Comprehensive river water quality monitoring using convolutional neural networks and gated recurrent units: A case study along the Vaigai River

• Published in: Journal of environmental management 2024-08, Vol.365, p.121567, Article 121567
• Main Authors: S, Geetha T., G, Gandhimathi, C, Chellaswamy, Selvan P, Thiruvalar
• Format: Article
• Language: English
• Subjects: Environmental Monitoring - methods; Gated recurrent unit; Neural Networks, Computer; River water monitoring; Rivers - chemistry; Water Quality; Water quality parameters; Water resource management
• ISSN: 0301-4797; 1095-8630; 1095-8630
• DOI: 10.1016/j.jenvman.2024.121567

Effective monitoring of river water quality is required for long-term water resource management. Convolutional Neural Networks and Gated Recurrent Unit-based water quality monitoring (CNGRU-WQM) were used in this investigation to develop a comprehensive monitoring system along the Vaigai River. The system was designed to collect real-time data on several crucial water quality parameters. The collected characteristics encompassed factors like water pollution levels, turbidity, pH readings, temperature, and total dissolved solids, offering a comprehensive view of river water quality. The monitoring system was methodically set up, with sensors strategically positioned at various locations along the river. This ensured that data collection would take place at regular intervals. The CNGRU-WQM model achieved a validation accuracy of 97.86%, surpassing the performance of other state-of-the-art approaches. Particularly noteworthy is the fact that the actual use of this system incorporates real-time warnings, which enable stakeholders to be instantly informed if water quality measurements surpass pre-set criteria. The study's contributions include its efficient river water quality monitoring system, which encompasses a variety of indicators, and its ability to significantly affect environmental conservation efforts by offering a helpful tool for informed decision-making and timely interventions. •A hybrid approach has been introduced for monitoring the quality of river water.•Real-time data has been collected from different locations.•Various water quality parameters has been measured and estimated.•Compare the performance of CRU-WQM with other machine learning methods.