The performance of classification and forecasting Dong Nai River water quality for sustainable water resources management using neural network techniques

Than, Nguyen Hien; Ly, Che Dinh; Tất, Phạm Văn

doi:10.1016/j.jhydrol.2021.126099

Cited by 37 publications

(17 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Water quality is an important indicator to determine the potential use of water and is one of the most important factors contributing to human health [28,41]. Currently, various water quality assessment methods have been developed to support water quality management, such as water quality index, pollution index, fuzzy comprehensive evaluation, multivariate analysis, and others [42,43]. Assessment of the water quality can also be done using the STORET method.…”

Section: Resultsmentioning

confidence: 99%

Water Quality Assessment of West Tarum Canal for Drinking Water Supply

2021

View full text Add to dashboard Cite

West Tarum Canal (WTC) is a canal that drains raw water from the Jatiluhur Dam. Nearly 81% of the raw water for drinking water used by Jakarta’s people comes from this canal. However, various land uses such as agriculture, industry, settlements, and infrastructure development impact WTC’s water quality. This research aims to assess WTC’s water quality in 2016-2020 based on water quality standards set by the Government and using the STORET method. The results of this research indicate that the concentrations of TDS (142-351 mg/L), Fe (0.1-0.15 mg/L), Mn (0.03-0.1 mg/L) are meet the standards, while DO (3.6-4.9 mg/L), BOD (4-10 mg/L), COD (13-30 mg/L) are not meet the standards. Almost all monitoring points have pH values between 5.75-7.68 that are meet the standards. The STORET score of WTC is from -26 to -38 with an average of -30, which indicates that WTC’s water quality is moderately polluted. Water contamination in WTC will burden the drinking water processing and ultimately affect the community’s ability to pay for drinking water. This research also shows the need for integrated management of WTC from upstream to downstream and the need to increase collaboration between stakeholders in carrying out this management.

show abstract

Section: Resultsmentioning

confidence: 99%

Water Quality Assessment of West Tarum Canal for Drinking Water Supply

2021

View full text Add to dashboard Cite

show abstract

“…Among them, recurrent neural networks (RNN) show excellent performance in dealing with time series (Zhang et al, 2022). Long short-term memory (LSTM) is used to predict the classification of the Dong Nai River (Than et al, 2021). A model based on bi-directional long short-term memory (Bi-LSTM) is proposed to assess the water quality factors of a river in India (Khullar & Singh, 2022).…”

Section: Related Workmentioning

confidence: 99%

A spatiotemporal dissolved oxygen prediction model based on graph attention networks suitable for missing data

Fang

Liu

2023

Preprint

View full text Add to dashboard Cite

The evaluation and forecasting of water quality are crucial for the prevention and management of water pollution. The majority of the current water quality prediction techniques only take into account a single site. Moreover, there are a huge number of missing values in the monitored water quality data, which prevents the model from functioning correctly. To address these problems, a spatiotemporal prediction framework based on graph attention network (GAN) appropriate for missing data is developed. First, a module based on Neural Controlled Differential Equations (NCDE) is employed to deal with missing data. Then, an improved GAN is utilized to establish the spatiotemporal link of the data on water quality. A new iterative optimization method based on a k-nearest neighbor graph is proposed to enhance the performance of the graph model. To enhance the model's ability to handle multiple features, Shapley additive explanations (SHAP) is applied to evaluate the importance of features, and then multiple features is screened, and they are fused through GRU. Moreover, a fused graph attention mechanism is proposed to improve the robustness of GAN against noise. According to the experimental results, the proposed framework outperforms other models, and the model maintains stable performance for missing data. The spatiotemporal prediction method established in this paper can be a potential strategy to expand the water quality prediction model library.

show abstract

“…Zhou et al (2018) proposed an improved and integrated method called, Improved Grey Relational Analysis and Long Short-Term Memory to attain better performance in water quality prediction. Than et al (2021) proposed a neural network model for accurate water quality prediction in the Dong Nai River. Yet another water quality prediction method using machine learning methods using artificial neural networks and support vector machines was investigated in Deng et al (2021) and Mostafa et al (2022).…”

Section: Related Workmentioning

confidence: 99%

Normalized hellinger feature selection and soft margin boosting classification for water quality prediction

Karuppannan

Jayagopal

2023

Expert Systems

View full text Add to dashboard Cite

Several research works have been introduced using classification methods for predicting water quality. However, the accuracy level with which the prediction was made and the time consumed in the prediction was not found to be satisfactory. In order to solve these problems, water quality forecasting is done using machine learning and deep learning techniques. The study introduces an efficient water quality prediction called, Normalized Hellinger Feature Selection based Soft Margin Boosting (NHFS‐SMB) is developed based on pre‐processing, feature selection, and classification with higher accuracy and lesser time consumption. The Hellinger Similarity Feature selection model is used to select relevant features for water quality prediction. SMBC is an ensemble classifier that combines several weak classifiers (i.e., associative classifier) to provide the final prediction results as normal or contaminated conditions with lesser error rates. Experimental evaluation is determined by the factors such as accuracy, precision, recall, F1‐score, and ROC Curve with respect to the number of data points.

show abstract

The performance of classification and forecasting Dong Nai River water quality for sustainable water resources management using neural network techniques

Cited by 37 publications

References 44 publications

Water Quality Assessment of West Tarum Canal for Drinking Water Supply

Water Quality Assessment of West Tarum Canal for Drinking Water Supply

A spatiotemporal dissolved oxygen prediction model based on graph attention networks suitable for missing data

Normalized hellinger feature selection and soft margin boosting classification for water quality prediction

Contact Info

Product

Resources

About