A management-oriented water quality model for data scarce catchments

Slaughter, Andrew; Hughes, Denis A.; Retief, D.C.H.; Mantel, Sukhmani

doi:10.1016/j.envsoft.2017.07.015

Cited by 39 publications

(21 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Liu & Lu (2014) compared support vector machine (SVM) models with ANNs to predict WQ in a river. Moreover, there are several studies based on multiple linear regression methods combined with AI to develop WQ models (Slaughter et al, 2017;Tomas et al, 2017). An AI system was developed (Ji et al, 2017) by combining multiple models based on SVM, ANNs, LR to prove the supremacy of SVM in predicting dissolved oxygen (DO) concentration in Wen-Rui Tang River, China.…”

Section: Machine Learning For Water Quality Evaluationmentioning

confidence: 99%

Dynamic multi-objective optimisation using deep reinforcement learning: benchmark, algorithm and an application to identify vulnerable zones based on water quality

Hasan

Lwin

Imani

et al. 2019

Engineering Applications of Artificial Intelligence

View full text Add to dashboard Cite

Dynamic multi-objective optimisation problem (DMOP) has brought a great challenge to the reinforcement learning (RL) research area due to its dynamic nature such as objective functions, constraints and problem parameters that may change over time. This study aims to identify the lacking in the existing benchmarks for multi-objective optimisation for the dynamic environment in the RL settings. Hence, a dynamic multi-objective testbed has been created which is a modified version of the conventional deep-sea treasure (DST) hunt testbed. This modified testbed fulfils the changing aspects of the dynamic environment in terms of the characteristics where the changes occur based on time. To the authors' knowledge, this is the first dynamic multi-objective testbed for RL research, especially for deep reinforcement learning. In addition to that, a generic algorithm is proposed to solve the multi-objective optimisation problem in a dynamic constrained environment that maintains equilibrium by mapping different objectives simultaneously to provide the most compromised solution that closed to the true Pareto front (PF). As a proof of concept, the developed algorithm has been implemented to build an expert system for a real-world scenario using Markov decision process to identify the vulnerable zones based on water quality resilience in São Paulo, Brazil. The outcome of the implementation reveals that the proposed parity-Q deep Q network (PQDQN) algorithm is an efficient way to optimise the decision in a dynamic environment. Moreover, the result shows PQDQN algorithm performs better compared to the other state-of-the-art solutions both in the simulated and the real-world scenario.

show abstract

Section: Machine Learning For Water Quality Evaluationmentioning

confidence: 99%

Dynamic multi-objective optimisation using deep reinforcement learning: benchmark, algorithm and an application to identify vulnerable zones based on water quality

Hasan

Lwin

Imani

et al. 2019

Engineering Applications of Artificial Intelligence

View full text Add to dashboard Cite

show abstract

“…Therefore, several specialized memory units have been developed-for example, the long short term memory cell [32] and gated recurrent unit [33]. In this study, in order to make full use of the time series characteristics (TSC) of water quality parameters and avoid the long-term dependence problem of ordinary RNN, the LSTM network model is established to predict the concentration of major pollutants in lakes and the changing trend of lake water quality, so as to provide reference for water quality control and water resources development and utilization [34].…”

Section: Introductionmentioning

confidence: 99%

Monitoring Water Quality Parameters of Taihu Lake Based on Remote Sensing Images and LSTM-RNN

Huang

Wang

2020

IEEE Access

View full text Add to dashboard Cite

Long-term dynamic monitoring of the water quality of freshwater resources is of great significance to the stable and orderly operation of human society. Most studies only use one of the measured data from the monitoring station and the remote sensing satellite data as the data source. However, a single data source will cause inaccuracy and incompatibility of the water quality monitoring results. Few studies start from practical applications to generate digital images of water quality changes. Furthermore, the performance of shallow neural networks in water quality monitoring is not often ideal. Considering the above problems, we proposed a long short-term memory network model (LSTM) to invert four key water parameters including pondus hydrogenii (PH), dissolved oxygen (DO), chemical oxygen demand (CODMn) and ammonia-nitrogen (NH3-H). Moreover, the model was applied to the satellite images of various periods to generate the inverted image of each water quality parameter. The proposed model has exhibited excellen t performance in the water quality ass essment of the project, with the coefficient of determination (R 2), the relative root-mean-square error (rRMSE), and the mean relative error (MRE) values of 0.83, 0.16, and 0.18, respectively. And the inverted images are also consistent with the official information. INDEX TERMS Remote sensing, water quality parameters, water quality monitoring, LSTM network.

show abstract

“…The increasing demand for modeling integrated physio-chemical and hydro-biological processes in typical ecosystems requires the inclusion of additional fluxes to simulate mutually dependent complex processes such as nutrient generation, transport, transformation, and recirculation in hydrological systems. Such high expectation output from the advanced and coupled WQM models is limited due to over-parameterization and associated assumptions in the process [ 28 , 29 ]. Opinions remain divided on whether higher dimensional complex models (two-dimensional or three-dimensional) or simple models based on appropriate theories and logic is the best approach to water quality modeling [ 30 , 31 , 32 ].…”

Section: Introductionmentioning

confidence: 99%

Assessment of Water Quality Profile Using Numerical Modeling Approach in Major Climate Classes of Asia

Iqbal

Shoaib

Farid

et al. 2018

IJERPH

View full text Add to dashboard Cite

A river water quality spatial profile has a diverse pattern of variation over different climatic regions. To comprehend this phenomenon, our study evaluated the spatial scale variation of the Water Quality Index (WQI). The study was carried out over four main climatic classes in Asia based on the Koppen-Geiger climate classification system: tropical, temperate, cold, and arid. The one-dimensional surface water quality model, QUAL2Kw was selected and compared for water quality simulations. Calibration and validation were separately performed for the model predictions over different climate classes. The accuracy of the water quality model was assessed using different statistical analyses. The spatial profile of WQI was calculated using model predictions based on dissolved oxygen (DO), biological oxygen demand (BOD), nitrate (NO3), and pH. The results showed that there is a smaller longitudinal variation of WQI in the cold climatic regions than other regions, which does not change the status of WQI. Streams from arid, temperate, and tropical climatic regions show a decreasing trend of DO with respect to the longitudinal profiles of main river flows. Since this study found that each climate zone has the different impact on DO dynamics such as reaeration rate, reoxygenation, and oxygen solubility. The outcomes obtained in this study are expected to provide the impetus for developing a strategy for the viable improvement of the water environment.

show abstract

A management-oriented water quality model for data scarce catchments

Cited by 39 publications

References 38 publications

Dynamic multi-objective optimisation using deep reinforcement learning: benchmark, algorithm and an application to identify vulnerable zones based on water quality

Dynamic multi-objective optimisation using deep reinforcement learning: benchmark, algorithm and an application to identify vulnerable zones based on water quality

Monitoring Water Quality Parameters of Taihu Lake Based on Remote Sensing Images and LSTM-RNN

Assessment of Water Quality Profile Using Numerical Modeling Approach in Major Climate Classes of Asia

Contact Info

Product

Resources

About