water quality in distribution networks with dynamically adaptive connectivity and hydraulic control.

Evaluation of water quality fluctuation in the tidal reach under the impact of on shore wastewater discharges based on MIKE 21 model in dongguan, China

The text discusses the study of water quality simulation in the lower reaches of the Dongjiang River basin—tidal river section. Utilizing MIKE 21 software, an integrated 2D hydrodynamic-water quality coupled mathematical model was constructed for the lower reaches and estuary area of the Dongjiang River. The model simulated the hydrodynamic characteristics near the wastewater outlet and the impact of pollutant discharge on the water quality in the study area. The simulation results indicate: (1) During wet and dry seasons, the overall flow pattern remains consistent, with the ebb current velocity slightly greater than the flood tide current velocity beside the outlet. The flow velocity is generally greater during wet season compared to dry season. Interestingly, the flow direction during flood and ebb tides is completely opposite and tides have a great impact on the maximum flow velocity in Xialu Connecting channel. (2) Under normal discharge conditions, the increment of COD, NH3-N, and TP in most areas of Zhongtang Waterway, Daoyun Waterway, and Hengchong Waterway is less than 5 mg/L, 0.1 mg/L, and 0.02 mg/L, respectively, indicating a minor impact on water quality; While in the event of an accidental discharge, the concentration increment is in the range of 5–30 mg/L, 0.1–0.25 mg/L, and 0.02–0.05 mg/L, respectively. COD causes the most significant water quality changes and will severely impact the water environment of the Dongjiang River basin. Furthermore, pollutant dispersion is more extensive during dry seasons due to reduced river flow and weaker dilution capacity. (3) Overall, the results underscore the critical importance of accounting for both tidal and seasonal dynamics in simulation of estuarine water quality.

Characterizing and tracing the heavy metals' spatial distribution in karst surface river affected by acid mine drainage

A series of pollution problems have been caused by the drainage of closed coal mines. This study focuses on the XZ River in Guizhou Province, employing a comprehensive approach involving hydrochemical analysis, mathematical statistics, and one-dimensional water quality modeling to investigate the spatial evolution and source apportionment of heavy metals in karst surface rivers affected by AMD. The average pH, EC, and Eh values of groundwater and surface water in the XZ river basin were 6.27, 331 <i>μ</i>s/cm, 230 mV and 3.86, 1671.57 <i>μ</i>s/cm, 412.71 mV, respectively, indicating severe pollution of surface water by AMD. The study reveals that groundwater in the XZ River Basin is predominantly of the HCO<sub>3</sub>-Ca-Mg type, shifting to HCO<sub>3</sub>-SO<sub>4</sub>-Ca-Mg type with the influx of acid mine wastewater. Utilizing statistical methods for source apportionment analysis indicated that downstream concentrations of Fe, Mn, Zn, Cu, As, and Sr most likely originate from tributaries. Further analysis through one-dimensional water quality simulation confirmed that downstream concentrations of Fe, Mn, Zn, Cu, As, and Sr are attributed to tributaries. This suggests that XZ River is affected by pollution from different coal mines, primarily from coal mine M4, followed by coal mine M1.

Simulation of Water Quality in Bung Binh Thien Lake, An Giang Province, Vietnam, Using the Delft3D Model

Simulation of Water Quality in Bung Binh Thien Lake, An Giang Province, Vietnam, Using the Delft3D Model

Evaluation of fish habitat suitability based on stream hydrodynamics and water quality using SWAT and HEC-RAS linked simulation

The objective of this study was to evaluate fish habitat suitability by simulating hydrodynamic and water quality factors using SWAT and HEC-RAS linked simulation considering time-series analysis. A 2.9 km reach of the Bokha stream was selected for the habitat evaluation of Zacco platypus, with hydrodynamic and water quality simulations performed using the SWAT and HEC-RAS linked approach. Based on simulated 10-year data, the aquatic habitat was assessed using the weighted usable area (WUA), and minimum ecological streamflow was proposed from continuous above threshold (CAT) analysis. High water temperature was identified as the most influential habitat indicator, with its impact being particularly pronounced in shallow streamflow areas during hot summer seasons. The time-series analysis identified a 28% threshold of WUA/WUAmax, equivalent to a streamflow of 0.48 m3/s, as the minimum ecological streamflow necessary to mitigate the impact of rising water temperatures. The proposed habitat modeling method, linking watershed-stream models, could serve as a useful tool for ecological stream management.

Simulation and Optimal Scheduling of Water Quality in Urban and Rural Water Supply Systems: A Case Study in the Northwest Arid Region of China

Simulation and Optimal Scheduling of Water Quality in Urban and Rural Water Supply Systems: A Case Study in the Northwest Arid Region of China

A hybrid approach to improvement of watershed water quality modeling by coupling process-based and deep learning models.

Watershed water quality modeling to predict changing water quality is an essential tool for devising effective management strategies within watersheds. Process-based models (PBMs) are typically used to simulate water quality modeling. In watershed modeling utilizing PBMs, it is crucial to effectively reflect the actual watershed conditions by appropriately setting the model parameters. However, parameter calibration and validation are time-consuming processes with inherent uncertainties. Addressing these challenges, this research aims to address various challenges encountered in the calibration and validation processes of PBMs. To achieve this, the development of a hybrid model, combining uncalibrated PBMs with data-driven models (DDMs) such as deep learning algorithms is proposed. This hybrid model is intended to enhance watershed modeling by integrating the strengths of both PBMs and DDMs. The hybrid model is constructed by coupling an uncalibrated Soil and Water Assessment Tool (SWAT) with a Long Short-Term Memory (LSTM). SWAT, a representative PBM, is constructed using geographical information and 5-year observed data from the Yeongsan River Watershed. The output variables of the uncalibrated SWAT, such as streamflow, suspended solids (SS), total nitrogen (TN), and total phosphorus (TP), as well as observed precipitation for the day and previous day, are used as training data for the deep learning model to predict the TP load. For the comparison, the conventional SWAT model is calibrated and validated to predict the TP load. The results revealed that TP load simulated by the hybrid model predicted the observed TP better than that predicted by the calibrated SWAT model. Also, the hybrid model reflects seasonal variations in the TP load, including peak events. Remarkably, when applied to other sub-basins without specific training, the hybrid model consistently outperformed the calibrated SWAT model. In conclusion, application of the SWAT-LSTM hybrid model could be a useful tool for decreasing uncertainties in model calibration and improving the overall predictive performance in watershed modeling. PRACTITIONER POINTS: We aimed to enhance process-based models for watershed water-quality modeling. The Soil and Water Assessment Tool-Long Short-Term Memory hybrid model's predicted and total phosphorus (TP) matched the observed TP. It exhibited superior predictive performance when applied to other sub-basins. The hybrid model will overcome the constraints of conventional modeling. It will also enable more effective and efficient modeling.

Exploring PCSWMM for Large Mixed Land Use Watershed by Establishing Monitoring Sites to Evaluate Stream Water Quality

Extensive hydrologic and water quality modeling within a watershed benefits from long-term flow and nutrient data sets for appropriate model calibration and validation. However, due to a lack of local water quality data, simpler water quality modeling techniques are generally adopted. In this study, the monitoring sites were established at two different locations to collect hydraulic data for the hydraulic calibration and validation of the model. In addition, water quality samples were collected at eight monitoring sites and analyzed in the lab for various parameters for calibration. This includes total suspended solids (TSS), soluble phosphorus, five-day biochemical oxygen demand (BOD5), and dissolved oxygen (DO). The Personal Computer Storm Water Management Model (PCSWMM) 7.6 software was used to simulate all the pollutant loads using event mean concentrations (EMCs). The performance of the model for streamflow calibration at the two USGS gauging stations was satisfactory, with Nash–Sutcliffe Efficiency (NSE) values ranging from 0.51 to 0.54 and coefficients of determination (R2) ranging from 0.71 to 0.72. The model was also validated with the help of historical flow data with NSE values ranging from 0.5 to 0.79, and R2 values ranging from 0.6 to 0.95. The hydraulic calibration also showed acceptable results with reasonable NSE and R2 values. The water quality data recorded at the monitoring stations were then compared with the simulated water quality modeling results. The model reasonably simulated the water quality, which was evaluated through visual inspection using a scatter plot. Our analysis showed that the upstream tributaries, particularly from agricultural areas, were contributing more pollutants than the downstream tributaries. Overall, this study demonstrates that the PCSWMM, which was typically used for modeling urban watersheds, could also be used for modeling larger mixed land use watersheds with reasonable accuracy.

Integration of hydrodynamic and water quality modeling to mitigate the effects of spill pollution into the Nile River, Egypt.

Mitigating spill pollution in the Nile River is crucial to protecting aquatic life, water quality, and public health. Extensive studies focused on the assessment of water quality and hydrodynamics of the Nile River, but there have been relatively few studies that have applied integrated hydrodynamic and water quality modeling approaches to simulate actual accidents in the Nile Fourth Reach. The goal of this study is to develop advanced computational models to simulate accidental spills in the Nile River and track the resulting impacts on water quality. Hydrodynamic and water quality simulations were performed using Delft3D software for 144km of the Nile River, Egypt, from El-Menia to Assuit. Once the hydrodynamic and water quality models were calibrated, two phosphate spill scenarios were modeled under maximum and minimum flow conditions. The spatial distribution of the spill plume along the studied river section was visualized every 12h following the spill occurrence for both scenarios. The results of the research were calibrated and validated against measured field data. In addition, various error and performance indicators were calculated to thoroughly assess the rigor and reliability of the results. The results demonstrated that flow velocity was the main factor influencing the spill plume characteristics and behavior. Initially, advection force plays a significant role after a spill occurs. After that, phosphate becomes mixed and diluted through dispersion. The spill plume took less time to reach downstream areas during the period of maximum flow compared to minimum flow. Additionally, the concentration of phosphate decreased as the water flowed downstream. The spatial distribution of the spill over time can assist water treatment facilities in developing mitigation strategies to address the spill impacts. However, complex Nile River dynamics demand extensive computational power. Therefore, the model was simplified for spill events, using the modeling capabilities to analyze hypothetical spills and contaminant spread in the absence of real data.

Review paper on applications of the HEC-RAS model for flooding, agriculture, and water quality simulation

ABSTRACT The Hydrologic Engineering Center (HEC) of the United States Army Corps of Engineers (USACE) has developed the HEC-River Analysis System (RAS) hydraulic channel flow model as part of its portfolio of hydrologic and hydraulic modeling tools. HEC-RAS is a new version with capabilities to model flooding, modeling for agriculture production computations for water quality, and many more related to hydraulic modeling. The analysis components of the HEC-RAS system include one-dimensional steady flow water surface profile computations, one-dimensional and or two-dimensional unsteady flow simulation, one-dimensional and two-dimensional computations of quasi-unsteady or fully unsteady flow movable boundary sediment transport, and one-dimensional water quality analysis. The main objective of this paper is to review the use of HEC-RAS on flooding, agriculture, and water quality. The three primary components of the data input are plan data, geometry data, and flow data. Three of these scenarios applied to HEC-RAS and were proven by previous reviewers.

MODCEL-MHUS: a comprehensive multilayer hydrodynamic unified simulation for stormwater, sanitary sewer systems, and urban surface.

Numerous countries and regions have embraced implementing a separate sewer system, segregating sanitary and storm sewers into distinct systems. However, the functionality of these systems often needs to improve due to irregular interconnections, resulting in a mixed and malfunctioning system. Sewage collection is crucial for residential sanitation, but untreated collection significantly contributes to environmental degradation. Analyzing the simultaneous operation of both systems becomes vital for effective management. Using mathematical tools for precise and unified diagnosis and prognosis becomes imperative. However, municipal professionals and companies need more tools specifically designed to evaluate these systems in a unified way, mapping all the hydraulic connections observed in practice. This study proposes a unified simulation method for stormwater and sanitary sewer urban systems, addressing real-world scenarios and potential interferences. The primary goal is to develop a simulation method for both systems, considering system interconnections and urban layouts, involving hydrodynamic and water quality simulations. The practical application of this method, the Multilayer Hydrodynamic Simulation Method (MODCEL-MHUS), successfully identifies issues in urban water networks and suggests solutions, making it a valuable tool for urban water management and environmental engineering professionals.

Development of a multi-objective reservoir operation model for water quality-quantity management

This study aims to develop a multi-objective quantitative-qualitative reservoir operation model (MOQQROM) by a simulation-optimization approach. However, the main challenge of these models is their computational complexity. The simulation-optimization method used in this study consists of CE-QUAL-W2 as a hydrodynamic and water quality simulation model and a multi-objective firefly algorithm-k nearest neighbor (MOFA-KNN) as an optimization algorithm which is an efficient algorithm to overcome the computational burden in simulation-optimization approaches by decreasing simulation model calls. MOFA-KNN was expanded for this study, and its performance was evaluated in the MOQQROM. Three objectives were considered in this study, including (1) the sum of the squared mass of total dissolved solids (TDS), (2) the sum of the squared temperature difference between reservoir inflow and outflow as water quality objectives, and (3) the vulnerability index as a water quantity objective. Aidoghmoush reservoir was employed as a case study, and the model was investigated under three scenarios, including the normal, wet, and dry years. Results showed the expanded MOFA-KNN reduced the number of original simulation model calls compared to the total number of simulations in MOQQROM by more than 99%, indicating its efficacy in significantly reducing execution time. The three most desired operating policies for meeting each objective were selected for investigation. Results showed that the operation policy with the best value for the second objective could be chosen as a compromise policy to balance the two conflicting goals of improving quality and supplying the demand in normal and wet scenarios. In terms of contamination mass, this policy was, on average, 16% worse than the first policy and 40% better than the third policy in the normal scenario. In the wet scenario, it was, on average, 55% worse than the first policy and 16% better than the third policy. The outflow temperature of this policy was, on average, only 8.35% different from the inflow temperature in the normal scenario and 0.93% different in the wet scenario. The performance of the developed model is satisfactory.

A real-time simulation model of water quality with the impact of internal pollution for water source reservoir.

The water source reservoirs are the important urban water source in northern China. Although external pollution has been greatly improved, the internal pollutants in reservoirs continue to accumulate with the complex deposition and release processes, resulting in potential risks to water supply safety. To address the aforementioned issue, this paper proposed a simulation model of water quality named ECOlab EU1-WSR to simulate the spatio-temporal changes of water quality under the influence of internal pollution for the water source reservoirs. Based on the analysis of the water quality characteristics and the distribution of benthic vegetation in the reservoir, a three-dimensional hydrodynamic model was established based on MIKE3, the corresponding parameters and the related state variables were set, the ECOlab EU1-WSR model was established by secondly developing the original ECOlab EU1 template, and the real-time dynamic outputs of pollutant content in sediment were added to link the water quality index with sediment nutrition index for better revealing the impact of the internal pollution on the water quality. The performance of the model was evaluated by the case application on the water quality simulation of Daye reservoir and the optimization of the connection project between Daye reservoir and Xueye reservoir in Shandong Province China. The results showed that the model can accurately and simultaneously simulate the pollution in water and sediment by the comparative verification of hydrodynamics, water temperature, and water quality. Moreover, the model can effectively reflect the influence of the accumulation, deposition, and release of internal pollution on water quality by analyzing the correlation between the content of various pollution in water body and those in sediment, such as the total nitrogen and total phosphorus in the water body at the bottom of the water intake, were negatively correlated with the total nitrogen and total phosphorus in the sediments with correlation coefficients of 0.538 and 0.917, respectively. In addition, the optimal water inlet position and water flow rate of the connection project can be optimized and determined by using the model to effectively control water quality. The established model will be a useful tool for the design and management of a reservoir, the interconnection projects, and other water bodies by adaptively recoded.

Method for analyzing urban waterlogging mechanisms based on a 1D-2D water environment dynamic bidirectional coupling model

Urban waterlogging is a significant global issue. To achieve precisely control urban waterlogging and enhance our understanding of its causes, a novel study method was introduced. This method is based on a dynamic bidirectional coupling model that combines 1D-2D hydrodynamic and water quality simulations. The waterlogging phenomenon in densely populated metropolitan areas of Changzhi city, China, was studied. This study focused on investigating the process involved in waterlogging formation, particularly overflow at nodes induced by the design of the topological structure of the pipe network, constraints on the capacity of the underground drainage system, and the surface runoff accumulation. The complex interplay among these elements and their possible influences on waterlogging formation were clarified. The results indicated notable spatial and temporal variation in the waterlogging formation process in densely populated urban areas. Node overflow in the drainage system emerged as the key influencing factor in the waterlogging formation process, accounting for up to 71% of the total water accumulation at the peak time. The peak lag time of waterlogging during events with short return periods was primarily determined by the rainfall peak moment. In contrast, the peak time of waterlogging during events with long return periods was influenced by the rainfall peak moment, drainage capacity and topological structure of the pipe network. Notably, the access of inflow from both upstream and downstream segments of the pipe network drainage system significantly impacted the peak time of waterlogging, with upstream water potentially delaying the peak time substantially. This study not only provides new insights into urban waterlogging mechanisms but also provides practical guidance for optimizing urban drainage systems, urban planning, and disaster risk management.

A Variational Mode Decomposition Analysis and Prediction Simulation of DO in the Water Environment of the Chengdu Area, China

DO is an important index to characterize environmental water quality. The time series fluctuation of DO can be analyzed via frequency band decomposition, which is very valuable for water quality simulations. In this paper, DO in the Chengdu area of China was studied using variational mode decomposition with daily meteorological data and water quality data from 2020 to 2022. After variable decomposition, the DO data were first decomposed into different frequency band signals named IMF1, IMF2, IMF3, IMF4, and IMF5. IMF1 represented the low-frequency signal with long-term trend characteristics of the data. IMF2 to IMF5 represented the high-frequency signal with short-term mutation characteristics of the data. By combining the variable decomposition results with the correlation analysis, it was found that the long-term trend characteristics of DO are affected by the superposition of meteorological factors, hydrological factors, and water pollution factors but have a weak correlation with any single determining factor. The air temperature, water temperature, phosphorus, air pressure, pH value, chemical oxygen demand, and nitrogen were relatively strongly correlated with the long-term trend characteristics of DO. The short-term mutation characteristics of DO were mainly determined using the characteristics of the water body itself, while the influence of the meteorological factors could basically be ignored. The water temperature, pH value, and eutrophication were the biggest influencing factors. Then, a predictive framework combining frequency division with a deep learning model or a machine learning model was constructed to predict DO. The predicted results of GRU, random forest, and XGBoost with and without the framework were compared. It was shown that, after removing the interference factors with correlations less than 0.3, the predicted value of DO was much closer to the actual value. The XGBoost and random forest models with decomposed signals had a high degree of simulation fitting and could be used to predict DO in the Chengdu area. The above research approach can be applied to further explore the prediction of various pollution factors in different areas of China.

Analysis of the water temperature and quality characteristics of the Daheiting Reservoir based on coupled hydrology–hydrodynamic–water quality simulations

To analyse the influence of nonpoint source pollution on the water quality in the Daheiting Reservoir, based on measured hydrological and water quality data, a coupled hydrological model for the Panjiakou-Daheiting interval basin and a two-dimensional vertical hydrodynamic and water quality model for the Daheiting Reservoir were constructed. The water quantity and water quality output results of the upstream hydrological model (Soil and Water Assessment Tool, SWAT) were input into the reservoir hydrodynamic-water quality model (CE-QUAL-W2) as boundary conditions. The annual distribution of pollutants in the reservoir was simulated and analysed. The results showed that nutrients such as nitrate nitrogen and phosphate in the reservoir area were stratified, and there were seasonal differences due to the water temperature, dissolved oxygen and upstream inflow. After lowering upstream pollution emissions, the water quality in the reservoir was improved significantly, and compared to that of reducing point source pollution, the effect of reducing nonpoint source pollution is more obvious.

Multi-step ahead dissolved oxygen concentration prediction based on knowledge guided ensemble learning and explainable artificial intelligence

Accurate water quality prediction is crucial for effective environmental management and decision-making. However, previous studies have solely relied on historical data to simulate water quality, overlooking the potential discrepancies between predicted values and actual observations. Additionally, the opacity of machine learning models has posed challenges to the credibility of their predictions. Hence, considering the excellent nonlinear fitting ability of ensemble tree models, especially the Categorical Boosting (Catboost) model, this study proposes a knowledge-guided Catboost (KGCatboost) model for predicting the dissolved oxygen concentration, one of the vital water quality indicators, in 15 river sections of the Yangtze River Basin in Yunnan Province, China. Furthermore, to enhance the model’s interpretability, we employ the SHapley Additive exPlanations (SHAP) method to analyze the contributions of each input variable within the water body. The results demonstrate that on the test set of each dataset, the mean Nash-Sutcliffe Efficiency (NSE) value of KGCatboost is 0.874, which has improved by 0.34% and 3.07% compared to Catboost and eXtreme Gradient Boosting (Xgboost). In addition, this study reveals that pH has the most significant impact on DO concentrations. Specifically, as the pH increased, the DO concentration increased significantly. A regulatory mechanism has also been developed to alleviate the hazards caused by low DO concentrations. The KGCatboost model can provide valuable guidance for water resource management processes.