Fig 1 - available from: Neural Computing and Applications
This content is subject to copyright. Terms and conditions apply.
Source publication
The management of river water quality is one the most significant environmental challenges. Water quality index (WQI) describes several water quality variables at a certain aquatic environment and time. Classically, WQI is commonly computed using the traditional methods which involved lengthy computation, consume timing and occasionally associated...
Similar publications
Air quality status on the East Coast of Peninsular Malaysia is dominated by Particulate Matter (PM10) throughout the years. Studies have affirmed that PM10 influence human health and the environment. Therefore, precise forecasting algorithms are urgently needed to determine the PM10 status for mitigation plan and early warning purposes. This study...
Citations
... Regarding the simulation and prediction of WQ, adaptive boosting (Adaboost) [26], gradient boosting (GBM) [27], extreme gradient boosting (XGBoost) [28], decision tree (DT) [29,30], extra trees (ExT) [31], random forest (RF) [26], multilayer perceptron (MLP) [32], radial basis function (RBF) [33], deep feed-forward neural network (DFNN) [34] and convolutional neural network (CNN) [17] have been reported. Among these techniques, the multilayer perceptron (MLP) model has mostly outperformed the others in precision and accuracy [35] and is the most widely used architecture [36]. ...
Identifying and measuring potential sources of pollution is essential for water management and pollution control. Using a range of artificial intelligence models to analyze water quality (WQ) is one of the most effective techniques for estimating water quality index (WQI). In this context, machine learning–based models are introduced to predict the WQ factors of Southeastern Black Sea Basin. The data comprising monthly samples of different WQ factors were collected for 12 months at eight locations of the Türkiye region in Southeastern Black Sea. The traditional evaluation with WQI of surface water was calculated as average (i.e. good WQ). Single multiplicative neuron (SMN) model, multilayer perceptron (MLP) and pi-sigma artificial neural networks (PS-ANNs) were used to predict WQI, and the accuracy of the proposed algorithms were compared. SMN model and PS-ANNs were used for WQ prediction modeling for the first time in the literature. According to the results obtained from the proposed ANN models, it was found to provide a highly reliable modeling approach that allows capturing the nonlinear structure of complex time series and thus to generate more accurate predictions. The results of the analyses demonstrate the applicability of the proposed pi-sigma model instead of using other computational methods to predict WQ both in particular and other surface water resources in general.
... The traditional approach to water quality assessment is the use of water quality indices (WQIs) (Khuan et al. 2002;Asadollah et al. 2021). WQIs are an important tool for water resources management as they represent a single measure that encompasses various physical and chemical parameters of water quality (Zali et al. 2011;Hameed et al. 2017). However, the calculation of these indices is often associated with challenges, such as being time-consuming, complex and prone to inconsistencies due to the use of different equations and methods (Kouadri et al. 2021). ...
In Saudi Arabia, water pollution and drinking water scarcity pose a major challenge and jeopardise the achievement of sustainable development goals. The urgent need for rapid and accurate monitoring and assessment of water quality requires sophisticated, data-driven solutions for better decision-making in water management. This study aims to develop optimised data-driven models for comprehensive water quality assessment to enable informed decisions that are critical for sustainable water resources management. We used an entropy-weighted arithmetic technique to calculate the Water Quality Index (WQI), which integrates the World Health Organization (WHO) standards for various water quality parameters. Our methodology incorporated advanced machine learning (ML) models, including decision trees, random forests (RF) and correlation analyses to select features essential for identifying critical water quality parameters. We developed and optimised data-driven models such as gradient boosting machines (GBM), deep neural networks (DNN) and RF within the H2O API framework to ensure efficient data processing and handling. Interpretation of these models was achieved through a three-pronged explainable artificial intelligence (XAI) approach: model diagnosis with residual analysis, model parts with permutation-based feature importance and model profiling with partial dependence plots (PDP), accumulated local effects (ALE) plots and individual conditional expectation (ICE) plots. The quantitative results revealed insightful findings: fluoride and residual chlorine had the highest and lowest entropy weights, respectively, indicating their differential effects on water quality. Over 35% of the water samples were categorised as ‘unsuitable’ for consumption, highlighting the urgency of taking action to improve water quality. Amongst the optimised models, the Random Forest (model 79) and the Deep Neural Network (model 81) proved to be the most effective and showed robust predictive abilities with R² values of 0.96 and 0.97 respectively for testing dataset. Model profiling as XAI highlighted the significant influence of key parameters such as nitrate, total hardness and pH on WQI predictions. These findings enable targeted water quality improvement measures that are in line with sustainable water management goals. Therefore, our study demonstrates the potential of advanced, data-driven methods to revolutionise water quality assessment in Saudi Arabia. By providing a more nuanced understanding of water quality dynamics and enabling effective decision-making, these models contribute significantly to the sustainable management of valuable water resources.
... It concluded that DO, suspended solids and nitrates are the key input parameters to predict WQI using ML models. For describing the relationship between WQI and various chemical parameters (e.g., BOD, chemical oxygen demand (COD), DO, nitrate, pH, suspended solids) in tropical environments, Hameed et al. (2017) used two ANN models, namely Back-Propagation Neural Network (BPNN) and Radial Basis Function Neural Network (RBFNN), and the RBFNN model produced better prediction results. Ho et al. (2019) focused on the WQI classification of Malaysian river water using six water quality parameters (Ammonical nitrogen, Biological Oxygen Demand, Chemical Oxygen Demand, Dissolved Oxygen, pH, and suspended solids). ...
Despite high pollution levels in Indian rivers, a comprehensive study on the water quality index (WQI) remains elusive. WQI values were computed, and their classes were determined using six water quality parameters from an available decennial dataset (n = 3595) on Indian Rivers. This study aims to assess the spatial distribution of WQI values and their classes across Indian River systems while exploring the application of machine learning (ML) based models in predicting WQI classes using a reduced number of input parameters.
Modeling experiments were designed on five models- Decision Tree (DT), Random Forest (RF), Gradient Boosted Trees (GBT), Artificial Neural Network (ANN), and Support Vector Machine (SVM) for predicting WQI classes. Each model was trained with input parameters and WQI classes with 2990 datasets. Testing of WQI classes by each model was made on 605 datasets under different framework sets. Models’ performance metrics were evaluated by accuracy, weighted mean recall and precision, and F-score.
Our study demonstrates that the two largest systems, Ganga and Brahmaputra, lie on the extremes of the WQI (mean) spectrum, reflecting the impact of contrasting population density, industrial activities, change in land-use-land-cover pattern, and agricultural use on the riverine WQI. Our modeling experiments underscore that with only three input parameters, GBT can predict WQI classes with > 80% of performance metrics. With only two input parameters, GBT, RF, and ANN, all can provide reliable estimates. Our study highlights that ML models can serve as decision-supporting tools for water resource policymakers and managers in making effective pollution control and water resource management decisions.
... There are various artificial intelligence models for predicting air pollution, such as ANN, support vector machines (SVM), logistic fuzzy deep neural networks, etc. Deep Neural Networks (DNN) is a more complex version of an artificial simulation neural network, which is considered to be the most accurate model for predicting air pollution [2]. ...
In the face of escalating global environmental issues, the role of artificial intelligence (AI) in environmental protection has become increasingly significant. One of the key areas of focus is water quality monitoring in sewage treatment. AI has been instrumental in predicting and identifying potential contaminants, thus ensuring the safety and cleanliness of water resources. This not only aids in maintaining public health but also contributes to the preservation of aquatic ecosystems. Another crucial aspect is the prediction of air pollution. AI algorithms have been developed to forecast air quality, enabling timely measures to mitigate the adverse effects of air pollution. This is particularly beneficial in urban areas where air pollution levels are typically high. The paper also explores AI's role in soil protection, specifically in detecting soil quality. AI can help identify nutrient deficiencies, monitor moisture levels, and detect the presence of harmful substances, thereby promoting sustainable agriculture and preserving soil health. Furthermore, the paper extends its scope to specific applications of AI in other areas of environmental protection. These include wildlife conservation, waste management, and climate change mitigation, among others. Finally, the paper looks forward to the future impact and development of AI in the field of environmental protection. It discusses the potential challenges and opportunities, and how AI can be leveraged to create a more sustainable and environmentally-friendly future. The paper underscores the importance of continued research, and development in AI to fully harness its potential in environmental protection.
... Machine learning algorithms such as the artificial neural network (ANN) were widely adopted for WQ and WQI modeling [2][3][4][5]. To improve the model's capability in handling temporal features, the recurrent neural network (RNN) was designed, where the neurons within the RNN layer are interconnected to allow feedback and self-learning during backpropagation. ...
... These models had difficulties in predicting the future trend of the WQ parameter and lack practicability in real-life applications; given that WQ management personnel cannot react immediately to an expected water pollution event. Most WQI modeling works focused on WQI estimation based on different WQ input combinations [2,3,5], while direct forecasting of WQI time series is rarely reported. A WQI estimation work may save the hassles of manual calculations or reduce the number of sensors required but is unable to predict the future WQI for better remedy planning. ...
... By separating the complex input signal into IMFs oscillating at fixed frequencies, the non-stationary behavior of the input data was diminished. The decomposed IMFs came in a more Fig. 4. Unlike past studies that typically employed the DOE data collected on a bimonthly basis [3,5], the IoT-based WQ data now allow researchers to forecast WQI changes at higher resolution, hence leading to the potential of applying the recurrent-based DL models. 17,544 sets of observations were collected from 1st January 2019 to 31st December 2020, which comprised AN, BOD, COD, DO, pH and SS at hourly intervals. ...
The water quality index (WQI) serves as a global representation of river water quality (WQ). Existing studies related to the WQI have mainly focused on two aspects: (i) a WQI point estimation using multiple WQ inputs; and (ii) a one-step-ahead WQI forecasting with datasets of lower temporal resolution. These approaches, however, are limited in their ability to forecast future trends of the WQI for timely and prompt responses to pollution events. In this study, the deep learning algorithms, namely the long short-term memory (LSTM) and the gated recurrent unit (GRU), were selected for direct multi-step-ahead WQI forecasting. To enhance the capability of the models in capturing their temporal patterns, the input signal was pre-decomposed using the empirical mode decomposition (EMD) and variational mode decomposition (VMD) into several intrinsic mode functions (IMFs). The characteristics of these IMFs were then analyzed and used to ease model learning on capturing their temporal patterns. Our study shows that the selection of signal decomposition strategies significantly impact the model performance. Both deep learning algorithms offered comparable performances, with the VMD-LSTM exhibiting the lowest prediction errors (MAPE = 1.9237%) and the highest Kling–Gupta efficiency (KGE = 0.6761) over a two-month test period. To the best of our knowledge, this is the first paper that applies a direct forecast approach for the multi-step-ahead WQI forecasting, using the IMFs obtained from the EMD and VMD decompositions. The performance of the model was evaluated through a rolling forward setup to ensure its consistency across different test periods. The proposed modeling framework holds the potential to assist policymakers and stakeholders in decision-making, particularly in planning remedies and efficient water resource management.
... The selection of suitable model inputs was based on criteria that prioritized the lowest root mean square error (RMSE) and the highest coefficient of determination (R 2 ). This research methodology shares similar objectives with the studies conducted by (Gazzaz et al., 2012;Hameed et al., 2017). However, it introduces a novel approach involving the use of standardized regression coefficients to evaluate the most influential independent parameters across all six predictive models. ...
... Moreover, the choice of input features plays a pivotal role in the stability and resilience of the prediction model. By thoughtfully examining and choosing relevant input variables, you can enhance the models' stability, ensuring consistent and dependable predictions across different scenarios (Gültekin and Sakar, 2018;Hameed et al., 2017;Singha et al., 2021;Wong et al., 2022). The relative importance score of each input feature is determined by the optimization algorithm used by the respective prediction models. ...
The Water Quality Index (WQI) is a primary metric used to evaluate and categorize surface water quality which plays a crucial role in the management of fresh water resources. Machine Learning (ML) modeling offers potential insights into water quality index prediction. This study employed advanced ML models to get potential insights into the prediction of water quality index for the Aik-Stream, an industrially polluted natural water resource in Pakistan with 19 input water quality variables aligning them with surrounding land use and anthropogenic activities. Six machine learning algorithms, i.e. Adaptive Boosting (AdaBoost), K-Nearest Neighbors (K-NN), Gradient Boosting (GB), Random Forests (RF), Support Vector Regression (SVR), and Bayesian Regression (BR) were employed as benchmark models to predict the Water Quality Index (WQI) values of the polluted stream to achieve our objectives. For model calibration, 80% of the dataset was reserved for training, while 20% was set aside for testing. In our comparative analyses of predictive models for water quality index, the Gradient Boost (GB) model stood out the fittest for its precision, utilizing a combination of just seven parameters (chemical oxygen demand, total organic carbon, oil & grease, Ammonia- nitrogen, arsenic, nickel and zinc), surpassing other models by achieving better results in both training (R2 = 0.88, RMSE = 7.24) and testing (R2 = 0.85, RMSE = 8.67). Analyzing feature importance showed that all the selected variables, except for NO3 N, TDS and temperature had an impact on the accuracy of the models predictions. It is concluded that the application of machine learning to assess water quality in polluted environments enhances accuracy and facilitates real-time tracking, enabling proactive risk mitigations.
... Notably, in 1999, alarming 42% of Malaysia's river basins exhibited pollution attributable to suspended solids (SS), a consequence of poorly planned and unregulated land clearing activities. An additional 30% of these river basins were contaminated by organic matter originating from industrial effluents, whereas 28% were contaminated with ammoniacal nitrogen (NH 3 -N) resulting from sewage disposal and animal husbandry practices [10,11]. The adverse consequences of this pervasive degradation of river water quality extend far beyond the environmental domain, encompassing significant ramifications for human health, ecological stability, and national economic well-being [12][13][14]. ...
... ANNs, which have witnessed significant advancements since their inception, stand at the forefront of data-driven modelling methodologies. Their capacity to model intricate nonlinear interactions between multiple variables in complex systems renders them invaluable tools [11,[22][23][24][25]. In recent decades, ANNs have been demonstrated to predict a diverse array of water quality parameters across various aquatic environments, including salinity, total dissolved solids (TDS), electrical conductivity, turbidity, chemical oxygen demand (COD), biochemical oxygen demand (BOD), dissolved oxygen (DO), ammoniacal nitrogen (NH 3 -N), and water quality indices (WQI) [20,21,24,26,27]. ...
... The RB NN model architecture for riverine load prediction was developed based on the broad conceptual ANN developed by Basant et al. [26]. The nonlinear Gaussian function equation is given by Eq. (4), where μ is the centre of the Gaussian function (mean value of x) and d is the distance (radius) from the centre, providing a measure of the spread of the Gaussian curve [11]. The spread constant d is used in this study as the maximum spread constant for the radial basis function [52]. ...
... Zubaidi et al. [126] 2020 Hybrid Artificial Neural Network Model / Urban stochastic water demand prediction Local water managers can use this study to efficiently manage their existing water system and plan extensions to meet the increasing demand for water. Hameed et al. [127] 2023 Review/ development of stormwater runoff management The evolution of stormwater runoff management concepts was examined through reference emergence analysis, keyword clustering, and keyword emergence analysis. In addition to identifying hotspots, trends, and technological advancements, this approach also highlighted limitations. ...
Digital twins provide insights into physical objects by serving as advanced virtual representations. Their sensors capture detailed information about an object's functionality through their use of various sensors. It is possible to gain a deep understanding of the object's performance and potential areas for improvement by collecting data, which includes metrics such as energy output, temperature, and weather conditions. Digital twins are becoming important in a variety of research and industrial application sectors as production lines and processes become more digitalized and as improved data analysis techniques such as machine learning and enhanced visualization techniques are used. There is no unified definition of the digital twin concept in scientific literature, which results in imprecise applications and the weakening of its terminology. However, this study demonstrates how digital twin models can be applied to urban drainage systems. As a result, this highlights the relatively novel use of digital twins within the field of urban water system engineering. Our review of the language, practices, and directions in smart stormwater management provides a framework to organize and comprehend the current research landscape while highlighting crucial areas for future research. Our results demonstrate that there is near-unanimous agreement within the literature that smart technology has been, or will be, advantageous for stormwater management. However, while some progress has been made in terms of quantity management, maturity in water quality management has not yet been achieved. This study examines the scientific literature on digital twins in the application of artificial intelligence for smart city stormwater infrastructure systems, specifically focusing on urban drainage systems. A demonstration of the workflow and features of current digital twin applications in urban drainage systems is also presented, providing valuable insights and guidance for future research and development in this field.
... Similarly, a robust artificial intelligence framework facilitated the multi-hour advance prediction of surface ozone gas concentration (AlOmar et al., 2020). In tropical regions, such as Malaysia, machine learning played a pivotal role in predicting the Water Quality Index (WQI) (Hameed et al., 2017). Moreover, the deployment of advanced techniques like Radial Basis Function Neural Networks (RBFNN) and Feed Forward Neural Networks (FFNN) enabled precise forecasting of daily river flow levels in the Johor River, Malaysia (Yaseen et al., 2016). ...
Predicting groundwater levels is pivotal in curbing overexploitation and ensuring effective water resource governance. However, groundwater level prediction is intricate, driven by dynamic nonlinear factors. To comprehend the dynamic interaction among these drivers, leveraging machine learning models can provide valuable insights. The drastic increase in computational capabilities has catalysed a substantial surge in the utilisation of machine learning-based solutions for effective groundwater management. The performance of these models highly depends on the selection of hyperparameters. The optimisation of hyperparameters is a complex process that often requires application-specific expertise for a skillful prediction. To mitigate the challenge posed by hyperparameter tuning's problem-specific nature, we present an innovative approach by introducing the automated machine learning (AutoML-GWL) framework. This framework is specifically designed for precise groundwater level mapping. It seamlessly integrates the selection of best machine learning model and adeptly fine-tunes its hyperparameters by using Bayesian optimisation. We used long time series (1997-2018) data of precipitation, temperature, evaporation, soil type, relative humidity, and lag of groundwater level as input features to train the AutoML-GWL model while considering the influence of Land Use Land Cover (LULC) as a contextual factor. Among these input features, the lag of groundwater level emerged as the most relevant input feature. Once the model is trained, it performs well over the unseen data with a strong correlation of coefficient (R = 0.90), low root mean square error (RMSE = 1.22), and minimal bias = 0.23. Further, we compared the performance of the proposed AutoML-GWL with sixteen benchmark algorithms comprising baseline and novel algorithms. The AutoML-GWL outperforms all the benchmark algorithms. Furthermore, the proposed algorithm ranked first in Friedman's statistical test, confirming its reliability. Moreover, we conducted a spatial distribution and uncertainty analysis for the proposed algorithm. The outcomes of this analysis affirmed that the AutoML-GWL can effectively manage data with spatial variations and demonstrates remarkable stability when faced with small uncertainties in the input parameters. This study holds significant promise in revolutionising groundwater management practices by establishing an automated framework for simulating groundwater levels for sustainable water resource management.
... This compelling relationship served as the foundation for the de-velopment of a machine learning model capable of predicting oxygen levels based on temperature variations. This innovative approach allows for the prediction of oxygen concentrations in the water, offering insights into the interrelationship between temperature and dissolved oxygen crucial for the management of water quality.The quality of the water becomes a growing concern throughout the developing world [4]. The integration of machine learning techniques into water quality assessment is an important step toward revolutionizing predictive modeling in environmental science. ...
This study delves into how AI can be applied within water and environmental engineering research, particu- larly emphasizing the utilization of machine learning models to advance the accuracy of water quality predictions in the Delaware River. Through an analysis of time series data spanning from 2020 to 2022 and the utilization of exploratory data analysis methods, this investigation scrutinizes various elements influenc- ing the dynamics of water quality. For water quality time series analysis, identifying changes in long-term trends is important, yet identifying specific change-points is also important[3]. A strong correlation is notably detected between levels of dissolved oxygen and recorded temperatures. Leveraging this correlation, an intricate polynomial regression model is crafted to forecast dissolved oxygen concentrations based on expected temperature values. This predictive model not only clarifies the inherent link between dissolved oxygen and temperature but also offers insights into projecting future dissolved oxygen levels in the Delaware River, considering anticipated temperature fluctuations. These findings hold significant promise, potentially enhancing ecologi- cal evaluations and the development of impactful management strategies, specifically designed for water quality monitoring and conservation efforts within the Delaware River basin. Hence, eval- uation of water quality of groundwater is extremely important to prepare for remedial measures[1].
