Applicability of water quality models around the world-a review

Data-driven evolution of water quality models: An in-depth investigation of innovative outlier detection approaches-A case study of Irish Water Quality Index (IEWQI) model

Recently, there has been a significant advancement in the water quality index (WQI) models utilizing data-driven approaches, especially those integrating machine learning and artificial intelligence (ML/AI) technology. Although, several recent studies have revealed that the data-driven model has produced inconsistent results due to the data outliers, which significantly impact model reliability and accuracy. The present study was carried out to assess the impact of data outliers on a recently developed Irish Water Quality Index (IEWQI) model, which relies on data-driven techniques. To the author's best knowledge, there has been no systematic framework for evaluating the influence of data outliers on such models. For the purposes of assessing the outlier impact of the data outliers on the water quality (WQ) model, this was the first initiative in research to introduce a comprehensive approach that combines machine learning with advanced statistical techniques. The proposed framework was implemented in Cork Harbour, Ireland, to evaluate the IEWQI model's sensitivity to outliers in input indicators to assess the water quality. In order to detect the data outlier, the study utilized two widely used ML techniques, including Isolation Forest (IF) and Kernel Density Estimation (KDE) within the dataset, for predicting WQ with and without these outliers. For validating the ML results, the study used five commonly used statistical measures. The performance metric (R2) indicates that the model performance improved slightly (R2 increased from 0.92 to 0.95) in predicting WQ after removing the data outlier from the input. But the IEWQI scores revealed that there were no statistically significant differences among the actual values, predictions with outliers, and predictions without outliers, with a 95 % confidence interval at p < 0.05. The results of model uncertainty also revealed that the model contributed <1 % uncertainty to the final assessment results for using both datasets (with and without outliers). In addition, all statistical measures indicated that the ML techniques provided reliable results that can be utilized for detecting outliers and their impacts on the IEWQI model. The findings of the research reveal that although the data outliers had no significant impact on the IEWQI model architecture, they had moderate impacts on the rating schemes' of the model. This finding indicated that detecting the data outliers could improve the accuracy of the IEWQI model in rating WQ as well as be helpful in mitigating the model eclipsing problem. In addition, the results of the research provide evidence of how the data outliers influenced the data-driven model in predicting WQ and reliability, particularly since the study confirmed that the IEWQI model's could be effective for accurately rating WQ despite the presence of the data outliers in the input. It could occur due to the spatio-temporal variability inherent in WQ indicators. However, the research assesses the influence of data input outliers on the IEWQI model and underscores important areas for future investigation. These areas include expanding temporal analysis using multi-year data, examining spatial outlier patterns, and evaluating detection methods. Moreover, it is essential to explore the real-world impacts of revised rating categories, involve stakeholders in outlier management, and fine-tune model parameters. Analysing model performance across varying temporal and spatial resolutions and incorporating additional environmental data can significantly enhance the accuracy of WQ assessment. Consequently, this study offers valuable insights to strengthen the IEWQI model's robustness and provides avenues for enhancing its utility in broader WQ assessment applications. Moreover, the study successfully adopted the framework for evaluating how data input outliers affect the data-driven model, such as the IEWQI model. The current study has been carried out in Cork Harbour for only a single year of WQ data. The framework should be tested across various domains for evaluating the response of the IEWQI model's in terms of the spatio-temporal resolution of the domain. Nevertheless, the study recommended that future research should be conducted to adjust or revise the IEWQI model's rating schemes and investigate the practical effects of data outliers on updated rating categories. However, the study provides potential recommendations for enhancing the IEWQI model's adaptability and reveals its effectiveness in expanding its applicability in more general WQ assessment scenarios.

Bottled water safety evaluation: A comprehensive health risk assessment of oral exposure to heavy metals through deterministic and probabilistic approaches by Monte Carlo simulation

The consumption of bottled water has witnessed substantial global expansion in recent times. This study aimed to quantitatively evaluate the concentrations of eight heavy metals (As, Ba, Cd, Cr, Mn, Mo, Ni, and Zn) in 71 high-consumption bottled water brands in Iran. Non-carcinogenic and carcinogenic risk assessments were conducted using both deterministic and probabilistic approaches. Point estimation utilizing the Hazard Quotient (HQ) formula and sensitivity analysis employing the Monte Carlo Simulation (MCS) method through 10,000 repetitions in Oracle Crystal Ball® was used to ascertain the health risks associated with heavy metal exposure. Heavy metal concentrations were quantified through Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES). HQ point estimation results indicated that Cr exhibited the highest mean HQ value, whereas Cd demonstrated the lowest. In the probabilistic approach, the highest 95 percentile values were observed for Cr and Mo at 3.9E-01, while the lowest values were recorded for Cr and Mn at 1.10E-02. Heavy metal concentrations emerged as critical factors influencing non-carcinogenic and carcinogenic risks across all groups in the sensitivity analysis. The findings highlight the need for ongoing monitoring, research, and targeted regulations to address health risks from heavy metal exposure in bottled water and ensure public well-being.

Improving nitrate load simulation of the SWAT model in an extensively tile-drained watershed

Subsurface drainage systems are effective management practices employed to remove excess soil water, thereby improving soil aeration and crop productivity. However, these systems can also contribute to water quality issues by enhancing nitrate leaching and loads from agricultural fields. The Soil and Water Assessment Tool (SWAT) is commonly used to assess nitrate loads and long-term water quality impacts from agricultural watersheds. However, the current SWAT model oversimplifies nitrate transport processes by assuming a linear relationship between nitrate concentrations in tile flow and soil nitrate content. It also neglects the time lag between nitrate loading and transport with the flow. This study aimed to enhance the accuracy of nitrate load prediction by revising the subsurface drainage routine in the SWAT model. The revised routine was tested using flow and nitrate load measurements from a typical tile-drained watershed in east-central Illinois, U.S. The results demonstrated that the revised SWAT nitrate routine outperformed the current one in simulating nitrate transport at field and watershed scales. The revised routine improved the nitrate load prediction from an "unacceptable" to a "satisfactory" or "good" rating on the field scale. A sensitivity analysis conducted using the revised nitrate module showed the parameters directly associated with transpiration, groundwater discharge to the reach, the lag time of tile flow, and channel flow hydraulics were the most sensitive in nitrate load simulation. In addition, different tile depth scenarios were modeled to evaluate variation in the amount of surface runoff, tile flow, and nitrate loads by the surface flow and tile flow. The results of tile configuration scenarios agreed with understanding the tile flow process. The test results demonstrated the potential of the revised SWAT nitrate module as a tool to accurately evaluate the effects of tile drainage systems on water quality.

Improving algal bloom detection using spectroscopic analysis and machine learning: A case study in a large artificial reservoir, South Korea

The prediction of algal blooms using traditional water quality indicators is expensive, labor-intensive, and time-consuming, making it challenging to meet the critical requirement of timely monitoring for prompt management. Using optical measures for forecasting algal blooms is a feasible and useful method to overcome these problems. This study explores the potential application of optical measures to enhance algal bloom prediction in terms of prediction accuracy and workload reduction, aided by machine learning (ML) models. Compared to absorption-derived parameters, commonly used fluorescence indices such as the fluorescence index (FI), humification index (HIX), biological index (BIX), and protein-like component improved the prediction accuracy. However, the prediction accuracy was decreased when all optical indices were considered for computation due to increased noise and uncertainty in the models. With the exception of chemical oxygen demand (COD), this study successfully replaced biochemical oxygen demand (BOD), dissolved organic carbon (DOC), and nutrients with selected fluorescence indices, demonstrating relatively analogous performance in either training or testing data, with consistent and good coefficient of determination (R2) values of approximately 0.85 and 0.74, respectively. Among all models considered, ensemble learning models consistently outperformed conventional regression models and artificial neural networks (ANNs). However, there was a trade-off between accuracy and computation efficiency among the ensemble learning models (i.e., Stacking and XGBoost) for algal bloom prediction. Our study offers a glimpse of the potential application of spectroscopic measures to improve accuracy and efficiency in algal bloom prediction, but further work should be carried out in other water bodies to further validate our proposed hypothesis.

Forested watersheds provide the highest water quality among all land cover types, but the benefit of this ecosystem service depends on landscape context

Conversion of natural land cover can degrade water quality in water supply watersheds and increase treatment costs for Public Water Systems (PWSs), but there are few studies that have fully evaluated land cover and water quality relationships in mixed use watersheds across broad hydroclimatic settings. We related upstream land cover (forest, other natural land covers, development, and agriculture) to observed and modeled water quality across the southeastern US and specifically at 1746 PWS drinking water intake facilities. While there was considerable complexity and variability in the relationship between land cover and water quality, results suggest that Total Nitrogen (TN), Total Phosphorus (TP) and Suspended Sediment (SS) concentrations decrease significantly with increasing forest cover, and increase with increasing developed or agricultural cover. Catchments with dominant (>90 %) agricultural land cover had the greatest export rates for TN, TP, and SS based on SPARROW model estimates, followed by developed-dominant, then forest- and other-natural-dominant catchments. Variability in modeled TN, TP, and SS export rates by land cover type was driven by variability in natural background sources and catchment characteristics that affected water quality even in forest-dominated catchments. Both intake setting (i.e., run-of-river or reservoir) and upstream land cover were important determinants of water quality at PWS intakes. Of all PWS intakes, 15 % had high raw water quality, and 85 % of those were on reservoirs. Of the run-of-river intakes with high raw water quality, 75 % had at least 50 % forest land cover upstream. In addition, PWS intakes obtaining surface water supply from smaller upstream catchments may experience the largest losses of natural land cover based on projections of land cover in 2070. These results illustrate the complexity and variability in the relationship between land cover and water quality at broad scales, but also suggest that forest conservation can enhance the resilience of drinking water supplies.

What drives the change of nitrogen and phosphorus loads in the Yellow River Basin during 2006-2017?

The Yellow River Basin (YRB) plays a very important role in China's economic and social development and ecological security. In particular, the ecosystem of the YRB is sensitive to climate change. However, the change of nutrient fluxes in this region during the past years and its main driving forces remain unclear. In this study, a hydrologic model R System for Spatially Referenced Regressions on Watershed Attributes (RSPARROW) was employed to simulate the spatio-temporal variations in the fluxes of total nitrogen (TN) and total phosphorus (TP) during the period of 2006-2017. The results suggested that the TN and TP loads increased by 138% and 38% during 2006-2014, respectively, and decreased by 66% and 71% from 2015 to 2017, respectively. During the period of 2006-2017, the annual mean fluxes of TN and TP in the YRB were in the range of 3.9 to 591.6 kg/km²/year and 1.7 to 12.0 kg/km²/year, respectively. TN flux was low in the upstream area of the Yellow River, and presented a high level in the middle and lower reaches. However, the flux of TP in Gansu and Ningxia section was slightly higher than that in the lower reaches of the Yellow River. Precipitation and point source are the key drivers for the inter-annual changes of TN loads in most regions of the YRB. While the inter-annual variations of TP loads in the whole basin are mainly driven by the point source. This study demonstrates the important impacts of climate change on nutrient loads in the YRB. Moreover, management measures should be taken to reduce pollution sources and thus provide solid basis for control of nitrogen and phosphorus in the YRB.

Modeling nutrient flows from land to rivers and seas - A review and synthesis

Water quality modeling facilitates management of nutrient flows from land to rivers and seas, in addition to environmental pollution management in watersheds. In the present paper, we review advances made in the development of seven water quality models and highlight their respective strengths and weaknesses. Afterward, we propose their future development directions, with distinct characteristics for different scenarios. We also discuss the practical problems that such models address in the same region, China, and summarize their different characteristics based on their performance. We focus on the temporal and geographical scales of the models, sources of pollution considered, and the main problems that can be addressed. Summary of such characteristics could facilitate the selection of appropriate models for resolving practical challenges on nutrient pollution in the corresponding scenarios globally by stakeholders. We also make recommendations for model enhancement to expand their capabilities.

Response Model for Urban Area Source Pollution and Water Environmental Quality in a River Network Region

With the development of cities, urban area source pollution has become more severe and a significant source of water pollution. To study the relationship between urban area source pollution and water environmental quality in a river network, this study uses a city in the Yangtze River Delta, China, as an example. The Storm Water Management Model (SWMM) model and the MIKE11 model were combined into a unified modeling framework and used to simulate dynamic changes in the water quality of a river network under light rain, moderate rain, and heavy rain. In the study period, the annual urban area source input loads of potassium permanganate (COD_Mn), total phosphorus (TP), and ammonia nitrogen were 29.8, 0.9, and 4.8 t, respectively. The influence of light rain on the water quality of the river network was lagging and temporary, and rainfall area pollution was the primary contributor. Under the scenario of moderate rain, overflow from a pipeline network compounded rainfall runoff, resulting in a longer duration of impact on the water quality in the river. Additionally, the water quality in the river course was worse under moderate rain than under light or heavy rain. Under the scenario of heavy rain, rain mainly served a dilutive function. This research can provide support for urban area source pollution control and management.

Exploring the type and strength of nonlinearity in water quality responses to nutrient loading reduction in shallow eutrophic water bodies: Insights from a large number of numerical simulations

Reducing the load of nutrients is essential to improve water quality while water quality may not respond to the load reduction in a linear way. Despite nonlinear water quality responses being widely mentioned by studies, there is a lack of comprehensive assessment on the extent and type of nonlinear responses considering the seasonal changes. This study aimed to measure the strength of nonlinearity of theoretically possible water quality responses and explore their potential types in shallow eutrophic water bodies. Hereto, we generated 14,710 numerical water body cases that describe the water quality processes using the Environmental Fluid Dynamics Code (EFDC) and applied eight load reduction scenarios on each water body case. Inflows are simplified from Lake Dianchi. The climate conditions consider three cases: Lake Dianchi, Wissahickon Creek, and Famosa Slough. We then developed a nonlinearity strength indicator to quantify the strength and frequency of nonlinear water quality responses. Based on the quantification of nonlinearity, we clustered all the samples of water quality responses using K-Means, an unsupervised Machine Learning algorithm, to find the potential types of nonlinear water quality responses for TN (total nitrogen), TP (total phosphorus), and Chla (chlorophyll a). Results show linear or near-linear response types account for 90%, 69%, and 20% of TN, TP, and Chla samples respectively. TP and Chla could perform more types of nonlinearity. Representative nonlinear water quality responses include disproportional improvement, peak change (disappear, move forwards or afterward), and seasonal deterioration of TN after load reduction. This study would contribute to the current understanding of nonlinear water quality responses to load reduction and provide a basis to study under which conditions the nonlinear responses may emerge.

Modeling Chlorophyll a with Use of the SWAT Tool for the Nielba River (West-Central Poland) as an Example of an Unmonitored Watercourse

The majority of eutrophication studies focuses on lacustrine processes, thus riverine systems remain less recognized in this context. Moreover, since the availability of data related to parameters affecting this phenomenon is quite limited, modeling efforts should be considered. The current study verifies the SWAT model’s capability to simulate chlorophyll a loads for unmonitored watercourse. The analyses of the relationships between individual parameters, directly involved in the eutrophication process, help in the exploration of its dominant trends in SWAT modeling. The results obtained for the Nielba River pilot catchment (west-central Poland) showed a strong correlation of chlorophyll a with flow and surface runoff, but no relationship with temperature or solar radiation. Moreover, an impact of local conditions (hydrological features) on chlorophyll a load simulation could be traced in detail. The research specified the limitations and impact of generalization in the SWAT model on the results. Furthermore, intricacies related to the dataset statistical treatment (e.g., outliers) have been presented.

Spatial characteristics of nutrient budget on town scale in the Three Gorges Reservoir area, China

Accurately quantifying nutrient budget is an essential step toward sustainable nutrient management in large watersheds increasingly disturbed by human activity. A town-scale nutrient budget framework based on the Soil and Water Assessment Tool was developed for 2010-2012 in the Three Gorges Reservoir area in China (TGRA). Moran's I spatial correlation test and Geodetector spatial heterogeneity test were employed to systematically analyze the spatial characteristics of the resulting nutrient budget. The Moran's I value of total nitrogen (TN) and total phosphorus (TP) gradually increased from input to output in the range of 0.091-0.232 and 0.102-0.484, respectively. Towns with higher TN and TP inputs were largely concentrated in the main urban area of Chongqing because of its high population density. By contrast, towns with higher TN and TP outputs were concentrated in the head of the TGRA. The Moran's I values of the TN and TP retention coefficients (R) were 0.433 and 0.524, respectively, demonstrating clear spatial consistency. Towns with a "High-high" spatial consistency pattern and positive R value were concentrated in the tail and hinterland, while those with a "Low-low" spatial consistency pattern and negative coefficient value were located mainly in the head of the TGRA. This phenomenon was mostly caused by differences in regional elevation, the normalized difference vegetation index, and soil erosion factor. The interaction effect between any two of these three factors on nutrient retention (Geodetector q-value) was greater than 60%. Therefore, future nutrient management should be based on a full understanding of regional biophysical conditions, especially in large areas. These findings provide a new perspective on fine nutrient management.

Using physicochemical and biological parameters for the evaluation of water quality and environmental conditions in international wetlands on the southern part of Lake Urmia, Iran

The Kani Barazan and Yadegarlou wetlands in the southern part of Lake Urmia (Iran) have been substantially modified due to human activities and anthropogenic use. In recent years, freshwater-based eco-biological studies to recognize the quality of water resources have been greatly expanded. Microalgae and Cyanophyta are considered important bioindicators for the evaluation of water quality and wetland health worldwide. Herein, 22 microalgae and 5 Cyanophyta genera were identified in both wetlands, in which Cyanophyta has mainly caused blooms. Principal components analysis (PCA) was carried out based on links between the distribution of microalgae and Cyanophyta with physical and chemical parameters. The data showed that depth, turbidity, and the temperature had a significant influence on the microalga and Cyanophyta communities in both wetlands. Based on the biological properties, it seems that the Kani Barazan and Yadegarlou international wetlands experience meso-eutrophic conditions. The integration of the physical, chemical and biological parameters with the water quality index (WQI) revealed that both wetlands were polluted as a consequence of human activities. Moreover, a close relationship between WQI and the biological parameters was documented. Thus, we concluded that microalgae and Cyanophyta communities, their abundance patterns, and water quality changes could provide valuable data for the conservation of the Kani Barazan and Yadegarlou international wetlands.

Modeling and assessing water and nutrient balances in a tile-drained agricultural watershed in the U.S. Corn Belt

Identifying the key processes and primary sources of water and nutrient losses is essential for water quantity and quality management in watersheds. This is especially true in the U.S. Corn Belt, which has been recognized as the primary region contributing nutrient loads to the Great Lakes and the Gulf of Mexico. A SWAT (Soil and Water Assessment Tool) model simulation was set up in an agricultural watershed with about 50% tile drainage area in the U.S. Corn Belt to study the water and nutrient balance components for the whole watershed and the corn-soybean rotation system. The SWAT model was improved to consider additional nitrogen and phosphorus loss paths from the soil. The model was comprehensively calibrated and validated for simulating monthly stream flow, total suspended solids (TSS), nutrient loads (including total Kjeldahl nitrogen (TKN), nitrate and nitrite nitrogen (NO_x-N), total phosphorus (TP) and orthophosphate phosphorus (orthoP)), actual evapotranspiration (ET_a), leaf area index (LAI) and annual crop yields in the watershed from 2011 to 2019. Results showed the model performance was very good for simulating the stream flow, TSS and ET_a, and acceptable for nutrient loads, LAI and crop yields. ET_a, surface runoff, lateral soil flow, tile drainage and percolation respectively accounted for 65%, 15%, 2%, 8% and 9% of the precipitation. Fertilizer was the main source of nitrogen and phosphorus input to the watershed, and harvested crops were the main paths removing nutrients. Surface runoff, tile drainage and percolation each contributed about 30% of total nitrogen losses to water, with surface runoff being dominated by organic nitrogen while tile drainage and percolation were dominated by nitrate nitrogen. Phosphorus losses were mainly through surface runoff, which resulted in 66% of the total losses and was dominated by organic phosphorus and soluble phosphorus. Representing about 49% of the watershed area, the corn-soybean rotation system contributed 83% and 88% of the total nitrogen and phosphorus inputs, respectively, to the watershed, as well as 64% and 46% of the nitrogen and phosphorus losses to the water system, respectively. The non-growing season (October to the next April) was identified as the critical period resulting in water and nutrient losses due to low evapotranspiration and plant uptake. Targeted management strategies for reducing nutrient loads in key hydrological paths were suggested.

Analysis of indicators of surface water pollution in Atlantic Forest preservation areas

Managing water resources in regions with scarce data, like most developing countries, is still one of the major challenges around the world. Analysis of water quality parameters can provide important information for understanding the current status of water resources and their surroundings, including the changes that have occurred over time. This study aims to evaluate the influence of preservation areas on surface water quality in the Atlantic Forest biome. For this purpose, water quality monitoring sites with a greater number of parameters and longer monitoring time, located in six basins in the Atlantic Forest region of Brazil near preservation areas, were selected. This study employs seven statistical methods, such as cluster and principal component analysis (PCA), and promotes a robust analysis of the pollution of water resources in the Atlantic Forest. The most preserved basins, with more than 87% preservation area, have lower levels of pollution. The second most degraded basin, with 56% preservation area, presents intermediate pollution levels. The most degraded basin has the highest level of pollution. The basin with the lowest area of native vegetation is considered a degraded basin. Finally, non-point sources of pollution from agricultural activities were identified as the main sources of pollution in the region. The cophenetic correlation of 0.97 indicates a good performance of the cluster analysis. In addition, the pre-tests of PCA showed the suitability of the data for performing the test (Bartlett test, < 2.2e^-16 and KMO, P= 0.7). The first principal component in the PCA, which accounts for 31.4% of the total variation, is associated with strong ammonia nitrogen and total Kjeldahl nitrogen loads, and moderate biological oxygen demand and nitrite loads. The second component, representing 13.6% of the total variation, indicates periods of self-cleaning of water resources after contamination. The results indicate the importance of maintaining preservation areas in the watershed contribution areas for the improvement of surface water quality in the Atlantic Forest.

Application of Machine Learning for eutrophication analysis and algal bloom prediction in an urban river: A 10-year study of the Han River, South Korea

The increasing release of nutrients to aquatic environments has led to great concern regarding eutrophication and the risk of unwanted algal blooms. Based on observational data of 20 water quality parameters measured on a monthly basis at 40 stations from 2011 to 2020, this study applied different Machine Learning (ML) algorithms to suggest the best option for algal bloom prediction in the Han River, a large river in South Korea. Eight different ML algorithms were categorized into several groups of statistical learning, regression family, and deep learning, and were then compared for their suitability to predict the chlorophyll-derived trophic index (TSI-Chla). ML algorithms helped identify the most important water quality parameters contributing to algal bloom prediction. The ML results confirmed that eutrophication and algal proliferation were governed by the complex interplay between nutrients (nitrogen and phosphorus), organic contaminants, and environmental factors. Of the models tested, the adaptive neuro-fuzzy inference system (ANFIS) exhibited the best performance owing to its consistent and outperforming prediction both quantitatively (i.e., via regression) and qualitatively (i.e., via classification), which was evidenced by the lowest value of mean absolute error (MAE) of 0.09, and the highest F1-score, Recall and Precision of 0.97, 0.98 and 0.96, respectively. In a further step, a representative web application was constructed to assist common users to predict the trophic status of the Han River. This study demonstrated that ML techniques are not only promising for highly accurate water quality modeling of urban rivers, but also reduce time and labor intensity for experiments, which decreases the number of monitored water quality parameters, providing further insights into the driving factors of water quality deterioration. They ultimately help devise proactive strategies for sustainable water management.

From Monitoring and Modeling to Management: How to Improve Water Quality in Brazilian Rivers? A Case Study: Piabanha River Watershed

Water quality has been a global concern, as evidenced by UN Sustainable Development Goals. The current paper has focused on the Piabanha River rehabilitation as a case study which can be generalized to other similar watersheds. A monitoring program during a hydrological year was carried out, and different databases were used to calibrate and validate the QUAL-UFMG water quality model. Sanitation is the major problem in the watershed, notably in its headwater catchments, which concentrate the most urbanized regions where water quality is worse in the dry season due to low river flows. Thus, simulations of the river water quality have been performed through computational modeling suggesting organic load reductions in some sub-basins. In conclusion, some strategies to improve water quality have been discussed: (i) The water quality rehabilitation must consider progressive goals of pollution reduction starting with an initial implementation in a reduced area. The monitoring should be based on a few parameters relevant and simple to monitor. (ii) Pollution reduction ought to be carried out strategically with deadlines and intermediate goals that must be agreed upon between the stakeholders in the watershed. (iii) Watershed committees should supervise projects to improve water quality in partnership with the State Prosecutor’s Office.

Choosing an appropriate water quality model-a review

Water quality models are quite complex to use even for scientists, requiring knowledge in different areas such as biology, chemistry, physics, and engineering. Hence, the use of these models by a non-specialist is quite complicated, demanding considerable time and research, particularly to choose which model is the most appropriate for a given situation. In this study, a comparative guide is suggested, which can help users select the appropriate water quality model for certain systems and variables. Five models were considered as follows: AQUATOX, CE-QUAL-W2, Spatially Referenced Regression Model on Watershed Attributes (SPARROW), Soil and Water Assessment Tool (SWAT), and Water Quality Analysis Simulation Program 7 (WASP7), which have been widely used during the last 5 years. All of these selected models are free and easily available. It was verified that each model has its particularities and applications; however, the AQUATOX model has several advantages compared with the other models analyzed. In addition, to illustrate the availability of the proposed comparative guide, a case study was carried out to demonstrating the selection process of the selected models.